Research guide and technical manual
Korn FerryAssessment of Leadership Potential
Korn Ferry Assessment of Leadership PotentialResearch guide and technical manual
© Korn Ferry 2015-2016. All rights reserved.
No part of this work may be copied or transferred to any other expression or form without a license from Korn Ferry.
For the sake of linguistic simplicity in this product, where the masculine form is used, the feminine form should always be understood to be included.
www.kornferry.com
Korn Ferry Assessment of Leadership Potential Research guide and technical manual
Version 15.2a—05/2016
Research guide and technical manual
Korn FerryAssessment of Leadership Potential
Table of contents
Section IIntroduction ................................................................................................................................................................................................. 1
Section IIIntroduction to high potential identification .............................................................................................................................3
Business need............................................................................................................................................................................................3
What is leadership potential? ...........................................................................................................................................................3
Potential and Korn Ferry’s Four Dimensions of Leadership and Talent ......................................................................6
Korn Ferry’s model of leadership potential ...............................................................................................................................8
Section IIIThe Korn Ferry Assessment of Leadership Potential..........................................................................................................10
Overview ....................................................................................................................................................................................................10
Section IVThe Korn Ferry Assessment of Leadership Potential: Measurement approach ......................................................11
Background ................................................................................................................................................................................................11
Section VThe Korn Ferry Assessment of Leadership Potential: What it measures ................................................................. 16
Drivers ......................................................................................................................................................................................................... 16
Experience ................................................................................................................................................................................................ 21
Traits............................................................................................................................................................................................................ 23
Awareness ................................................................................................................................................................................................ 24
Learning agility ...................................................................................................................................................................................... 26
Leadership traits ................................................................................................................................................................................... 27
Capacity ..................................................................................................................................................................................................... 31
Derailment risks ..................................................................................................................................................................................... 32
Section VIThe Korn Ferry Assessment of Leadership Potential: Uses, administration, scoring, and reporting overview ..............................................................................................................................................................................34
Intended uses .........................................................................................................................................................................................34
KFALP technical characteristics and validity .........................................................................................................................36
Norms .........................................................................................................................................................................................................48
Appendix A. Frequently asked questions ....................................................................................................................................50
Appendix B. Inter-method correlations ........................................................................................................................................ 55
Appendix C. Norm descriptions .......................................................................................................................................................56
Appendix D. Global Personality Inventory (GPI) definitions ..............................................................................................68
Appendix E. GPI and KFALP correlations ....................................................................................................................................70
References .................................................................................................................................................................................................... 71
© Korn Ferry 2015–2016. All rights reserved. 1
Section I Introduction
The Korn Ferry Assessment of Leadership Potential (KFALP) is a world-class, science-driven, and comprehensive
self-assessment for measuring leadership potential. The assessment measures an individual’s Drivers, Experience,
Awareness, Learning agility, Leadership traits, Capacity, and Derailment risks. Norms are applied to provide
information related to advancement to more senior leadership levels.
Leadership potential is about what could be at some point in the future, not what is currently. By focusing on
measures related to what could be, the tool has been carefully conceived and empirically designed to provide
critical data about people—data proven to differentiate those who have successfully advanced from those who
have not advanced.
The KFALP is designed to provide data important for individuals and organizations to consider as they think
about leadership potential. It is not designed for selection of individuals into specific jobs today, however, it can
be used as a data point for selection into roles where most hires are expected to further advance into leadership
within a relatively short time.
This manual is designed as a technical reference to help deepen your understanding of the research behind the
KFALP.
You can refer to this manual for a variety of purposes:
• Build your knowledge regarding the research studies on high potential identification.
• Enhance your understanding of the research foundation of the KFALP.
• Review some key findings from psychometric analyses of KFALP data, including sub-group differences.
• Find answers to some frequently asked questions.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved. 3
Section II Introduction to high potential identification
Business needToday, organizations face a unique and unprecedented set of challenges and potential opportunities. The pace of
market change, speed of innovation, global dynamics, and changing demographics generate many opportunities
to both create and extract value, but it is often more difficult to locate those opportunities and act upon them.
Thus, how do companies compete in this increasingly complex and volatile environment? One of the central
differentiators for companies is a strong human capital foundation: the right leaders in the right places.
To succeed in driving business strategy, it is imperative for companies to have a future-focused talent strategy.
Organizations need to develop and sustain a pipeline of the right leaders, with the right abilities, in the right
roles, and at the right times to ensure a sustainable competitive advantage. The idea of identifying and
managing high potential talent has become increasingly essential for organizations.
Most organizations have recognized the need for and have implemented a formal process to identify and assess
high potential talent (Church & Rotolo, 2013; Silzer & Church, 2009). The construct of leadership potential, as
used by many organizations, refers to the possibility that individuals have the qualities (e.g., motivation, skills,
abilities, experiences, and characteristics) to advance in their careers and perform effectively in future roles. It
implies further growth and development to reach some desired end state.
According to several studies, only about one-half of companies report having a high potential identification
program (Howard, 2009; Slan-Jerusalim & Hausdorf, 2007; Wells, 2003). Companies which do have programs
frequently select individuals based on factors not necessarily related to potential, such as personal experience
with the person, performance review ratings, and past performance results (Slan-Jerusalim & Hausdorf, 2007;
Pepermans, Vloeberghs, & Perkisas, 2003). In addition, Martin and Schmidt (2010) indicated that based on
their research on leadership transitions, nearly 40% of internal job moves made by people identified by their
companies as “high potentials” end in failure.
What is leadership potential?When comparing the way in which 13 major organizations defined potential, Karaevli and Hall (2003) found
that every organization had a differing definition of the construct. Silzer and Church (2009) formed several
common potential definitions based on the results of a survey of organizations, finding that 35% defined
potential in terms of moving into a top leadership role, 25% defined it as having the potential to take on broader
responsibilities and leadership duties, 10% viewed potential as having a history of being a high performer, and
25% of organizations surveyed defined potential as successfully moving up two levels from the current level.
With so many different perspectives and definitions of potential in use, it may be most useful to start by defining
what potential is not.
Potential is not the same as current job performance. Current performance is directly visible, but potential is
a prediction about the future. Not all high performers are high potentials. Research suggests that only about
30% of high performers should be classified as high potentials (Corporate Leadership Council, 2005). Although
previous performance can be a useful piece of information in the identification of high potential individuals, it
should not be (though, often is) confused with potential (Corporate Leadership Council, 2005; Slan-Jerusalim &
Hausdorf, 2007; Pepermans, Vloeberghs, & Perkisas, 2003).
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved.4
The distinction between performance and potential is highlighted by the fact that the factors required for
successful performance changes from one organizational level to the next. The Charan, Drotter, and Noel (2011)
six-passage model is often used to describe the leadership requirements throughout the various organizational
levels within a company. This “Pipeline Model of Leadership Development” defines the crucial skills for successful
management transitions from the very bottom of an organization (managing oneself) to the very top (managing
the enterprise). Each of the six management transitions in this model, illustrated in Figure 1, involves a major
change in job requirements, demanding new skills, time applications, and work values.
Figure 1. The changing requirements of leadership.
INDIVIDUALCONTRIBUTOR
FRONT-LINEMANAGER
MANAGER OFMANAGERS
BUSINESSUNIT LEADER
SENIOREXECUTIVE
CHIEFEXECUTIVE
Short–term Long–term
Limited stakeholders Multiple stakeholders
Manage tasks Manage portfolio
Get the job done Maximize shareholder value
Transactional Transformational
TECHNICAL SKILLS LEADERSHIP AND MANAGEMENT SKILLS STRATEGIC BUSINESS ACUMEN
When advancing to leadership positions of greater responsibility, leadership roles increase in their challenge,
breadth, and complexity. As leaders advance, they must reallocate their focus so that they can help others to
perform effectively. They must learn to value the work of leadership and believe that making time for others,
planning, coordinating, and coaching are imperative in their new responsibility. Strong performance at one level
does not necessarily indicate the capability to successfully progress to more complex future leadership roles.
Potential is also not a person’s readiness at present to move into an immediate next-level position or promotion.
Potential is a more long-term concept, and individuals being considered for an immediate next level should be
assessed using Korn Ferry’s readiness assessments instead.
Another important question frequently asked when discussing the identification of high potential individuals is
potential for what? Everyone has potential, but to what end (Silzer & Church, 2009). High potential employees
are typically selected earlier on for a certain level or type of role in the future. Some organizations have sub-
pools of their high potentials for future marketing executives or engineering executives (Dowell, 2010), but over
time both the individual and the role itself is likely to evolve into something a little different, and thus more
general positions or leadership levels tend to be better targets for high potentials rather than very specific
roles. Some of the common job band levels used by organizations include global leaders/senior executives,
mid-management/technical-functional, and high-value individual contributors or HiPros (Silzer & Church, 2009).
The construct of leadership potential, as used by many organizations, refers to the possibility that individuals
have the qualities (e.g., motivation, skills, abilities, experiences, and characteristics) to effectively perform and
advance in their careers. It implies further growth and development to reach some desired end state.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved. 5
We define potential as “the capacity and interest to develop the qualities required for effective performance in significantly more challenging leadership roles.”
As organizations identify high potential employees and invest copious amounts of time and money to develop these
individuals, including providing them with the most challenging assignments, accurate identification is critical.
Several researchers have worked to identify individual attributes that are related to long-term potential
(Corporate Leadership Council, 2005; Lombardo & Eichinger, 2000; McCall, Lombardo, & Morrison, 1988;
Peterson & Erdahl, 2007; Silzer & Church, 2009). Based on an extensive literature review of nine external high
potential models from consulting firms and two corporate surveys, Silzer and Church (2009) identified seven
characteristics that are commonly viewed as indicators of high potential employees:
• Cognitive skills include conceptual or strategic thinking, breadth of thinking, cognitive ability, and dealing with ambiguity.
• Personality variables include interpersonal skills, dominance, stability, resilience, and maturity.
• Learning ability includes adaptability, learning orientation, learning agility, and openness to feedback.
• Leadership skills include developing others, leading and managing others, and influencing and inspiring.
• Motivation variables include energy, engagement, drive for advancement, career drive, interests, career aspirations, results orientation, and risk taking.
• Performance record includes leadership experiences and performance track record.
• Knowledge and values include cultural fit and technical/functional skills and knowledge.
They further boiled these down into three types of dimensions of potential, which could each be measured at
different stages during an individual’s career (Silzer & Church, 2009).
• Foundational dimensions include more stable attributes such as cognitive skills and personality, which could be measured even at very early career stages.
• Growth dimensions involve abilities to learn, as well as motivational variables, which may appear more clearly in certain job contexts over others.
• Career dimensions include the more level/function-specific predictors, such as performance records, leadership skills, and experiences, which should be measured and normed against the different stages during career progression.
Korn Ferry’s model of leadership potential was formed on the basis of previous findings and science and
includes predictors from each of these three broad types of dimensions found in the literature. In Section V
of this technical manual, we define and describe each predictor, summarizing key literature relevant to each
construct. Before turning to this review, we outline how the capabilities measured by the KFALP relate to
broader approaches to describing and understanding leaders’ capabilities.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved.6
Potential and Korn Ferry’s Four Dimensions of Leadership and TalentKorn Ferry organizes the qualities that lead to success into four distinct categories to describe a whole-person
view of individuals: Drivers and Traits—which describe “who you are,” and Experiences and Competencies—
which describe “what you do” (see Figure 2). The four dimensions influence one another and interact within
each person.
Figure 2. Four Dimensions of Leadership and Talent.
ExperiencesCompetencies
Traits Drivers
Skills and behaviors required for success that can be observed.
Inclinations, aptitudes and natural tendencies a person leans toward, including personality traits and intellectual capacity.
Assignments or roles that prepare a person for future roles.
Values and interests that influence a person’s career path, motivation and engagement.
What you do
Who you are
Each dimension plays a distinct role in performance, engagement, potential, and personal career development.
As a measure of potential, the KFALP focuses heavily on Traits and Drivers—who the person is and may become
in general—rather than what the person has done or can do today. The focus is potential to develop for the
future, rather than readiness or fit for a specific job today.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved. 7
Feedback from the KFALP is organized into seven categories, or signposts. The seven signposts are a rational
organizing framework and not an empirical one. They are designed to assist clients in discussion and considering
critical components of potential. How the signposts relate to Korn Ferry’s Four Dimensions of Leadership and
Talent is shown in Figure 3.
Figure 3. Four Dimensions of Leadership and Talent.
ExperiencesCompetencies
Traits Drivers
• Track record of formative experiences.
• Engaged by leadership.• Learns from experience.• Self–awarness.• Leadership dispositions.• Minimal derailment risks.• Aptitude for logic and
reason.
What you do
Who you are
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved.8
Korn Ferry’s model of leadership potentialDefinitions of the attributes assessed by the KFALP are shown in Table 1. These drivers, experiences, and traits
reflect leadership potential and are related to advancement in the leadership pipeline.
Table 1. Model of the Korn Ferry Assessment of Leadership Potential.
DriversMotivation to be a leader – successful leaders are driven by and prefer the challenges and work of leaders.
The Drivers signpost includes three sub-dimensions:Advancement drive: drive to advance through collaboration, ambition, challenge.
Career planning: how narrowly or broadly focused are the participant’s career goals and how specific is his/her career plan.
Role preferences: preference for the work of roles requiring versatility and achievement through others vs. professional mastery and expertise.
ExperienceTrack record of formative experiences – successful leaders have a foundation of experiences that enable the
acquisition of new skills.
The Experience signpost includes three sub-dimensions:Core experience: what a leader has learned in the course of his/her day-to-day leadership career.
Perspective: the diversity of a leader’s experience across organizations, industries, functional areas, and countries.
Key challenges: a leader’s experience with a number of seminal developmental challenges.
AwarenessSelf-knowledge and insight – leaders exhibit awareness of self in the context of their work.
The Awareness signpost includes two sub-dimensions:Self-awareness: the extent to which the leader is aware of his/her strengths and development needs.
Situational self-awareness: the extent to which the leader monitors and is aware of how events impact his/her performance.
Learning agilityAbility to learn from experience – leaders extract lessons from experience and leverage them in novel situations.
The Learning agility signpost includes four sub-dimensions:Mental agility: a leader’s tendency to be inquisitive and approach problems in novel ways.
People agility: a leader’s skill in reading others and applying the insights gained in people-related matters.
Change agility: a leader’s tendency to promote new possibilities and to take ideas from vision to reality.
Results agility: a leader’s propensity to deliver outstanding results in new and tough situations.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved. 9
Leadership traitsDispositions held by leaders – leaders possess traits that make them naturally inclined to lead.
The Leadership traits signpost includes five sub-dimensions:Focus: the balance between attending to details and keeping an eye on the big picture.
Persistence: the passionate pursuit of personally valued long-term goals.
Tolerance of ambiguity: a leader’s capacity to deal effectively with uncertainty or confusing situations.
Assertiveness: the willingness to assume a leader role and comfort with leadership.
Optimism: a leader’s tendency to have a positive outlook.
CapacityAptitude for logic and reasoning.
The Capacity signpost includes one sub-dimension:Problem solving: the ability to spot trends and patterns and draw correct conclusions from confusing or ambiguous data.
Derailment risksManaged derailment risks – leaders manage those configurations of traits which might lead them to derail.
The Derailment risks signpost includes three sub-dimensions:Volatile: a risk toward being mercurial, erratic, or unpredictable.
Micromanaging: a risk toward controlling the work of direct reports.
Closed: a risk toward being closed to alternative perspectives and opportunities.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved.10
Section III The Korn Ferry Assessment of Leadership Potential
OverviewThe Korn Ferry Assessment of Leadership Potential (KFALP) helps organizations assess and identify talent who
have the characteristics needed to develop the competencies and gain the experience to succeed in leadership
roles.
• It gives a complete view of a person’s leadership potential, no matter where they are in the organization.
• It accurately identifies high potentials with seven key signposts of leadership potential proven by research.
• It helps organizations invest in the right talent and target the right areas for development.
• It provides organizations with the ability to objectively and accurately identify people with high leadership potential.
• It gives a complete view of a person’s leadership potential, no matter where they are in the organization.
• It accurately identifies high potentials in seven key facets, or signposts, proven by research to be related to advancement in leadership roles.
The KFALP was built by leveraging expertise from the combined decades of knowledge and the hundreds of
thousands of leadership assessments Korn Ferry has amassed. The framework was developed based on rigorous
analysis using a combination of quantitative, qualitative, and market-based data, sourcing from both Korn Ferry’s
own extensive data stores and external literature review:
• Research analyses based on a variety of Korn Ferry intellectual property, such as Leadership Experience Inventory (LEI) and Korn Ferry Four Dimensional Executive Assessments, TalentView® of Leadership Transitions, and viaEDGE®.
• Extensive review of the scientific literature on high potential identification and leadership pipeline.
• Expert input representing decades of work in leadership research and development.
• Customer input.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved. 11
Section IV The Korn Ferry Assessment of Leadership Potential: Measurement approach
The KFALP uses state-of-the-art measurement technology in this assessment: Forced-Choice Item Response
Theory (Brown, 2016; Brown & Maydeu-Olivares, 2011). Forced-Choice Item Response Theory (FC-IRT) offers a
number of advantages over more traditional approaches to measurement, such as Likert-type response scales or
Forced-Choice measures grounded in Classical Test Theory (CTT).
Background
Response distortion and fakingOne of the challenges in the measurement of the characteristics of potential by self-assessment is controlling
intentional response distortion—faking. To the extent possible, it is very important to capture data that reflects
a participant’s true standing on characteristics related to advancement. Psychometricians have become
increasingly concerned with the response distortions associated with the Likert-style response formats
commonly used in psychological measurement (Stark, Chernyshenko, Chan, Lee, & Drasgow, 2001). As use of
self-assessments of personality has increased, respondents’ desire to present themselves in a positive light
to gain advantage has become of greater concern. The growing availability of self-coaching materials and
use of unproctored Internet-based tests has further contributed to the potential for faking to be increasingly
problematic (Sliter & Christiansen, 2012). Past approaches to handling response distortion have not been highly
satisfactory.
Social desirability and/or “faking scales” have long been deployed in order to detect and address faking.
However, this approach suffers from limitations. When faking is indicated, it is difficult to know how to proceed.
In research settings, a completed assessment may simply be thrown out. In applied settings, coaches and
decision makers may simply be warned that the results are perhaps untrustworthy and to proceed with caution.
Yet others have attempted to use results from faking detection or social desirability scales to adjust observed
scores in diverse ways (Goffin & Christiansen, 2003). Such methods, however, have been repeatedly criticized
as being arbitrary and difficult to validate (McCrae & Costa, 1983; Goffin & Christiansen, 2003). Sacket (2011),
summarizing the observations contained in an edited book on faking in CTT, concludes, “In sum, we are not yet
at the point at which methods of identifying individuals engaging in faking are available for use in operational
settings,” and “So, without a reliable and valid detection method, scores cannot be adjusted to remove the
effects of faking.”
Forced-choice response formats also have been developed and employed to combat faking. These formats
force respondents to endorse or choose one item over another. When scored traditionally, these formats make
extreme high or extreme low endorsement of every item impossible. In addition to combating faking (Sackett
& Lievins, 2008), forced-choice measurement can markedly reduce response bias, “halo” or leniency effects,
and response variance attributable to individual response styles not immediately associated with item content
(Bartram, 2007; Cheung & Chan, 2002). However, when used in combination with classical test theory scoring
methods, traditional forced choice response formats always produce scale scores which are known as ipsative,
that are problematically auto-correlated and interdependent (Brown & Maydeu-Olivares, 2011). That is, an
individual’s scores on one scale is, in very large part, a direct and artificial reflection of person-scores on the
other scales included in the measure (Heggestad, Morrison, Reeve, & McCloy, 2006).
This dependency makes normative comparisons across individuals improper and violates the assumptions of
many common psychometric statistics (Blinkhorn, Johnson, & Wood, 1988; Hammond & Barrett, 1996; Hough &
Ones, 2001; Meade, 2004).
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved.12
Forced-choice IRT modelsResearchers in psychological measurement have sought to tackle the many problems associated with forced-
choice measures. Stark, Chernyshenko, and Drasgow (2005) developed a pairwise preference ideal point model
that addresses most related problems by pairing and presenting items with similar levels of social desirability
and by employing scoring and parameter estimation methods that are shown to perform well under certain
conditions vis-à-vis eliminating ipsative auto-correlation. Stark and Chernyshenko (2007) point out that number
of pairwise preference ratings needed to obtain reasonable person-score standard errors using this approach
may be particularly high in non-adaptive testing situations. Hence, the Stark et al. (2005) pairwise preference
model works and is markedly more efficient with computer assisted adaptive testing administration, wherein
item presentation is customized according to real-time response patterns. Where fixed form administration
is optimal or necessary, test administration and reliability using the Stark et al. (2005) model may require
many more items than desirable, limiting its feasibility in applied organizational settings. Additional limitations
associated with this model’s reliance on an ideal point measurement framework include the relative difficulty of
writing items, the lack of invariance in parameter estimates and model fit when reversing item coding, and the
apparent reduced accuracy of item parameter estimation (Brown & Maydeu-Olivares, 2011; Maydeu-Olivares,
Hernandez, & McDonald, 2006).
As an alternative, Brown and Maydeu-Olivares (2011) developed a structured multidimensional forced-choice IRT
model that addresses problems associated with faking, response bias, and ipsativity while also addressing some
of the limitations of the paired preference Stark et al. (2005) model. The authors describe a model that is linear
in differences between latent traits.
Briefly, and in very straightforward terms, we can now simultaneously overcome the limitations and maintain the
benefits of forced-choice measurement by employing recent advances in Item Response Theory (IRT) modeling
that takes advantage of the item format and produces scale scores that are truly normative. This model is the
FC-IRT model (see Brown & Maydeu-Olivares, 2011).
The method uses large samples of people and statistical models to identify 1) item loadings, which indicate
the degree to which each item is related to the underlying trait it is intended to measure, and 2) item difficulty,
which relates to the likelihood of a person with a given level of a trait ranking the item as more like them vs.
other items. These two pieces of information are used, along with the rankings a person provides, to generate
scores. A more detailed and statistical explication of the methodology can be found in subsequent paragraphs
and sections.
The latent traits are manifest by binary comparisons of items that are presented in the forced-choice blocks.
The model rearranges forced-choice responses into a series of exhaustive binary comparisons, thereby allowing
for components of non-ipsative trait measures to drive parameter estimation, scoring, and interpretation of
person-scores. The model is novel in that it creates a relative independence among otherwise predictably auto-
correlated forced-choice based construct scores. It is flexible in terms of forced-choice block sizes and is feasible
in that parameters and scores can be estimated using existing popular statistical software packages, including
Mplus (Muthen & Muthen, 2010). We also have developed a related R package (Zes, 2015; Zes, Lewis, & Landis,
2014) that similarly estimates the Brown and Maydeu-Olivares (2011) model and related extensions of it.
FC-IRT uses a two-stage approach. If a person prefers or ranks item 1 higher than item 2, then that person’s
preference for item 1 is higher than item 2. For instance, in a block of four items, there are six possible
comparisons to make between the items. After a series of blocks are administered, we can use the information
provided by these comparisons to fit an IRT model. We first use the IRT model to obtain information about
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved. 13
how difficult the items are to endorse (called item difficulty), and how well the items relate to the traits being
measured (called item loadings). Then, using item difficulty and item loading data, we can accurately estimate
a person’s normative scale score for each of the underlying traits measured by the assessment. By abstracting
the ranking information into two separate yet linked statistical models, this two-stage approach allows us to take
forced-choice item responses and produce normative scale scores free of the ipsative limitations of traditional
forced-choice measurement.
Using Forced-Choice Item Response Theory, we can take advantage of the bias and faking resistance of the
forced-choice item response format while eliminating the psychometric limitations of the classical test theory
scoring methods.
FC-IRT in KFALPIn brief, the use of FC-IRT (Brown, 2016; Brown & Maydeu-Olivares, 2011) in KFALP provides a methodology with
the following advantages:
• Removes response styles and scale anchor issues by eliminating the Likert rating scale.
• Thwarts faking by eliminating the transparency of the Likert rating scale.
• Frees interpretation from specific scale content by estimating true trait level.
• Provides superior true score estimates of error.
• Takes item content into consideration when estimating true score, thus improving fidelity of the score to the trait construct.
Many participants report the forced-choice response format to be more engaging and challenging than typical
single-item rating assessments—some even describe it as difficult. This is as intended. Most understand that this
is a characteristic of an assessment designed to be more incisive. It takes unlimited control of responses away
from the participant and requires hard choices. It is this feature that inhibits faking and improves the quality of
the data.
FC-IRT measurement technology is applied to the Awareness, Learning agility, Leadership traits, and Derailment
risks signposts and to part of the Drivers signpost. This state-of-the-art technology allows true normative
comparison of individuals while effectively controlling error resulting from idiosyncratic use of a response scale
or from intentional faking (Brown & Maydeu-Olivares, 2011; Drasgow, Chernyshenko, & Stark, 2010).
Item administrationThe KFALP traits and drivers scored with FC-IRT are administered in forced-choice response format. Each
construct type is grouped together in its own test form. Traits are measured with traits in blocks of four items,
and drivers with drivers in blocks of six items. Construct scores are estimated using a modification (Zes, 2015;
Zes et al., 2014) of the Brown and Maydeu-Olivares (2011) Forced-Choice Item Response Theory (FC-IRT) model
to arrive at construct estimates whose correlations are based on the nature of the constructs and not according
to forced-choice item response format artifacts.1 Eight items were designed to tap each trait, and trait response
blocks contain four items each. Each driver is measured using ten items, with items presented in response blocks
of six items.
1 In early developmental efforts, we administered and scored forced-choice based trait scales and Likert-based trait scales of the same items and constructs to the same individuals and found, much as Brown and Maydeu-Olivares (2011) did, that alternate-form correlations between the same constructs typically had magnitudes consistent with most conceptualizations of alternate test form construct convergence (e.g., r > .70 in every case). The full correlations are shown in Table 17 in Appendix B.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved.14
An example of a forced-choice multi-item block from the Drivers test form is shown in Table 2. This example
illustrates that each response block is comprised of items measuring multiple scales within the domain. That is,
for each trait or driver response block, there is no more than one item from each scale.
Table 2. Example six-item block.
Item 1 Well-defined work objectives.
Item 2 Situations without a winner and a loser.
Item 3 Having high status within the organization.
Item 4 Avoiding meetings so I can focus on my work.
Item 5 Developing myself beyond work.
Item 6 Consistent direction in my career.
Upon seeing a block of items, candidates are tasked with ranking the items from “Most” to “Least” on some
continuum. Specifically, in this example, candidates would be asked to rank the items from “Most preferred” to
“Least preferred.”2
ScoringTo set the stage for the FC-IRT model, assume that we have a test composed of several six-item blocks (as
with the Drivers test form, which has 10 six-item blocks) where each item in a given block measures a unique
construct or dimension (much like the example in Table 2). Further, assume that candidates are asked to rank the
items from “Most preferred” to “Least preferred.” To model this setup using FC-IRT, we first employ a Thurstonian
Comparative Model (Brown & Maydeu-Olivares, 2011; for the origin of this model, see Thurstone, 1927). Using this
model, for a given block of six items there are six latent utilities/thresholds, ti. If a candidate prefers or ranks item
i larger than item j, then the utility for item i, ti is larger than the utility for item j, tj. This information can be coded
in a comparative task as
Forced Choice Item Response Theory
To assess a candidate’s standing on scales from each of the three dimensions (Drivers, Competencies,
Traits), items are presented in multi-item blocks as shown below.
Item 1 Well defined work objectivesItem 2 Situations without a winner and a loser.Item 3 Having high status within the organization.Item 4 Avoiding meetings so I can focus on my work.Item 5 Developing myself beyond work.Item 6 Consistent direction in my career.
Figure 1: Block of Six Items
Upon seeing a block of items, candidates are tasked with ranking the items from “Most” to “Least” on some
continuum. Specifically, in this example, candidates would be asked to rank the items from “Most Preferred”
to “Least Preferred”1. Traditionally—that is, in Classical Test Theory (CTT)—information gathered in this
way induces a dependence among the items and scales known as ipsativity (Brown & Maydeu-Olivares, 2011).
One attractive feature of ipsative measurement is that response biases and “halo” effects are diminished
(Bartram, 2007; Cheung & Chan, 2002). However, due to the inherent item dependence, ipsative scales
are also known to produce problems for “score interpretation and for almost every conventional type of
psychometric analysis” (Brown & Maydeu-Olivares, 2011, p. 461; see also Baron, 1996). We can overcome
this limitation of ipsative measurement and still reap the benefits by employing a Forced Choice Item
Response Theory (FCIRT) method of modeling the items and scale relationships (Brown, 2010; Brown &
Maydeu-Olivares, 2011, 2012, 2013; Maydeu-Olivares & Brown, 2010).
To set the stage for the FCIRT model, assume that we have a test composed of several six item blocks
where each item in a given block measures a unique construct or dimension (much like the example in Figure
1). Further, assume that candidates are asked to rank the items from “Most Preferred” to “Least Preferred”.
To model this setup using FCIRT, we first employ a Thurstonian Comparative Model (Brown & Maydeu-
Olivares; for the origin of this model, see Thurstone, 1927). Using this model, for a give block of six items,
there are six latent utilities, ti. If a candidate prefers or ranks item i larger than item j, then the utility for
item i, ti is larger than the utility for item j, tj . This information can be coded in a comparative task as
yl =
1 if ti ≥ tj
0 if ti < tj
. (1)
1Note - for the Competency and Trait Dimensions, candidates are asked to rank blocks of items from “Most Like Me” to“Least Like Me”.
1
Using this coding, if the latent utility for item i is larger than the latent utility for item j, the observed response,
yl is represented by 1 (i.e., yl = 1 denotes that item i is ranked higher than item j). Then, for a block of six items,
there are fifteen possible comparative tasks. With this setup, we can model the comparative tasks a latent factor
model. To do so, we first note that the observed response, yl, is dependent on a difference of two item utilities.
This difference can be represented as a latent comparative response, yl* = ti - tj, such that
Using this coding, if the latent utility for item i is larger than the latent utility for item j, the observed
response, yl is represented by 1 (i.e., yl = 1 denotes that item i is ranked higher than item j). Then, for a
block of six items, there are 15 possible comparative tasks. With this setup, we can model the comparative
tasks a latent factor model. To do so, we first note that the observed response, yl, is dependent on a difference
of two item utilities. This difference can be represented as a latent comparative response, y∗l = ti − tj , such
that
yl =
1 if y∗l ≥ 0
0 if y∗l < 0. (2)
Because we are assuming that the items measure a latent construct we can model each item’s utility as a
linear function of the underlying latent construct as
ti = µi + λiηa + εi, (3)
where µidenotes the mean of the latent utility, λi denotes a factor loading, and ηa denotes a common
latent factor underlying the utility ti, and ei denotes a unique factor. Moreover, we assume that each
item measures one and only one latent trait, the common and unique latent constructs are orthogonal and
normally distributed, and that unique factors across items are orthogonal.
Notice from (??)–(??) that we have a nested latent structure. Specifically, we have modeled each observed
binary response as being dependent on a latent comparative response, which in turn is dependent on a
linear combination of an underlying latent trait. This nested latent structure is typically referred to as a
second-order factor model. As is well known (Takane & De Leeuw, 1987), many IRT models are equivalent
to factor models of dichotomous variables. As shown by Brown and Maydeu-Olivares (2012), through
reparameterization, we can recast the second-order factor model as a first-order Thurstonian IRT model.
To reparameterize the model, we rewrite each latent comparative response as
y∗l = (ui − uj) + λiηa − λjηb + ei − ej . (4)
If we assume for the moment that the two traits are known, or conditioned on, and recalling the assumption
that the traits and unique factors are normally distributed, then the item characteristic function for preferring
item i over item j for a person can be written as
P (yl = 1|ηa, ηb) = NCDF
(ui − uj) + λiηa − λjηb√
Ψ2i +Ψ2
j
, (5)
2
2 For the trait dimensions, candidates are asked to rank blocks of items from “Most like me” to “Least like me.”
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved. 15
Because we are assuming that the items measure a latent construct, we can model each item’s utility as a linear
function of the underlying latent construct as
Using this coding, if the latent utility for item i is larger than the latent utility for item j, the observed
response, yl is represented by 1 (i.e., yl = 1 denotes that item i is ranked higher than item j). Then, for a
block of six items, there are 15 possible comparative tasks. With this setup, we can model the comparative
tasks a latent factor model. To do so, we first note that the observed response, yl, is dependent on a difference
of two item utilities. This difference can be represented as a latent comparative response, y∗l = ti − tj , such
that
yl =
1 if y∗l ≥ 0
0 if y∗l < 0. (2)
Because we are assuming that the items measure a latent construct we can model each item’s utility as a
linear function of the underlying latent construct as
ti = µi + λiηa + εi, (3)
where µidenotes the mean of the latent utility, λi denotes a factor loading, and ηa denotes a common
latent factor underlying the utility ti, and ei denotes a unique factor. Moreover, we assume that each
item measures one and only one latent trait, the common and unique latent constructs are orthogonal and
normally distributed, and that unique factors across items are orthogonal.
Notice from (??)–(??) that we have a nested latent structure. Specifically, we have modeled each observed
binary response as being dependent on a latent comparative response, which in turn is dependent on a
linear combination of an underlying latent trait. This nested latent structure is typically referred to as a
second-order factor model. As is well known (Takane & De Leeuw, 1987), many IRT models are equivalent
to factor models of dichotomous variables. As shown by Brown and Maydeu-Olivares (2012), through
reparameterization, we can recast the second-order factor model as a first-order Thurstonian IRT model.
To reparameterize the model, we rewrite each latent comparative response as
y∗l = (ui − uj) + λiηa − λjηb + ei − ej . (4)
If we assume for the moment that the two traits are known, or conditioned on, and recalling the assumption
that the traits and unique factors are normally distributed, then the item characteristic function for preferring
item i over item j for a person can be written as
P (yl = 1|ηa, ηb) = NCDF
(ui − uj) + λiηa − λjηb√
Ψ2i +Ψ2
j
, (5)
2
where μi denotes the mean of the latent utility, λi denotes a factor loading/discrimination, ηa denotes a common
latent factor underlying the utility ti, and εi denotes a unique factor. Moreover, we assume that each item
measures one and only one latent trait, that the common and unique latent constructs are orthogonal and
normally distributed, and that unique factors across items are orthogonal.
Notice from (1), (2), and (3) that we have a nested latent structure. Specifically, we have modeled each observed
binary response as being dependent on a latent comparative response, which, in turn, is dependent on a linear
combination of an underlying latent trait. This nested latent structure is typically referred to as a second-order
factor model. As is well known (Takane & De Leeuw, 1987), many IRT models are equivalent to factor models of
dichotomous variables. As shown by Brown and Maydeu-Olivares (2012), we can recast the second-order factor
model as a first-order Thurstonian IRT model via reparameterization.
To reparameterize the model, we rewrite each latent comparative response as
Using this coding, if the latent utility for item i is larger than the latent utility for item j, the observed
response, yl is represented by 1 (i.e., yl = 1 denotes that item i is ranked higher than item j). Then, for a
block of six items, there are 15 possible comparative tasks. With this setup, we can model the comparative
tasks a latent factor model. To do so, we first note that the observed response, yl, is dependent on a difference
of two item utilities. This difference can be represented as a latent comparative response, y∗l = ti − tj , such
that
yl =
1 if y∗l ≥ 0
0 if y∗l < 0. (2)
Because we are assuming that the items measure a latent construct we can model each item’s utility as a
linear function of the underlying latent construct as
ti = µi + λiηa + εi, (3)
where µidenotes the mean of the latent utility, λi denotes a factor loading, and ηa denotes a common
latent factor underlying the utility ti, and ei denotes a unique factor. Moreover, we assume that each
item measures one and only one latent trait, the common and unique latent constructs are orthogonal and
normally distributed, and that unique factors across items are orthogonal.
Notice from (??)–(??) that we have a nested latent structure. Specifically, we have modeled each observed
binary response as being dependent on a latent comparative response, which in turn is dependent on a
linear combination of an underlying latent trait. This nested latent structure is typically referred to as a
second-order factor model. As is well known (Takane & De Leeuw, 1987), many IRT models are equivalent
to factor models of dichotomous variables. As shown by Brown and Maydeu-Olivares (2012), through
reparameterization, we can recast the second-order factor model as a first-order Thurstonian IRT model.
To reparameterize the model, we rewrite each latent comparative response as
y∗l = (ui − uj) + λiηa − λjηb + ei − ej . (4)
If we assume for the moment that the two traits are known, or conditioned on, and recalling the assumption
that the traits and unique factors are normally distributed, then the item characteristic function for preferring
item i over item j for a person can be written as
P (yl = 1|ηa, ηb) = NCDF
(ui − uj) + λiηa − λjηb√
Ψ2i +Ψ2
j
, (5)
2
If we assume for the moment that the two traits are known, or conditioned on, and recalling the assumption that
the traits and unique factors are normally distributed, then the item characteristic function for preferring item i
over item j for a person can be written as
Using this coding, if the latent utility for item i is larger than the latent utility for item j, the observed
response, yl is represented by 1 (i.e., yl = 1 denotes that item i is ranked higher than item j). Then, for a
block of six items, there are 15 possible comparative tasks. With this setup, we can model the comparative
tasks a latent factor model. To do so, we first note that the observed response, yl, is dependent on a difference
of two item utilities. This difference can be represented as a latent comparative response, y∗l = ti − tj , such
that
yl =
1 if y∗l ≥ 0
0 if y∗l < 0. (2)
Because we are assuming that the items measure a latent construct we can model each item’s utility as a
linear function of the underlying latent construct as
ti = µi + λiηa + εi, (3)
where µidenotes the mean of the latent utility, λi denotes a factor loading, and ηa denotes a common
latent factor underlying the utility ti, and ei denotes a unique factor. Moreover, we assume that each
item measures one and only one latent trait, the common and unique latent constructs are orthogonal and
normally distributed, and that unique factors across items are orthogonal.
Notice from (??)–(??) that we have a nested latent structure. Specifically, we have modeled each observed
binary response as being dependent on a latent comparative response, which in turn is dependent on a
linear combination of an underlying latent trait. This nested latent structure is typically referred to as a
second-order factor model. As is well known (Takane & De Leeuw, 1987), many IRT models are equivalent
to factor models of dichotomous variables. As shown by Brown and Maydeu-Olivares (2012), through
reparameterization, we can recast the second-order factor model as a first-order Thurstonian IRT model.
To reparameterize the model, we rewrite each latent comparative response as
y∗l = (ui − uj) + λiηa − λjηb + ei − ej . (4)
If we assume for the moment that the two traits are known, or conditioned on, and recalling the assumption
that the traits and unique factors are normally distributed, then the item characteristic function for preferring
item i over item j for a person can be written as
P (yl = 1|ηa, ηb) = NCDF
(ui − uj) + λiηa − λjηb√
Ψ2i +Ψ2
j
, (5)
2where ψi² is the variance of item i uniqueness. Notice that the function is a standard normal ogive, which in this
case is an IRT model that is dependent on two latent traits. Using this setup, the observed ipsative measurement
model is transformed, loses its ipsative characteristics, and becomes a normative latent IRT model.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved.16
Section V The Korn Ferry Assessment of Leadership Potential: What it measures
As mentioned previously, the KFALP measures Drivers, Experiences, and Traits that reflect leadership potential.
Feedback from the KFALP is organized into seven categories. All of the characteristics assessed by the KFALP
are related to advancement in the leadership pipeline. Relevant literature is summarized in this section of the
manual to provide deeper descriptions of each of these characteristics.
DriversDrivers are the preferences, values, and motivations that influence career aspirations. They provide the “will do”
that creates engagement and energy for a task or role. To the extent that a person’s drivers are aligned with the
role, they will be energized by it.
People with leadership potential find the role of a leader interesting and the work of leading motivating, which is
crucial to being effective. Leadership becomes progressively more difficult at every level, and the demands upon
time and energy increase. If the work doesn’t align to what drives them, it is unlikely that any leader will have
the energy and resilience needed to thrive or even to just survive. According to Silzer and Church (2010), 90% of
organizations now use an individual’s career drive as one predictor to identify high potential.
High potential leaders value the nature of leadership work, the opportunity to make a difference, having a
positive impact on their coworkers and organization, and having greater responsibility. This is evident in the
greater prevalence of goals and aspirations related to leadership at each career level.
Data collected over the past decade at Korn Ferry show that those who move up in leadership are marked by
having higher career aspirations, more specific career goals, a desire to take on general management and C-suite
positions, and are engaged by getting things done through others (see Table 3).
Table 3. Signals of leadership drive across management levels.
Percent choosing in the top three motivators:
First level leader Mid-level leaderFunctional or
business unit leader Senior/top
Influence on the direction of the organization. 38% 52% 61% 72%
Belief in the mission of the organization. 41% 41% 47% 56%
Responsibility for the performance of others and the results of the unit.
30% 42% 48% 49%
Source: Over 17,000 leaders, Career History Questionnaire (Gerstner, Hazucha, & Davies, 2012).
With KFALP, Drivers are measured in three facets, Advancement drive, Career planning, and Role preferences.
We will discuss each in turn.
Advancement driveAdvancement drive involves the motivation to advance through collaboration, ambition, and challenge.
Individuals high on Advancement drive are more willing to put time and energy into achieving greater
responsibility and power. Those low on Advancement drive are less motivated by the challenges of advancing
along the leadership pipeline.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved. 17
BackgroundIn general, motivation has been a central focus of organizational research for many years. The high level of
interest can be attributed to the long-held belief that work behavior is influenced by a mix of factors, including
ability, motivation, and situational constraints/facilitators (Campbell & Pritchard, 1976). In human resource
management, this is referred to as the AMO framework. In essence, the AMO framework proposes that employee
performance (P) is a function of the employee’s ability (A), motivation (M), and opportunity (O) to perform
(Boselie, Dietz, & Boon, 2005; Boxall & Purcell, 2008).
Work motivation is a set of forces that interact with the situation to initiate work-related behavior and to
determine its direction, intensity, and duration. This definition highlights the fact that motivation can be seen
in the choices individuals make among goals to pursue (i.e., direction), the amount of effort they put forth
toward attaining the goals (i.e., intensity), and persistence of action (i.e., duration). In the workplace, notable
achievements rarely are the outcome of random activities. Rather, they typically involve a combination of choices
concerning what to do, how much attentional effort to devote to specific activities, and when to shift direction
and levels of effort.
Numerous theories have been proposed to describe, explain, and investigate motivation. A particularly notable
framework that integrates different motivation theories involves Kanfer’s (1990) distinction between distal
and proximal motivations. Motivations that have immediate and direct impact on behaviors are proximal. For
instance, goal setting can focus attention toward goal-relevant activities and away from irrelevant activities
(Locke, 1978). When individuals are committed, goals will energize individuals and initiate the execution of
action plans toward attaining the goals. Goal commitment, therefore, is a proximal motivation. In contrast,
distal motivations affect action goals through proximal motivations. The impact of distal motivations tends to
span longer time frames and across situations. For example, needs represent a set of distal motivations. The
same need can be satisfied through the pursuit of different goals. Since the assessment of leadership potential
involves predicting relatively enduring behavioral patterns, we focus on measuring distal motivation.
There are different approaches to conceptualizing and measuring motivations. Needs are variable internal
states that, when activated or aroused, energize and direct behavior (Pittman & Zeigler, 2007). A need affects
behavior when there is a discrepancy between one’s current state and a desired state. The discrepancy leads
to the experience of an internal tension that energizes behavior, leading individuals to pursue things in their
environment that can help reduce the discrepancy. Although it is not always well supported, Maslow’s (1954)
need hierarchy is perhaps the most well-known needs theory. In contrast, values are standards or criteria for
selecting among alternatives. They serve as the base for making choices. Values underlie and affect attitudes,
which in turn underlie and affect behavior. To consider values in the workplace is to probe the very reasons
people work and why they behave in the ways they do in their jobs. A value is an enduring belief that a specific
mode of conduct or end state of existence is personally and socially preferable to alternative modes of conduct
or end states (Rokeach, 1973). Therefore, values entail attention to both means (how to do) and ends (what to
pursue). For instance, two individuals may both have a desire to influence others. However, one may choose
to rely on formal power, the other may take a participative or deferential approach. This implies the difference
between needs and values. Whereas needs are considered to be at least partially biologically based, values are
shaped to a larger extent by social factors such as perceived relative status and also by culture.
Needs and values, nonetheless, are closely related. Values represent the expression of needs. When an
individual has a strong need for something, the individual places high value on situations that enable them to
satisfy this need. As such, needs and values tend to be used interchangeably in the work motivation literature
(Kooij, De Lange, Jansen, Kanfer, & Dikkers, 2011). This is revealed by the fact that measures of needs and values
often contain the same test items. For this reason, we reviewed both lines of research to inform the architecture
of our framework with the purpose of establishing a taxonomy of drivers that sufficiently synthesizes existing
theories of needs and values.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved.18
Advancement drive composite detailsFour Drivers comprise the KFALP Advancement drive composite. These are Collaboration, Structure, Power, and
Challenge.
Broadly, Collaboration refers to communion striving. It describes actions directed toward obtaining acceptance
in personal relationships and getting along with others. Socioanalytic theorists (e.g., Hogan & Warrenfeltz, 2003)
have argued that people have innate biological needs for acceptance and approval. Being connected to others,
feeling a sense of relatedness, and desire for interpersonal attachment is a fundamental human motivation
(Baumeister & Leary, 1995). Collaboration is associated with conformity and conflict avoidance, which are not
generally typical of leaders. In fact, in one longitudinal study, McClelland and Boyatzis (1982) found that the need
for affiliation was negatively related to promotion and managerial level. In today’s organizations, however, the
pace of technological change, increased complexity, competitive demands, challenging economics, and risks
involved in decision making have made it difficult for individuals to act alone. Leadership research increasingly
emphasizes collaborative approaches to leadership effectiveness (Yammarino, Salas, Serban, Shirreffs, & Shuffler,
2012). Some scholars even suggest that leaders develop and adopt “collective identities,” which involve self-
definitions based on group membership (Venus, Mao, Lanaj, & Johnson, 2012). Highly collaborative leaders
are motivated by internalizing group values and norms, fulfilling social roles and obligations, and contributing
to the group’s welfare. This typically cultivates trust among team members, which in turn results in increased
team performance (Drescher, Korsgaard, Welpe, Picot, & Wigand, 2014). Collaborative leadership is increasingly
characterized as key for innovation management. In our own data (e.g., D’Mello, 2015), we have repeatedly found
Collaboration (albeit characterized as a behavior more than a motive) to be one of the most salient predictors of
innovation and related outcomes.
Structure refers to preference for work-related stability, routine, certainty, and predictability. Individuals closely
associate certainty with comfort and safety and, by nature, often prefer (fore)knowledge concerning “what
will happen next.” Certainty and predictability facilitate control and personal agency and, where rewarding,
will reinforce and stabilize behavior. Meeting and reaping rewards according to known and clear expectations
generate even physiologically measurable outcomes, including dopamine levels in the brain, which are typically
desirable (Schultz, 1999). In contrast, when patterns do not play out according to expectation, or when if-then
reinforcement schedules are erratic, people tend to sense instability and threats to well-being.
Nevertheless, contemporary organizational design emphasizes agility and adaptability, and increasingly rewards
cross-functional efforts, related synergies, and comfort with ambiguity (Worley & Lawler, 2010). In a 2009 survey,
90% of executives, spanning all regions and industry sectors, ranked organizational agility and adaptability
as crucial to business success and survival (Sull, 2009). Clearly, individuals who value routine, security, and
order are more resistant to and disconcerted by change (Oreg, Vakola, & Armenakis, 2011). For these and other
reasons, high scores on Structure-like measures are increasingly associated with decreased success, particularly
among high-level business executives. High Structure managers and leaders, however, will likely continue to
thrive and be preferable in certain roles and contexts, particularly those characterized by strict regulations,
well-defined processes, and where the effects of not being precise, correct, and thorough are negative and
relatively serious.
A drive for Power involves a strong desire to influence others. Individuals driven by power enjoy being held
responsible for other people and broader group results. They aspire to achieve higher status and even a
prestigious title or rank. They are energized by visibility and strive to gain rewards and recognition for their
efforts. Motivation for power is arguably among the most critical for leadership success. The essence of
leadership itself is embodied in the act of influencing others, and a weak drive for power means a lack of interest
in influence and impacting others (McClelland, 1965; McClelland & Burnham, 1976). In Winter’s (1987) study of
US presidents, power motivation was significantly correlated with historian ratings of “greatness.” The same
power motivation scores have also been linked to ratings of certain aspects of presidential performance, as
well as charisma (House, Spangler, & Woycke, 1991). After reviewing the literature, Zaccaro (2001) cited power
motivation as a key and incremental predictor of leadership charisma.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved. 19
Individuals driven by Challenge prefer new and difficult projects that stretch their abilities. They tend to thrive
on learning and pushing their limits to acquire new proficiencies. They are excited by the prospect of making
a difference and are typically willing and eager to put forth discretionary effort in pursuit of accomplishing
goals. Individuals who are high on Challenge are also typically driven by competition and the desire to win.
Meta-analytic evidence links Challenge-like ratings to a variety of outcomes including income, job performance,
community leadership, and sales success (Spangler, 1992). In their meta-analysis, Collins, Hanges, and Locke
(2004) found that Challenge was related to choosing an entrepreneurial career as well as entrepreneurial
performance. Some previous research has suggested that Challenge is more strongly linked to leadership
success at lower levels of leadership, where the contributions and accomplishments of individuals are seen as
more important than influence over others (McClelland & Boyatzis, 1982). Studies of higher-level managers have
presented mixed results. House, Spangler, and Woycke (1991) and Deluga (1998), for example, found negative
or zero relationships between Challenge and presidential performance and greatness. In contrast, Zaccaro,
White, and colleagues (1997) found that Challenge was positively linked to senior leadership-potential ratings,
career achievement, and organizational level in a sample of army civilian managers. Taken together, this body of
research suggests Challenge is an important indicator of potential, with positive relationships to leaders’ success
at most, if not all, levels of leadership.
Career planningIndividuals who are high on Career planning have strong career preferences and have specific plans on how to
achieve their career goals. Those who are low in Career planning tend to be more opportunistic in how they
approach their career.
Career planning is the only descriptive, non-normed measure in the instrument. It reflects back participant
responses in a fairly straightforward fashion. There is not a demonstrated relationship between Career planning
and leadership advancement or success. We do know, however, that some find it useful to consider future career
moves, while others prefer to be more opportunistic.
A career plan can be useful in laying out a path to gaining experiences that are demonstrated to relate to
advancement. (See Experience below).
Role preferencesIt has long been recognized that not all individuals are interested in becoming managers and leaders. While
some employees value advancement above all else, others value primarily the intrinsic excitement of work.
In a survey of engineers conducted in early 1980s (Guterl, 1984), approximately one-third of the respondents
indicated a preference for a management career, one-third for a non-management career, and one-third were
unsure. This study reported that many engineers preferred to continue doing engineering and to avoid taking
on administrative responsibilities. In another survey of close to 1,500 individuals employed in R&D, less than
one-third of the survey participants preferred the managerial career over alternative career paths (Allen & Katz,
1986). These earlier survey findings are consistent with a more recent survey which found that only about one-
third (34%) of workers aspire to leadership positions (CareerBuilder, 2014).
Schein (1987) concluded that individuals hold a wide variety of career interests, which he labeled career anchors,
or orientations. These career orientations influence career choices, employer changes, other career-related
decisions, and ultimately how the future of career is viewed. At a high level, these differences describe the
contradiction between specialists and generalists in organizations (Gouldner, 1957; Kornhauser, 1962; Schein,
1987). Specialists are characterized by strong professional identification, craftsman-like view of skill mastery,
and a singular area of expertise. Conversely, generalists are identified by a drive to ascend the managerial
hierarchy and breadth of functional-area expertise. Given these differences, organizations were recommended
to subdivide their high potential pools and provide different approaches to manage and engage specialists and
generalists (Cesare & Thornton, 1993; Dries & Pepermans, 2008).
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved.20
Therefore, we systematically conducted research to obtain a more thorough understanding of the motivations,
interests, and values of different types of talented employees in organizations. We were especially interested
in understanding the difference between high potential leaders and high potential professional employees. Our
research effort consists of two phases of investigation—one qualitative and one quantitative.
The qualitative researchThe research started with interviewing 31 high-performing, specialized employees recruited from several
organizations. They were nominated by HR professionals of the participants’ respective companies. Participants
were represented by a wealth of different industries and professions, including engineers, scientists, physicians,
financial analysts, musicians, and clinical psychologists. Respondents were asked a series of questions relating
to their role preferences, events that shaped their career decisions, things they like and don’t like at work,
challenging experiences and the lessons learned, most memorable achievements, and future aspirations.
All interviews were taped and transcribed into written documents. The transcripts were then analyzed for
emerging themes. An analysis of the findings provides a clear, descriptive picture of high professionals: most
tend to have (and desire) a linear career path; they are intrinsically motivated by their job; they have a strong
desire to be recognized and respected for their expertise. They cite administrative or “busy work” and office
politics as the biggest disruptors to their effectiveness and express a desire for autonomy and independent
decision making in their roles.
The quantitative researchBased on information from the first phase of research, Korn Ferry turned to statistically measuring differences
in role preference, motivation, and engagement for high potential leaders and high potential professionals. This
phase of research involved developing 30 forced-choice items for pilot testing. The instrument was designed to
capture the dichotomous relationship between having an orientation toward either a generalist role (i.e., a role
with leadership/management responsibility) or a specialist role (i.e., a role that requires functional expertise).
The results were compelling. Three factors that define a continuum of role preference emerged from analysis on
data collected from a global sample of 10,823 participants:
Factor I: How people perceive bounds of their role responsibilities. Some people, preferring roles in narrowly defined areas, commit to their profession and enjoy opportunities that best use their expertise. Others, believing they can be versatile and taking roles in different areas, explore and try different careers. They enjoy doing things they haven’t done before.
Factor II: Different approaches to success. Individuals on one end of the continuum want to develop expertise and pursue best outcomes. They go deep and enjoy working toward greater precision and accuracy. Individuals on the other end of the continuum prefer to leverage others’ expertise and strive for practical outcomes. They make timely decisions and like to get a lot of work done, even if the work is imperfect.
Factor III: How people engage in work. Some people want to concentrate on a few priorities for high productivity, believing it most effective to get one task done before starting another. Others enjoy multitasking, preferring to get involved in many different activities and shifting easily from priority to priority.
Twelve items were retained to assess the three factors. The sum of the three factors creates a scale that can
distinguish the career orientation between high potential leaders and high potential professional employees.
Clear response patterns emerged in our global study. Data from over 10,000 assessments show that the
specialist orientation declines steadily along the organizational hierarchy. Individual contributors (ICs) have the
highest specialist orientations. Top-level executives have the lowest specialist orientations. And this pattern
emerged consistently in different international regions.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved. 21
Figure 4. Preferences.
1.30
1.32
1.34
1.36
1.38
1.40
1.42
1.44
1.46
1.48
1.50
ICs
1.45
1.42
1.38
1.35
1.33
Supervisors Managers Directors Executives
People are selected to different types of jobs. We expect the alignment between employees’ career orientation
and the type of job they do. A sample of participants (N = 134) was asked to take a supplementary survey after
they completed the career orientation assessment. Part of the supplementary survey asked participants to
rate the level of volatility, ambiguity, uncertainty, and complexity of their job, as well as the level of functional
specialization required for performing their job. The results are affirming. Individuals who are more specialist
oriented tend to report their job to be less complex and volatile (r = -.32, p < .001), and require more functional
specialization (r = .22, p < .05).
ExperienceAs leaders progress through their careers, they gain a series of experiences. Through these experiences, leaders
gain motivation, attitudes, knowledge, skills, and other capabilities that enable them to perform effectively
in future leadership roles (Campbell, 2012; Tesluk & Jacobs, 1998). Experiences, therefore, are an important
foundation for moving to new, more challenging roles.
A leader who has honed skills through depth and breadth of experiences has much more bandwidth to learn
everything else they must conquer to succeed when promoted to the next level. A leader who is behind the
curve, who lacks one or more relevant experiences, will have to learn these lessons while they are also learning
the job. This extra demand, at a time of rapid change, makes the transition risky and more likely to go awry.
Work experiences are multi-faceted, with quantitative and qualitative components that interact with each
other. They vary in specificity (task, job, team, organizational, and occupational) as well as measurement mode
(amount, time, density, timing, type) (Quinones, Ford, & Teachout, 1995; Tesluk & Jacobs, 1998). Consequently,
experience is complex, unique to individuals, and far from universal in its nature and value.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved.22
To capture this complexity, in the KFALP, we use a short, structured career walk-through that contains verifiable
questions about what an individual has done to date. This is an area where we have chosen to be brief to meet
the client’s need for brevity and to allow more robust measurement in other areas. In the KFALP, Experience is
normed against the participant’s current level. Three aspects of Experience are captured:
• Core experience can be considered “vertical experience.” It is captured by counting the number of leadership levels in which an individual has spent two-plus years working, relative to where the person is today. Each level brings unique, day-to-day experience in leadership, collaboration, and leading both up and down.
• Perspective can be considered “horizontal experience.” It represents the leadership “variations” the participant has gained from working in different organizations, industries, roles, countries, etc. Perspective helps leaders move beyond “the way we do things here” thinking to encompass the lessons and experiences of alternative or diverse ways of approaching the role, gained through variety of perspective.
• Key challenges captures the participant’s experience across 10 key leadership challenges that have been found to be particularly developmental. Not every leader will face each challenge during their career. Key challenges also reflects the leader’s role in the challenge—e.g., participant, leader, or sponsor.
The importance of developmental experiences originally was surfaced in a series of studies conducted by the
Center for Creative Leadership. In these qualitative studies, executives were interviewed and asked to describe key
events in their careers that caused the most learning. The following two questions were probed: (1) what specifically
happened on the job, and (2) what did they learn from the event. Researchers interviewed 191 executives from six
major corporations. Descriptions of the 616 events and 1,547 corresponding lessons were tabulated. The analyses
and results are summarized in the book aptly titled The Lessons of Experience (McCall et al., 1988). These researchers
observed that the most developmental experiences are challenging, stretching, and difficult.
Work experiences are challenging when existing tactics and routines are inadequate for addressing work tasks,
and new ways of dealing with work situations are required (De Pater, Van Vianen, Bechtoldt, & Klehe, 2009).
By presenting ambiguous situations and exposing individuals to a great diversity of organizational stimuli,
challenging experiences provide critical opportunities for leaders to acquire new job knowledge and practice
new skills. Stimuli that are novel, uncertain, and difficult tend to create a heightened sense of arousal within
individuals. This heightened arousal will lead to a wide range of behavioral and cognitive processes, including
learning (DeRue & Wellman, 2009). Challenging experiences also shape leaders’ work attitudes and motivation
by uncovering gaps between individuals’ current capabilities and what is required for assignment success
(McCauley, Ruderman, Ohlott, & Morrow, 1994).
Korn Ferry research has identified key career experiences that differentiate leaders. The more of these key
developmental experiences a leader accumulates, the greater the possibility that the leader will be successful
after promotion to the next level. Working with research partners at well-known universities, Korn Ferry has
also found that experience helps leaders develop their strategic thinking skills (Dragoni, Oh, Van Katwyk,
& Tesluk, 2011). In general, research has found that challenging experiences contribute to the development
of a broad variety of leadership competencies including business knowledge, visioning, strategic thinking,
problem solving, decision making, change management, and interpersonal skill (DeRue & Wellman, 2009;
Dragoni, Oh, Van Katwyk, & Tesluk, 2011; Dragoni, Tesluk, Russell, & Oh, 2009). Not surprisingly, individuals who
have had more challenging experiences are more likely to be perceived as having high promotion potential
(De Pater et al., 2009).
Leadership Experience Inventory (LEI) data at Korn Ferry indicate that, on average, leaders at the highest levels
are more likely to have had a wide range of key developmental, career-building experiences (see Figure 5). Note
also that the largest jump is from Mid-level leader to Business unit leader.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved. 23
Figure 5. Prevalence of key formative career experiences by management.
0
10
20
30
40
50
60
70
80
90
100
Critica
l neg
otiatio
ns
Exter
nal r
elatio
ns
Inhe
rited
pro
blems a
nd ch
allen
ges
Crisis
man
agem
ent
Downtur
n an
d/or f
ailur
es
Busin
ess g
rowth
Start-
up b
usin
ess
Produc
t dev
elopm
ent
Inte
rnat
iona
l/cro
ss-c
ultu
ral
First level leader
Mid-level leader
Business unit leader
Senior executive
CEO
Even within leaders at the highest levels, the impact of formative, challenging experience is observed. Among
leaders evaluated by Korn Ferry for CEO roles, those who were judged most effective reported having more
challenging work experiences. A higher proportion of the most effective CEO candidates had challenging
experiences that were heavy in strategic and people demands (Crandell, Orr, & Urs, 2015).
TraitsAwareness, Learning agility, Leadership traits, and Derailment risks are all measured using FC-IRT trait measures.
Following is background on the development of the Trait Scale Bank used to support reporting on Awareness,
Learning agility, Leadership traits, and Derailment risks.
Traits include personality characteristics that exert a strong influence on behavior. Personality influences how
people develop—what is more natural for them and what is more of an effort. Traits are the enduring aspects
of a person, changing little over time. Decades of research efforts offer insight into the application of trait
measures among upper-level management and executive leaders. Personality is related to leadership emergence
and effectiveness (Judge, Bono, Ilies, & Gerhardt, 2002) as well as the changing nature of performance across
time (e.g., Thoresen, Bradley, Bliese, & Thoresen, 2004). Meta-analyses and meta-analyses of meta-analyses
(Judge, Bono et al., 2002; Ones, Dilchert, Viswesvaran, & Judge, 2007) have demonstrated that at least three of
the factors from the widely known and established FiveFactor model show moderated but generally consistent
relationships with success among upper-level management and executive leaders. According to Church and
Rotolo (2013), 66% of organizations used personality inventories to assess high potentials.
The scales underlying the Awareness, Learning agility, Leadership traits, and Derailment risks were chosen
from among the scales of the Korn Ferry Trait Scale Bank (KFTSB). The foundation of the KFTSB is the
well-established personality science of the Big Five. The Big Five framework is an extension of the lexical
tradition, which assumes the important descriptors differentiating persons will be represented in natural
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved.24
language. This framework is a hierarchical, descriptive conceptualization of personality as opposed to a psycho-
dynamic conceptualization. Modern Big Five research is exploring linkages to neural substrates and self-
regulatory processes in the brain (DeYoung, 2010; DeYoung, Hirsh, Shane, Papademetris, Rajeevan, & Gray, 2010).
The Big Five has demonstrated global applicability (Schmitt, Allik, McCrae, & Benet-Martínez, 2007). Moreover,
this model is clearly established as the premier descriptive framework for personality science (John, Naumann, &
Soto, 2008).
The KFTSB is a state-of-the-art personality instrument developed as a new measure of the Five-Factor
personality model. There are 21 total scales designed to capture individual differences on those aspects of
personality most related to on-the-job performance and organizational fit.
In the development of the trait bank, work proceeded using a Classical Test Theory (CTT) approach and was
followed by a Forced-Choice Item Response Theory (FC-IRT) approach. Items were written by a global team to
represent each trait. Items were refined using traditional item analysis through multiple waves of data collection
focusing on factor structure and internal consistency of scales. Retained items were then tested using FC-IRT.
Factor structure was tested using exploratory factor analysis and confirmatory factor analysis (CFA) for both the
CTT and IRT versions of the scales. The final item set resulted in scales with primary loadings on the intended
factor for all scales. To test whether the structure would hold at the latent trait level—that is, on the IRT trait
scores—we conducted a CFA. Using the Goodness of Fit (GFI) and Standardized Root Mean Square Residual
(SRMSR) fit indices, we found that the intended a priori structure adequately described the underlying structure
of the IRT trait scores. Specifically, we found that the GFI = .904 and the SRMSR = .064, both within the range of
acceptable model fit as discussed in the psychometric literature (see Hu & Bentler, 1999).
From among these trait scales, appropriate scales and composites were chosen to form the core of the KFALP,
with a focus on scales that differentiate across levels of leadership. The Awareness, Learning agility, Leadership
traits, and Derailment risks signposts include trait-based measures using FC-IRT measurement technology.
Persons further along the leadership pipeline tend to score higher (or lower on derailers) than persons earlier
in their leadership career. Each scale is reported relative to normative data appropriate to the target role of the
assessment.
AwarenessTo achieve high performance, leaders must begin with a clear-eyed view of their existing strengths and their
development needs. They need to know where they excel, when they can trust their instincts and abilities, and
when they need to rely on the insights and abilities of others. They must also have keen awareness of their
thoughts, feelings, and behaviors on a moment-to-moment basis, what is sometimes called mindfulness. This
allows them to manage themselves and others more effectively.
• Self-awareness is awareness of strengths and freedom from blind spots. It is known to be related to leadership success and inoculate, to some degree, against derailment. Self-awareness is a process of continuous reassessment of self-knowledge and refinement of true self-awareness.
• Situational self-awareness is the capability of being aware of present experience in a non-judgmental way, paying attention to the importance of a variety of demands, being more aware of one’s expert intuitions, and more able to improvise in a dynamic environment.
Self-awarenessResearch suggests that derailed managers and executives share a common attribute—a lack of self-awareness
(Quast, Wohkittel, Chung, Vue, Center, & Phillips, 2013; Sala, 2003). Leaders who lack self-awareness perceive
themselves differently than other people perceive them. They tend to perceive themselves more positively than
others, which in turn makes them less aware of the weaknesses that put them at odds with the demands of the
organization (Atwater, Waldman, Ostroff, Robie, & Johnson, 2005). This lack of self-awareness can result in many
destructive behaviors, which may lead to derailment (Lombardo & McCauley, 1988; McCall et al., 1988).
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved. 25
Evidence suggests that highly self-aware leaders have a positive impact on company performance; prevalence of
high self-awareness correlates with high rate of return. Korn Ferry’s Zes and Landis (2013) analyzed 6,977 self-
assessments from professionals at 486 publicly traded companies and found that those with high self-awareness
tend to be concentrated in companies with a robust rate of return, suggesting that they might contribute to
greater business outcomes (see Figure 6).
Figure 6. Stock performance and self-awareness.
Low self–awareness
High self–awareness
0% 5% 10% 15% 20% 25%
Percent chance that a person works for a high-performing company as a factor of their self-awareness level.
Situational self-awarenessSituational self-awareness is an emerging construct in the industrial/organizational psychology literature. It is sometimes referred to as mindfulness, and has been called a “western adaptation to an eastern way of thought” (Haigh, Moore, Kashdan, & Fresco, 2011). Situational self-awareness involves one’s ability to regulate emotions, anticipate and be proactive for change, accept circumstances, live in the moment, reserve judgment, and be aware of even subtle internal and external information. Low scorers on Situational self-awareness are more likely to be focused on past or future events, are less aware of their impact on situations as they occur, and are more likely to use strict and well-defined heuristics when making decisions or characterizing a situation. Across studies and measurement instruments, Situational self-awareness has repeatedly shown compelling evidence of construct validity and has displayed key correlations with many other psychological constructs and outcomes (Haigh et al., 2011; Feldman, Hayes, Kumar, Geeson, & Laurenceau, 2007). Together with its theoretical foundations, correlational patterns help to elucidate the nature of Situational self-awareness and its potential utility. It has shown considerable positive relationships with positive affect, curiosity and exploration, emotional regulation, mood repair, and cognitive flexibility. Conversely, it has shown substantial negative relationships with a variety of maladaptive and problematic emotional and affective states. Specifically, increases in Situational self-awareness are associated with decreased anxiety, distress, depression, worry, rumination, thought suppression, avoiding experiences, and brooding (Kumar, Feldman, & Hayes, 2008; Johnson, 2007).
For much of its history, Situational self-awareness has been used as part of developmental plans for designing psycho-social interventions in diverse clinical and non-clinical settings. These include acceptance and commitment, relational frame theory, and a host of other cognitive-behavioral interventions (Baer, 2003). Related interventions designed to boost scores on Situational self-awareness-like constructs are emerging rapidly in organizational contexts as well (Hayes, Bond, & Barnes-Holmes, 2006) and have been explicated specifically for high-level executives (Passmore, 2007; Passmore & Marianetti, 2007). The potential utility of Situational self-awareness measures in organizations extends beyond its promising application for predicting who will be successful in the executive ranks. Situational self-awareness also can provide a framework or otherwise assist in coaching and development activities that show indications of substantially helping organizational personnel to manage stress, take advantage of stress, produce results while learning on the job, and mitigate derailment (Lee, 2012).
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved.26
Emerging conceptualizations of emotional regulation increasingly embed Situational self-awareness in a larger
framework as a component and an antecedent to pro-social behavior. It has otherwise been associated with
effective strategic decision making, novelty seeking, adaptive risk taking, and awareness of key resources
among key players in organizations (Langer, 2009; Nadkarni & Barr, 2008; Weick & Roberts, 1993). Currently,
the consensus in the extant literature is that Situational self-awareness has unilaterally positive effects in
organizational contexts and beyond (Lee, 2012; Dane, 2011). Although no empirical work has shown otherwise,
this notion is not without critique (Dane, 2011). The paucity of skepticism on Situational self-awareness as
a strictly positive characteristic focuses mostly on its “wide attentional breadth” and how it might distract
skilled professionals whose charge is to focus on limited information and limited scope issues in considerable
depth (Dane, 2011). Ultimately, the criticism is that Situational self-awareness scores in executives may reach a
debilitating critical mass associated with indecision and failure to react, more especially among management
personnel whose roles involve more narrowly defined task expertise and relatively static task environments
(Chajut & Algom, 2003)—job characteristics that are notably more common at lower levels of management,
although not exclusively so. Still, given the clear and substantial positive relationships between Situational self-
awareness and many general measures of positive adaptive behavior, strategic coping, and emotional states, it
is likely that job roles and organizational context will only moderate the magnitude of its otherwise generally
positive effect in occupational contexts (Goleman, 1998). Interestingly, Situational self-awareness may also
moderate the link between other psychological constructs and ratings of job performance, such that higher
Situational self-awareness strengthens positive associations where applicable (Barrick, Parks, & Mount, 2005).
Learning agilityLearning and skill development play an important role in an individual’s long-term effectiveness and career
success (Silzer & Church, 2009; Tannenbaum, 1997). The most effective way to assess a person’s potential to
learn from experience is by measuring learning agility. Learning agility is defined as the willingness and ability
to learn from experience, and subsequently apply that learning to perform successfully under new or first-time
conditions (Lombardo & Eichinger, 2000). Learning agile individuals are nimble and adaptable in changing
environments; they are key players who fill the leadership bench. Their ability to learn from experiences and take
on novel challenges sets them apart as high potentials, as evidenced by their speedy career ascent (Dai, Tang, &
Feil, 2014; Dai, De Meuse, & Tang, 2013). Nearly 25% of the Fortune 100 assess learning agility as one component
of potential.
Learning agility is especially crucial during job transitions—such as a promotion—when an individual invariably
faces new and unfamiliar situations. Instead of automatically defaulting to favorite past solutions or problem-
solving tactics, learning agile leaders apply fresh and varied approaches, ideas, solutions, and techniques to
solve those new, tough problems. In short, learning agile leaders find new ways to successfully navigate unknown
and unforeseen challenges.
The KFALP uses four scales to assess learning agility:
• Mental agility is a broad curiosity about the complex issues, challenges, and novel situations that leaders face daily, and sets the stage for effective problem solving. This curiosity about issues and problems helps spot patterns, trends, and relationships.
• People agility is understanding the value of getting work done with and through people, being attuned to individuals’ needs and motivations, and typically skilled at reading people with an effective influencing style.
• Change agility is embracing change and taking well-reasoned risks even in the face of that change. It includes openness and acceptance of change and willingness to balance the risks and trade-offs vs. waiting.
• Results agility is being energized by new, tough assignments and overcoming obstacles to achieve stretch work objectives. It includes the enjoyment of being judged against external standards of achievement.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved. 27
These four aspects of learning agility were selected based on prior factor analysis results with multi-rater
measures. Each aspect is associated with others’ (e.g., bosses’, human resource professionals’) ratings of
potential and propensity to stay out of trouble (i.e., unlikely to derail) (Lombardo & Eichinger, 2000).
The ROI for organizations and leaders is clear. Research shows that learning agile leaders are rated more
competent, recognized as having the most potential for advancement, get promoted faster and more often than
their peers, and outperform their peers after a promotion (Dai et al., 2013; Dragoni, Tesluk, & Oh, 2009; Dries,
Vantilborgh, & Pepermans, 2012; Lombardo & Eichinger, 2000). Learning agility also accounts for incremental
validity in performance above and beyond intelligence and personality (De Meuse, Dai, & Hallenbeck, 2010).
Korn Ferry has the most extensive research for describing and measuring learning agility. This includes
observable competencies and a set of related traits. Korn Ferry research found that highly learning agile
people earn promotion much more quickly (Dai et al., 2014; Dai et al., 2013). After grouping individuals by low,
moderate, and high learning agility scores, our analysis found that managers with high learning agility received
twice as many promotions over the 10-year period as those with low learning agility (see Figure 7).
Figure 7. Number of promotions managers were likely to receive over 10 years.
Low learning agility
Moderate learning agility
High learning agility
0 0.5 1
Number of promotions
1.5 2 2.5
Leadership traitsThe more an individual’s traits align with the traits that are characteristic of successful leaders, the greater the
potential for future success at higher organizational levels. Traits factor heavily into questions of leadership
potential because personality profiles look substantially different at each progressive level of management
(Crandell, Hazucha, & Orr, 2014).
The KFALP reports leaders’ scores on five traits. Each is defined briefly below, and then discussed in more detail.
• Focus is the willingness to let go of personal attention to all details and seek a bigger-picture perspective rather than the pursuit of minutiae at the expense of the big picture.
• Persistence is having and pursuing closely held and personally valued long-term goals despite obstacles and distraction. This focus helps sustain leaders through difficulties and detours. This is different from focusing on near-term goal attainment or on goals set by others.
• Tolerance of ambiguity is the capability to thrive in a volatile, uncertain, complex, and ambiguous (VUCA) world. Some leaders find energy in these situations and can work productively despite a lack of a clear view of the future, while others are more comfortable with conditions providing clarity and certainty.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved.28
• Assertiveness is the willingness to take the lead if a clear leader is not apparent and to do so with little hesitation. Successful leaders are comfortable taking charge, and they find leading the way to feel natural and expect good outcomes when they do.
• Optimism is the degree to which people tend to disregard disappointment, are satisfied with who they are, and expect the future to be bright. Successful leadership requires a steady healthy optimism and good expectations for the future.
FocusFocus taps the extent to which individuals are detail oriented, thorough, and careful in decision making and
work processes. Very high scorers may even be described as dogmatic and/or problematically perfectionist.
Anecdotally, a problematically high Focus score can be described by the aphorism “the perfect can be the
enemy of good.” Attention to detail may contribute to early career success, but inhibit or even derail a top
executive. That is, Focus is positively associated with performance for lower-level managers and individual
contributors whose roles involve a notable degree of task orientation and applicability of expertise, perhaps
as well as deference to protocol and well-defined process standards. But, Focus and Focus-like scores tend
to decrease at higher levels of management (Brousseau, Driver, Hourihan, & Larsson, 2006; Lewis & Ream,
2012). Among executive leaders, Focus and Focus-like measures typically correlate negatively with executive
performance and other management outcomes including career success (Lewis, 2012). Conceptually
convergent or otherwise markedly correlated measures have even been characterized as derailers for executive
managers (viz., “dutiful” in Hogan & Hogan, 2009) and are negatively correlated with defining components of
transformational leadership behaviors and traits. This shift across management level accounts, in part, for the
paradox of a merely satisfactory new manager who simultaneously has the potential to be a superior-performing
executive. And it explains, in part, why some leaders plateau despite early success. Having the right level—not
too much, not too little—of these traits is one indicator of future high performance as a leader.
PersistencePersistence refers to a tendency toward passionate and steadfast pursuit of personally valued long-term or
lifetime goals or values, in spite of obstacles, discouragement, or distraction. High scorers tend to push through
adversity and tend not to give up on difficult tasks and pursuits. They are typically characterized as resilient and
as having stamina and long-term or stable focus. Low scorers are more likely to change course when faced with
adversity, while putting emphasis on emergent opportunities and short-term pursuits and accomplishments.
Persistence has reference to long-term goal or value perseverance, resilience to adversity, and is not primarily
maintained by short-term periodic and ongoing work-related feedback from others or from comparison with
easily defined standards of excellence.
Duckworth, Peterson, Matthews, and Kelly (2007) explain that Persistence as a construct has arguably one of
the longest histories in all of psychology and particularly in the “psychology of achievement.” Several early
researchers, going back as far as the late 19th century, were interested in variables that separated similar and
even similarly gifted individuals into levels of achievement. Many found that persistence, perseverance, and
resilience were often key differentiating traits among individuals who otherwise had similar ability levels or
similar IQ (Terman & Oden, 1947; Howe, 1999; as noted in Duckworth et al., 2007). Simonton (1994) concludes
that Persistence, or “grit,” is among the more certain and consistent variables that high-impact and notable
historical figures most often have in common. Persistence is typically found to be uncorrelated or slightly
negatively correlated with IQ levels, and its incremental utility (over IQ and aptitude) for predicting life and
occupational outcomes seems well established (Duckworth et al., 2007; Ackerman & Heggestad, 1997; Moutafi,
Furnham, & Paltiel, 2005; Eskreis-Winkler, Shulman, Beal, & Duckworth, 2014). In fact, its utility in predicting
success is sometimes seen as the cornerstone for understanding the differential and additive utility of natural
ability vs. disposition-related variables in understanding life’s outcomes—including work-related outcomes
(Ericsson & Charness, 1994). High Persistence scores are associated with increased emotional stability,
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved. 29
increased standardized test scores, achievement motivation, educational attainment, educational performance,
employment retention, and retention in challenging educational programs—including highly selective military
training programs (Duckworth et al., 2007; Eskreis-Winkler et al., 2014).
Persistence-like constructs are also associated with increased levels of EQ, learning agility, strategic vision,
adaptability, motivation to lead, and stakeholder sensitivity among leaders or potential leaders in organizations
(Dries & Pepermans, 2012). Persistence has also been positively associated with CEO and entrepreneurial
success (Baum & Locke, 2004). CEOs having higher levels of Persistence-like traits tend to be more resourceful
and confident. They are more effective at communicating, setting, and reaching goals, as well as growing
businesses (Baum & Locke, 2004). Persistence may be characterized as a component or expression of work-
related “passion” (Houlfort, Philippe, Vallerand, & Menard, 2014) which, when associated with other socio-
emotional adaptive states, is positively predictive of increased enthusiasm, discretionary effort, positive work-
related relationships, positive organizational outcomes, work satisfaction, and resilience to burnout (Cardon,
Wincent, Singh, & Drnovsek, 2009; Cardon, Zietsma, Saparito, Matherne, & Davis, 2005; Liu, Chen, & Yao, 2011;
Philippe, Vallerand, Houlfort, Lavigne, & Donahue, 2010).
Tolerance of ambiguityA comfort with uncertainty and a willingness to make decisions and plans in the face of incomplete information
are hallmarks of high scorers on measures of Tolerance of ambiguity. Tolerance of ambiguity is a common
and critical component of measures used in executive selection, development, and succession contexts
(Lewis & Ream, 2012). Although the strength of association may be moderated by the nature of job roles and
contexts, high Tolerance of ambiguity among executives has been almost unilaterally associated with positive
individual- and company-level outcomes (Yukl & Mashud, 2010). Business climate and organizational functioning
characterized by ambiguity and uncertainty has repeatedly been characterized as “the new normal” (Cone,
2013), and management professionals and managerial scientists include Tolerance of ambiguity among the top
characteristics of successful executive leaders into the foreseeable future (Gratton & Erickson, 2007; Gratton,
2010). High Tolerance of ambiguity is markedly associated with innovation and an entrepreneurial orientation to
vocational pursuits, whether within or without organizational contexts. High scorers on measures of Tolerance of
ambiguity are more likely to seek and value diverse feedback, experiment, seek opportunities for innovation, and
avoid micromanaging (Kirschkamp, 2007). For medical organizations, Tolerance of ambiguity has been called a
key indicator differentiating between physicians who can and cannot successfully make the difficult transition
from clinical to executive management functions (Sherrill, 2001). Interestingly, high scorers on measures of
Tolerance of ambiguity do not eschew data or avoid seeking information by which planning and executing
decisions can be guided. Rather, an effective executive with an ambiguity tolerant disposition typically has a
more adaptive and nimble sense of when a critical mass of key information has been gathered, and they proceed
without problematic anxiety in cases where others may not when faced with information that seems inadequate
or incomplete. Brainstorming to fill in data gaps, pragmatism, and contingency plans are usually key accessories
for effective and highly ambiguity-tolerant executives (Strosaker, 2010).
AssertivenessEmpirical findings show Assertiveness to be a key component of leadership emergence and potential as well
as results-drive and achievement orientation (Dries & Pepermans, 2012).3 Assertiveness measures whether
people are inclined to proactively assume wide responsibility, take charge, and lead others. A notably assertive
individual is convinced that she/he should be in charge, and that both individual and group outcomes will
be optimized when she/he is granted group-level decision-making discretion, leadership status, authority
to delegate, and authority to set or heavily influence organizational objectives. As such, high Assertiveness
might be characterized, at least in part, as self-efficacy for leadership in general (Amos & Klimoski, 2014). High
3 Dries & Pepermans (2012) separate components of Assertiveness into multiple constructs for which they argue conceptual divergence. Taking initiative, they assert, is a component of “drive,” assertiveness in decision making is a component of “analytical skill,” and actively looking for opportunities to lead, delegating, and objective setting are components of “emergent leadership.” In their study, these higher-order constructs, however, show markedly and arguably statistically convergent correlational patterns (all having r > .75).
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved.30
Assertiveness scorers may also tend to be seen as confident, aggressive, and decisive, while low scorers are
likely perceived as tentative, passive, reserved, or indecisive and more comfortable deferring to and following
the lead of other individuals or groups. Low scorers may also have and attain leadership roles, but this is far
more likely when leadership status has been formally assigned and is associated with known and explicated
relative managerial rank and job title. High scorers on Assertiveness-like measures, on the other hand, will take
charge because they feel like it will benefit organizational members and collective pursuits whether or not they
were told or were granted clearance to assume responsibility as such. In short, high Assertiveness individuals are,
to some extent, in charge because they have decided they are in charge, and not necessarily because somebody
else, with or without authority, has told them that they are in charge. Their leadership status and effective
leadership status often is or at least begins as a de facto more than a de jure leadership status.
In the extant Big Five personality literature, a construct similar to Assertiveness is sometimes conceptualized as
a component of the factor of Extraversion, and is often called Dominance (e.g., Costa & McCrae, 1992; Depue &
Collins, 1999). Ones et al. (2007), however, in a comprehensive meta-analytic review, show marked differential
predictive utility for these two components of Extraversion—Sociability and Dominance, particularly for
managerial professionals. They find the impact of Dominance on managerial performance is positive and notably
different and larger than the impact of Sociability. Judge, Bono et al. (2002) similarly found Sociability and
Dominance having separate effects on leadership. Others have conceptualized and supported Assertiveness-
like constructs as belonging to higher-order factors removed from Sociability or other social-behavior-related
measures (e.g., Dries & Pepermans, 2012; Northouse, 1997; Mann, 1959; Stogdill, 1948; Hogan, 1983; Wiggins,
1996). Hogan (1983) and others (e.g., King & Figueredo, 1997) in empirically-based higher-order personality
structures separate Dominance from Extraversion or Sociability, concluding that the latter is better dubbed
“Surgency”—having reference to general positive mood and sociability, whereas, Dominance emerges as its own
factor with primary reference to confidence, independence, and aversion to submissiveness or deference. Others
argue Assertiveness and social variables are clearly associated, but not necessarily conceptualized as sub-
components of a single common latent factor (Dries & Pepermans, 2012). Yet others (e.g., McCrae & Costa, 1987)
assert that Sociability is not best combined with Assertiveness in an Extraversion factor, but that Sociability
belongs with emotional and affective variables.
Assertiveness predicts both self and other rating of Sociability, as well as “competency” domains like creativity,
analytical thinking, and problem solving (e.g., Anderson & Kilduff, 2009). Interestingly, Assertiveness seems
to affect others’ perceptions of competence in various leadership domains incrementally in models also
containing scores of actual competence. As such, Anderson and Kilduff (2009), among others, show that high
Assertiveness leaders typically instill trust and confidence in others in ways that are not always directly linked
to rationality, truth, or more objective measures of actual leadership status or skill. Increased and even very high
marks on Assertiveness-like measures among CEOs are positively associated with company innovation and
company patent counts, and the effect is notably stronger for CEOs operating in highly competitive markets
(Galasso & Simcoe, 2011). Assertiveness to lead, however, can sometimes be associated with lack of receptivity,
micromanaging, and/or need for control in ways that create challenges for team performance, particularly when
high Assertiveness marks are present in individuals having low marks on affiliation-type measures or measures of
EQ and/or positive affect (Driskell & Salas, 1992).
OptimismLeaders who are high in Optimism expect the future to be bright, are likely to disregard disappointment, and
are satisfied with themselves. Optimists are likely to be warm and slightly dominant interpersonally, and tend to
handle relationships successfully (Carver & Scheier, 2014).
Optimism is a compound trait, related to multiple factors in the Five Factor model of personality (Hough & Ones,
2001). Originally thought to be a blend of Extraversion and Emotional Stability (Hough & Ones, 2001), Optimism
has consistently been found to have moderate to strong relationships with each (Carver & Scheier, 2014; Kam &
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved. 31
Meyer, 2012). More recently, Optimism has been shown to be related to four of the Big Five, having somewhat
smaller associations with Agreeableness and Conscientiousness (Sharpe, Martin, & Roth, 2011). Optimism also has
been conceptualized as a key component of mixed models of Emotional Intelligence (Livingstone & Day, 2005).
Although Optimism is a facet of personality, it also is cognitive in nature because it involves expectations for
the future. Through these expectancies, Optimism is also connected to motivation and self-regulation (Carver
& Scheier, 2014). Positive expectancies are associated with more persistent and effective pursuit of goals
(Segerstrom, 2007). Consequently, individuals high in Optimism appear to be highly effective at investing effort
in complex situations, effectively increasing their engagement for higher-priority goals and under favorable
circumstances (Carver & Scheier, 2014). Leaders who are confident about the future will continue to invest effort;
those who are doubtful about eventual success are likely to disengage (Carver & Scheier, 2014).
Optimism has been associated with future success in a variety of contexts. For example, one study found
Optimism measured during first semester in law school predicted higher salaries a decade later after controlling
for hours worked (Segerstrom, 2007). Personality traits that contribute to Optimism have been associated with
leadership emergence and effectiveness. For example, in one meta-analysis, Extraversion and Emotional Stability
were robustly related to leadership in business settings; Emotional Stability and Conscientiousness were robustly
related to leadership in military and government settings (Judge, Bono et al., 2002). Composites of Emotional
Intelligence that include Optimism are positively related to job and life satisfaction, accounting for incremental
validity in these outcomes beyond the Big Five personality traits (Livingstone & Day, 2005).
CapacityCapacity refers to logic and reasoning, or cognitive ability. Research has shown that cognitive ability influences
virtually every aspect of job performance and potential (Ones, Dilchert, Viswesvaran, & Salgado, 2010). It is
positively related to leadership emergence and effectiveness (Judge, Colbert, & Ilies, 2004). High-performing
leaders are effective analytical and conceptual thinkers. They are astute at spotting patterns or trends in data
that others miss. And they solve problems with aplomb—at first individually, and then as leaders—by marshaling
and focusing resources on the right challenges.
It is an individual skill and there is a subtle trap when thinking about Capacity as one moves up in leadership: a
person’s role changes from being the primary problem solver to ensuring that the problem gets solved. Leaders
who cannot shift out of individual problem-solving mode and into the job of coaching and mentoring others to
analyze problems will struggle beyond mid-level leadership roles.
Likewise, organizations that rely on individual problem solving as their sole or even primary indicator of high
leadership potential risk flooding their pipeline with people who will peak in mid-level roles because they revert
to solving complex problems themselves. For this reason, it’s risky to assess pure cognitive ability without
simultaneously considering how this cognitive ability is imparted in a leadership role. Cognitive ability is most
strongly related to leader effectiveness when stress is low and when leaders are directive (Judge, Colbert et al.,
2004).
There is a great deal of confusion and misinformation in the popular press about terms such as cognitive ability,
intelligence, reasoning, IQ, and the like. All tests of cognitive ability are related, that is, correlated with each other
to some greater or lesser degree. There are many tests of cognitive ability, each with strengths and limitations,
and each with a somewhat unique interpretation and meaning. Perhaps the best available survey of cognitive
ability tests, their relationship to each other, and how they fit in the overall domain of cognitive ability is Carroll
(1993).
Our approach is pragmatic. We define Capacity very specifically as skill at detecting patterns and trends even in
ambiguous, contradictory, or otherwise “noisy” environments, and we use a test that is independent of language
or tacit knowledge of a domain in order to provide global comparability of score. To measure Capacity, we rely
on the Raven’s APM Version 2 (NCS Pearson, 2007) because it is the only effectively language-free test available,
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved.32
a feature important to global clients who wish to apply the same test anywhere in the world. The Raven’s APM
is the most widely-used, high-level, language-independent cognitive ability option available. It is a non-verbal
measure of gf, fluid intelligence (Horn, 1980). Gf is the capacity to think logically and solve problems in novel
situations, independent of language or tacit knowledge of a domain.
Any strong measure of fluid intelligence like the Raven’s APM is likely to have a quirk to consider. It will be
negatively correlated with age. This is because when younger, we rely more on gf for performance in novel
learning situations and for adaptive behavior. As we get older, we rely more on crystalized intelligence, gc
(experience, knowledge, and deductive reasoning). As a result, we tend to lose gf reasoning skills represented
due to disuse—or at least we rely on them less. This is not always the case, if we keep practicing, but many
people increasingly rely on heuristics, tacit knowledge, and experience for adaptive problem solving and less of
gf over time.
Derailment risksDerailment is the failure to achieve one’s potential. The outcomes associated with leadership derailment can
be very costly on many dimensions. In addition to the millions of dollars of direct and indirect financial costs,
derailed managers can engender a negative impact at the individual, team, and organizational levels. Such
leaders don’t build cohesive teams, dwindle the morale of coworkers, damage customer relationships, and fail
to meet business objectives (Bunker, Kram, & Ting, 2002; Hughes, Ginnett, & Curphy, 2008). Some estimate that
30% to 50% of high potential managers and executives derail (Lombardo & Eichinger, 1989).
The risks related to derailment go up at higher job levels: expectations are higher and consequences of failure
are higher (Hogan & Hogan, 2001; Tang, Dai, & De Meuse, 2013). As shown in Table 4, the potential for derailment
is rated significantly higher for upper management than lower and middle management (Tang & Dai, 2013). The
reasons for this increase in rated likelihood of derailment by others by level include: (1) the strengths that propel
leaders to the top often have corollary weaknesses; and (2) increased demands and higher expectations yield
more focused scrutiny.
Table 4. Derailment risk ratings across management.
Derailment factor Individual contributorN = 1,256
First level leaderN = 3,957
Mid-level leaderN = 3,307
Senior executiveN = 1,005
Insensitive to others 1.34 1.51 1.60 1.62
Overmanaging 1.54 1.67 1.73 1.77
Unable to adapt to differences
1.42 1.54 1.54 1.55
The KFALP Derailment risks measures use profiles or configurations of scores on traits to highlight a propensity
to behave in a way that may be problematic in future situations. For example, in stressful, ambiguous, or complex
situations, leaders may have a propensity to perform in derailing ways. That is, to handle extremely challenging
tasks, leaders may act in a way that could have a negative impact on the people they work with and their teams.
It cannot be overemphasized that risks are not destiny. Feedback on these risks is provided to equip leaders with
the awareness needed to expand their repertoire of behaviors and avoid defaulting to potentially problematic
patterns when faced with challenging circumstances.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved. 33
The KFALP provides feedback on three derailment risks, chosen because scores differentiate those who have
advanced from those who have not:
• Volatile is behaving in unexpected or detrimental ways. Effective leaders tend to be steady, even-tempered, and composed. Leaders who behave in a volatile way find it more difficult to build trust and confidence among their people.
• Micromanaging is staying involved in too many decisions rather than passing on responsibility, doing detailed work rather than delegating it, and staying too involved with direct reports. Effective leaders allow their people to succeed through their own efforts and skills.
• Closed is being dismissive of differing perspectives. Being closed makes it more difficult to respond to the need for change or to cultivate new ideas that can improve performance of the leader or team. Effective leaders are open to the perspectives and ideas of others.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved.34
Section VI The Korn Ferry Assessment of Leadership Potential: Uses, administration, scoring, and reporting overview
Intended usesAt Korn Ferry, we are committed to offering science-based and experience-tested assessments that support the
success of our clients. The Korn Ferry Assessment of Leadership Potential (KFALP) is designed to provide data
important for individuals and organizations to consider as they think broadly about the leadership potential of
their internal talent. It is intended to evaluate long-term potential of individuals and leadership talent pools.
• It provides organizations with the ability to objectively and accurately identify people with high leadership potential.
• It gives a complete view of a person’s leadership potential, no matter where they are in the organization.
• It accurately identifies high potentials in seven key facets, or signposts, proven by research to be related to advancement in leadership roles.
The KFALP does not measure readiness to perform in a specific new job today nor does it measure fit with a
particular role. KFALP is designed to answer the question, “Who has the potential to take on higher-level, bigger
leadership roles in the future?” It was not developed or intended for use as a stand-alone potential screening
tool, but rather as a supplement to all available indicators of potential to assist with careful consideration and
identification of high potentials using the full range of information available to clients and as part of a complete
client process.
When used as a supplement to other measures for selection of persons into specific jobs, target roles are
assumed to be feeder roles for a leadership career progression where persons selected are expected within
a reasonable time to advance to higher leadership roles of greater responsibility. Use of the KFALP assumes
that this selection is for Professional/Individual contributor, First level leader, or Mid-level leader roles. Separate
Best Practices and Job Characteristics Questionnaires are available to assist in gaining maximum value from
the KFALP in selection contexts. The KFALP is best used in conjunction with all other data and information that
clients may have relevant to the potential of talent pools (education, specific domain knowledge, etc.).
The KFALP is not designed to supplement executive level leader selection, except for post-selection
development insight and onboarding support. It is not useful for evaluating potential at the top officer levels.
This is because the goal is to evaluate potential for developing, over time, the skills and experiences required in
roles of significantly greater breadth, complexity, and responsibility. Once a leader has attained an executive role,
evaluating their potential for advancement to the C-suite requires a more individualized level of granularity than
is provided by the KFALP. In addition, at this level, talent reviews seek to include a view of both potential and
readiness for the target role.
Administration and timingThe KFALP is available exclusively online and in many major business languages. The KFALP is extremely
efficient, taking 35–40 minutes to complete, exclusive of Raven’s Progressive Matrices, which adds up to 40
minutes. There are a number of branching items, and several items require multiple responses. The total number
of responses will be approximately 240–280 for most participants.
There are a variety of item formats included. The majority is the FC-IRT item administration which requires
choices to be made in a ranking format. It is more demanding than traditional rating formats. Many candidates
welcome this level of structure and the engagement required, correctly inferring that more accurate insights
are possible.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved. 35
Scale scoringWith the exceptions of Career planning and Experience, raw scores are normed against the target role level and
reported as percentiles on a scale such as the one in Figure 8. As noted above, Career planning is not normed;
however, higher scores indicate increasing focus of career planning and greater specificity of career plans.
Experience is normed against current level.
Persons reaching or exceeding the 50th percentile are scored as “green” for that sub-dimension. Individual
Reports contain specific feedback for scores at the very top of the range (yellow at the top of the range in
Figure 8), suggesting cautions regarding possible overuse of or dependence on a strong characteristic. These
are still positive scores.
Figure 8. Individual Report scale.
Tolerance of ambiguity
More likely to be disoriented or even immobilized by uncertainty or ambiguity in information or situations.
Comfortable with uncertain, vague, or contradictory information. Open to alternate solutions.
1 11 50 91 99
85
The meaning of yellow and green is simply whether or not the individual scored above or below the average of
persons who have already advanced to the target level. One should keep in mind that the scoring is against a
high target.
Signpost markersSignposts are marked “green” based on the number of sub-dimensions on which an individual has reached a
green threshold:
• Drivers – any two of three sub-dimensions.
• Experience – any two of three sub-dimensions.
• Awareness – both sub-dimensions.
• Learning agility – any three of four sub-dimensions.
• Leadership traits – any four of five sub-dimensions.
• Capacity – one (of one) sub-dimension.
• Derailment risks – none of the three Derailment risks factors.
The green and yellow signpost markers are provided as an “at a glance” view of the participants’ results. They
are also used to assist in sorting the Talent Grid Report.
No signpost or scale appearing yellow should be considered on its own as a “knock out” factor for consideration of potential. The KFALP data must be considered as a whole-person view and in conjunction with
everything job-relevant that is known about the person.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved.36
ReportingMultiple report and data formats are available:
• Individual Report – shows individual participants their results in each sub-dimension of each of the seven signposts against the chosen target level. The report provides insights to help map individual development strategies based on the individual’s pattern of responses. Based on the results, three specific development priorities with suggestions are provided.
• Talent Grid Report – provides an “at a glance” view of groups of individuals and how they compare with others within the organization, arming leadership with the ability to segment talent and make talent plans based on the pattern of results.
• Data Extract – contains exactly the same data as the Individual Report, extracted into a manipulable Excel format.
• Group Summary Report – a rich report comparing client participants against relevant global benchmarks.
KFALP technical characteristics and validity
ReliabilityReliability is the consistency or dependability of a score. It provides an estimate of the amount of error, or
noise, in the test’s measurement. In principle, if a test was completely reliable, all of the variation in people’s
scores would be due to their true scores—all signal. In practice, there is always some noise in test scores, similar
to there being some static on a phone call. So reliability can be thought of as the signal portion of a signal to
noise ratio for a test.
There is no exact determination of reliability, all estimates are—estimates. There are a number of ways
to estimate reliability, such as test-retest (multiple administrations), or various internal methods (single
administrations). The primary technical requirement is for a quality estimate of the proportion of true score
(non-error, non-noise) included in a measure. Test-retest is a relatively poor estimate of the proportion of true
score in a measure. This is because 1) it is not clear what retest time period is correct for this purpose, 2) it is not
possible to preclude true change on a measured characteristic during the interim period, and 3) it is not possible
to remove the effects of any process learning, assessment memory, retest boredom, or other effects from the
retest results. Crocker and Algina (1986) conclude, “In view of these issues, it is probably sensible to assume
that the test-retest coefficients probably present a somewhat inaccurate estimate of the theoretical reliability
coefficient.” We focus, then, on the superior single measurement event procedures.
In Classical Test Theory (CTT), a single measurement event indicator of reliability is used to characterize the
reliability of a test, typically, Cronbach’s a (Cronbach, 1951). It has been demonstrated that Cronbach’s a tends
to be a conservative, lower-bound estimate of reliability (Osburn, 2000), which is one of the reasons it is so
commonly used. A sufficiently high estimate of Cronbach’s a implies other types of reliability estimates will be
even higher. Typically, in CTT terms, a value of .70 or higher is considered good in personality testing (Nunnally &
Bernstein, 1994).
In Item Response Theory (IRT), there is no direct analog to Cronbach’s a, however, there is a similarly accurate
approach. In IRT, an estimate of error in a test can be calculated for each and every possible score along the
full range of scores. For a conservative estimate comparable to Cronbach’s a, we calculate the proportion of
true score included in the measure in scores across the full range of scores. We have computed from this an
“average” reliability for each trait scale. These averages were computed by estimating the IRT score and error
variances. The IRT score variance was estimated by computing the variance of computed IRT scores; the error
variance was estimated by averaging the squared Standard Error of Measurement across the trait range. With
an estimate of the IRT score and error variance in hand, the reliability, r’tt, is estimated as the ratio of true score
variance (IRT score variance minus error variance) to total score variance (IRT score variance). Nearly all of the
scales in the KFALP are quite high and most are in the .8s or even .9s.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved. 37
Reliability estimates are presented for each scale in Table 5 below. The FC-IRT scales are estimated as described
above. Role preferences is a CTT scale and is estimated as Cronbach’s alpha. Raven’s APM Version 2 is also a
CTT scale and estimates are taken from Raven’s Advanced Progressive Matrices (APM): Evidence of Reliability
and Validity (NCS Pearson, 2007). Note that the Experience scales and Career planning are not included. This is
because these measures do not involve estimates of a latent trait, but directly reflect back in summary form the
participant’s experiences and planning reported by the participant in the career walk-through.
Table 5. Reliability estimates.
r’tt, All N > 500
Drivers
Advancement drive
Role preferences
.92
.67
Awareness
Self-awareness
Situational self-awareness
.67
.72
Learning agility
Mental agility
People agility
Change agility
Results agility
.81
.89
.91
.81
Leadership traits
Focus
Persistence
Tolerance of ambiguity
Assertiveness
Optimism
.80
.84
.85
.88
.84
Capacity
Raven’s APM Version 2 .87
Derailment risks
Volatile
Micromanaging
Closed
.89
.91
.91
Relationship of KFALP FC-IRT measures with other measuresCorrelation of the scales of one instrument with similarly conceptualized and defined scales of another
instrument is considered evidence of construct validity (AERA, 2014). We have a sample of N = 230 persons
who have taken KFALP and the Global Personality Inventory (GPI) scales (Corporate Executive Board, 2001).
The Global Personality Inventory is a 37-scale assessment of adult personality. The Global Personality Inventory
scales are not necessarily similarly conceived or defined as KFALP scales and tend to be somewhat more
narrowly defined than scales included in the KFALP. Scale definitions are found in Appendix D.
CTT and FC-IRT self-assessment methods are quite different in response format and scoring procedures. One
would not necessarily expect high correlations between administrations of even the same scales and items using
the two methods due to 1) item ranking vs. Likert rating data collection, 2) simple arithmetic scoring of CTT vs.
two-stage linear model scoring in IRT, 3) practical caps on correlations due to imperfect reliability of scales, and
4) the fake resistance of FC-IRT methodology.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved.38
In our case, we would expect moderate correlations between conceptually related scales using different
instruments, items, and scoring methodologies and a pattern of correlations which generally support the
construct validity of KFALP scales. The full correlation matrix is 37 x 16. Due to its size, it is presented in
Appendix E, along with GPI scale definitions, found in Appendix D. The results, however, are recapped in Table 6
below with notable positive and negative correlations from the full matrix of correlations.
Table 6. Notable correlations between KFALP FC-IRT trait-based and GPI scales.
Global Personality Inventory scale
KFALP scale Most notable positive r Most notable negative r
Drivers
Advancement drive AdaptabilityDesire for AchievementEnergy LevelSociabilityCompetitivenessTaking ChargeRisk-taking
.37
.36
.35
.33
.32
.31
.30
Attention to DetailPassive-Aggressive
-.23-.28
Role preferences AdaptabilityEnergy LevelDesire for AchievementRisk-taking
.36
.25
.22
.22
Attention to DetailPassive-Aggressive
-.25-.25
Awareness
Self-awarenessSituational self-awareness
Stress ToleranceEmotional ControlSelf-AwarenessWork Focus
.26
.23
.22
.21
Learning agility
Mental agility InnovationAdaptabilityDesire for AchievementThought FocusVisionRisk-taking
.40
.36
.30
.29
.27
.26
Passive-Aggressive -.20
People agility InfluenceSociabilityEmpathyConsiderationSocial Astuteness
.37
.34
.32
.25
.22
Independence -.20
Change agility AdaptabilityRisk-takingDesire for AchievementEnergy LevelOpennessInitiative
.49
.40
.36
.32
.30
.28
Attention to DetailMicro-ManagingPassive-Aggressive
-.21-.24-.29
Results agility Energy LevelDesire for AchievementSelf-ConfidenceResponsibilityInitiative
.38
.36
.24
.23
.23
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved. 39
Global Personality Inventory scale
KFALP scale Most notable positive r Most notable negative r
Leadership traits
Focus Adaptability .24 Micro-ManagingWork FocusImpressingAttention to Detail
-.34-.35-.35-.61
Persistence Work FocusSelf-AwarenessResponsibilityEnergy LevelOptimism
.31
.29
.23
.22
.22
Tolerance of ambiguity AdaptabilityDesire for AchievementRisk-takingEnergy LevelInnovation
.49
.35
.32
.28
.26
Attention to DetailPassive-AggressiveMicro-Managing
-.26-.27-.27
Assertiveness Taking ChargeInfluenceSociabilityCompetitivenessInitiativeDesire for Achievement
.53
.34
.33
.29
.29
.27
Optimism Stress ToleranceOpennessSociabilityOptimism
.30
.25
.22
.20
Micro-ManagingNegative Affectivity
-.24-.26
Derailment risks
Volatile Emotional ControlStress ToleranceWork FocusSelf-Awareness
-.42-.39-.29-.23
Micromanaging Attention to DetailMicro-ManagingPassive-AggressiveWork FocusIndependenceEgo Centered
.51
.38
.26
.25
.21
.20
ConsiderationEmpathyInterdependenceSociabilityOpennessTrustAdaptability
-.20-.20-.20-.21-.21-.22-.34
Closed Passive-AggressiveMicro-ManagingAttention to Detail
.31
.29
.21
InitiativeEnergy LevelRisk-takingDesire for AchievementOpennessAdaptability
-.25-.31-.31-.32-.32-.52
(continued)
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved.40
Relationship with advancementCriterion-related validity is evidence for the relationship of a measure to desired outcomes. Table 7 below
displays effect sizes for the differences in score for Individual contributors against each higher organizational
level. This is the foundation of differentiation of those who have advanced from those who have not.
An effect size is a quantitative measure of the strength of a phenomenon, in this case, the standardized mean
difference between groups, Cohen’s δ. An effect size can be interpreted as small, medium, or large depending on
its context. A commonly used interpretation is as follows: an effect size of 0.2 is considered a small effect, 0.5 a
medium effect, and 0.8 and up a large effect (Cohen, 1988).
As shown in Table 7 below, bold indicates a statistically significant difference. The results show that there were
progressively higher average scores for persons who have advanced to the next level for most scales, and, as
expected, progressively lower scores for Derailment risks scales. Most measures have a medium to large effect size.
Table 7. Effect sizes for different position levels.
(Individual contributor N > 440 to 483)
Individual contributor/professional to First level leader(N = 1299 to 2001)
Individual contributor/professional to Mid-level leader(N = 1275 to 1958)
Individual contributor/professional to Functional leader(N = 1297 to 2218)
Individual contributor/professional to Business or organizational unit/division leader (N = 1110 to 1972)
Individual contributor/professional to Senior/top functional executive(N = 864 to 1593)
Individual contributor/professional to Top business or organizational group executive(N = 265 to 595)
Advancement drive .33 .38 .47 .62 .61 .70
Role preferences .15 .25 .33 .43 .42 .49
Situational self-awareness .14 .17 .21 .25 .30 .31
Self-awareness .05 .15 .19 .14 .20 .17
Mental agility .13 .21 .30 .35 .34 .35
People agility .18 .32 .35 .43 .50 .55
Change agility .28 .43 .56 .72 .75 .70
Results agility .20 .31 .36 .40 .44 .56
Focus .39 .53 .60 .60 .59 .52
Persistence .06 .15 .20 .27 .40 .40
Tolerance of ambiguity .42 .56 .73 .78 .85 .83
Assertiveness .33 .50 .52 .69 .69 .82
Optimism .12 .24 .28 .38 .37 .42
Volatile -.27 -.50 -.65 -.69 -.77 -.65
Micromanaging -.44 -.32 -.38 -.42 -.48 -.52
Closed -.12 -.61 -.73 -.75 -.79 -.73
The progression of score changes by level can be seen clearly in percentile terms in Figure 9 below. Individual
contributor is the baseline and is represented at the 50th percentile.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved. 41
Figure 9. Progressive increase (decrease) in scores with level.
0%
10%
20%
30%
40%
50%
60%
70%
80%
Advanc
emen
t driv
e
Role pre
fere
nces
Situat
iona
l self
-awar
enes
s
Self-a
waren
ess
Menta
l agilit
y
People
agilit
y
Chang
e ag
ility
Resul
ts a
gility
Optimism
Persis
tenc
e
Focus
Toler
ance
of a
mbig
uity
Microm
anag
ing
Close
d
Volatile
Asser
tiven
ess
Individualcontributor/professional
First levelleader
Mid-levelleader
Functionalleader
Business or organizational unit/division leader
Senior/top functional executive
Top business or organizational group executive
A note regarding Capacity: The Raven’s APM Version 2 (NCS Pearson, 2007) was chosen as the measure of Capacity because it is the only effectively language-free test available, a feature important to global clients. As is true of all tests of cognitive ability, the Raven’s APM Version 2 carries with it an elevated possibility of adverse impact against traditionally disadvantaged ethnic groups. It also has a modest negative correlation with age. This is why Raven’s scores go down with level, rather than up, as shown in Table 8.
Table 8. Raven’s APM effect sizes for different position levels.
(Individual contributor N > 440)
Individual contributor/professional to First level leader(N = 1299)
Individual contributor/professional to Mid-level leader(N = 1275)
Individual contributor/professional to Functional leader(N = 1297)
Individual contributor/professional to Business or organizational unit/division leader (N = 1110)
Individual contributor/professional toSenior/top functional executive(N = 864)
Individual contributor/professional to Top business or organizational group executive (N = 265)
Raven’s APM Total Score
-.37 -.50 -.65 -.69 -.77 -.65
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved.42
Reaching a minimum threshold for raw reasoning ability relative to other leaders is considered desirable for
leadership success and the Raven’s test can provide information in that area. However, the characteristics
mentioned above, and concern for clients’ desire to meet diversity and inclusion goals leads us to offer the
inclusion of Capacity in the KFALP as optional, and it is each client’s decision whether or not to do so.
Accuracy of classificationKFALP scale scores differentiate those who have repeatedly advanced from those who have not advanced. We
are able to use those scores and logistic regression to demonstrate the ability of KFALP scores in aggregate to
accurately classify people into their correct organization level using only those scores.
The analysis uses scores from 19 KFALP scales. Career planning is excluded because it is not a normed measure.
Core experience is based on the number of levels in which a leader has served a significant period of time and
is not included because it is so closely linked to level attainment. Perspective and Key challenges Experience
scores have been pre-processed for this analysis to statistically control for years of full-time employment. For
this analysis, they represent the range of experiences a person has gained, holding time constant. The analysis
displays classification with and without Raven’s APM.
In Table 9 below, there are 584 persons to be classified: 354 Individual contributors/ professionals and 230 Top
business or organizational group executives. In the Raven’s APM condition, among Individual contributors, 323 of
354 were correctly classified. Among Top business or organizational group executives, 182 of 230 were correctly
classified. Overall correct classification rate was 86%. In the case of KFALP without Raven’s APM, among Top
business or organizational group executives, 176 of 230 were correctly classified, and among the 354 Individual
contributors/ professionals, 322 were correctly classified. Overall correct classification was 84%. These are high
levels of correct classification.
Table 9. Accuracy of level classification using KFALP data only.
KFALP with Raven’s APM KFALP without Raven’s APM
Actual position Actual position
Individual contributor/professional
Top business or organizational
group executive
Individual contributor/professional
Top business or organizational
group executive
Pre
dic
ted
p
osi
tio
n
Individual contributor/professional
323 48 322 59
Top business or organizational group executive
31 182 32 171
Note: A statistic is available to characterize classification accuracy while taking chance into consideration.
It is Cohen’s kappa (Cohen, 1960). Kappa is defined as K = Po – Pe
1 – Pe , where Po is defined as the observed agreement, and Pe is defined as the expected probability of chance agreement. By using this formalization of classification, chance agreement is removed by “conditioning on the marginal distribution for the two classifications” (Thompson, 1990, p. 357). In Cohen’s kappa, a value of 0 is equal to chance classification. A value of 1 is equal to perfect classification. For the classifications above, kappa = .72 with Raven’s APM, and .67 without Raven’s APM. These values are regarded very good classification, knowing only the scores on KFALP.
Prediction of Work engagementWork engagement, the amount of discretionary effort a respondent is willing to expend toward their work,
is a critical component of leadership performance. In a recent meta-analysis, Work engagement has been
demonstrated to provide substantial prediction of task and contextual work performance (Mr = .43 and .34,
respectively), as well as prediction of organization commitment (Mr = .31) (Christian, Garza, & Slaughter, 2011).
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved. 43
In Table 10 below, correlations of KFALP scales and Work engagement are displayed. In addition to the
association with advancement described above, many of the scales included in the KFALP provide economically
valuable prediction of Work engagement. This indicates that, in addition to differentiating by organizational
level, many scales also differentiate those who are likely to be better, committed performers because they are
more engaged in their work. For Work engagement, correlations are reported both raw and corrected for the
reliability of the criterion variable (Spearman, 1904). With the exception of Raven’s APM and Work engagement,
all correlations are statistically significant p < .05.
Multiple regression using Work engagement as the dependent variable and the 19 scales in Table 10 below as the
predictors indicates that KFALP variables account for a meaningful part of the variance in Work engagement,
multiple R = .433, (F [18, 6429] = 82.387, p <.000).
Table 10. Correlations of KFALP scales with Work engagement.
Work engagement,r’tt = .72
N raw corrected
Advancement drive 10621 .23 .28
Role preferences 10621 .04 .06
Core experience 10621 .13 .17
Perspective 10621 .09 .11
Key challenges 10621 .14 .18
Self-awareness 10621 .08 .11
Situational self-awareness 10621 .05 .08
Mental agility 10621 .19 .25
People agility 10621 .13 .17
Change agility 10621 .22 .28
Results agility 10621 .36 .47
Optimism 10621 .16 .20
Persistence 10621 .22 .27
Focus 10621 .04 .05
Tolerance of ambiguity 10621 .22 .27
Assertiveness 10621 .17 .20
Raven’s APM 6451 -.01 -.01
Micromanaging 10621 -.09 -.11
Closed 10621 -.19 -.24
Volatile 10621 -.10 -.13
FairnessAn important question to examine is how various sub-groups score on assessment tools. This helps to anticipate
the expected effect of using the tools on the demographics of the workforce. Fairness of assessments is a
markedly important objective at Korn Ferry, and assessments are designed not to disadvantage any group.
Adverse impact occurs when employee selection procedures used in making employment decisions have the
effect of selecting persons belonging to a historically disadvantaged group at a rate that is substantially lower
than that of the group with the higher selection rate. Adverse impact may occur due to the characteristics of an
assessment tool or other components included in the selection process, or, due to characteristics of the labor
pool, recruitment practices, or other process factors.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved.44
Korn Ferry has carefully evaluated the scales in KFALP for the potential of adverse impact using 50th percentile
thresholds such as included in KFALP reports. A typical way of describing the potential for adverse impact is
in terms of effect size comparing individuals from historically disadvantaged groups with the majority group.
An effect size can be interpreted as a small, medium, or large difference in average score. A commonly used
interpretation is as follows: an effect size of +/- 0.2 is considered a small effect, +/- 0.5 a medium effect, and
+/- 0.8 a large effect (Cohen’s δ; Cohen, 1988).
Our goal is to keep group differences to a minimum. To place the effort in context, a review of the literature
(Hough, Oswald, & Ployhart, 2001) describes cognitive ability test effect sizes of up to -1.0, resulting in
substantial disadvantage to some minority groups. By contrast, non-cognitive, or trait-based measures, tend to
have far smaller effect sizes, with most near 0 and some ranging up toward absolute values of .30. In general,
these are far smaller effect sizes than cognitive tests. The KFALP uses one cognitive ability test, Raven’s APM,
as a measure of Capacity. We rely on tests of non-cognitive characteristics and experience for all other scales
in KFALP. In general, with standard and reasonable uses of assessments, Cohen’s δ effect sizes having absolute
values ≤ .25 are unlikely to provide either substantial advantage or disadvantage for any group. Mean and
median (Cohen’s δ) effect sizes for KFALP scale scores are displayed in Table 11 for trait-based and experience
scales. KFALP aggregate scores typically produce small or negligible overall effect sizes across gender and
ethnic groups, and aggregated effect sizes are never > +/- .09 for any scale type and are typically considerably
lower. An effect size of -.09 is roughly equivalent to an impact ratio of .95 with mean cut scores. This means that
the lower scoring group will be selected at a rate of 95% the rate of the higher scoring group, a rate far above
the 80% 4/5ths rule threshold.
Table 11. Mean and median effect sizes for ethnicity and gender contrasts across KFALP.
Score group African American Hispanic-Latino Asian Female
Mean δ Median δ Mean δ Median δ Mean δ Median δ Mean δ Median δ
Trait-based -.07 -.09 -.02 .02 -.07 -.05 -.03 -.02
Experience -.05 -.03 -.08 -.04 -.03 -.02 -.09 -.08
Note. White/Caucasian participants served as the reference group for the ethnic contrasts. Effect sizes are Cohen’s δ.
KFALP scales, when evaluated in aggregate as intended, are unlikely to disadvantage any traditionally
disadvantaged group to a significant degree.
Our examination of adverse impact is ultimately far more detailed than the aggregate findings shown in Table 11.
In the subsections that follow, we describe our in-depth examination and explicate related findings.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved. 45
Data sourcesAll KFALP data for which individuals reported Ethnicity and Gender were used in the analysis in order to
maximize sample sizes of underrepresented groups. Gender and Ethnicity were the primary variables of interest
in this analysis. In each analysis, though sample sizes vary, the maximum available cases with complete data
were used in order to maximize our ability to arrive at stable inferences. Reporting of ethnicity is required only in
the US and is optional for participants. Sample sizes are reported below in Table 12.
Table 12. Sample sizes for group differences analyses for ethnicity and gender across KFALP measures.
N
White/Caucasian 1172
Hispanic-Latino 103
African American 41
Asian 81
Male 919
Female 526
Analytic strategyData were analyzed using a repeated measures multivariate analysis of covariance (MANCOVA). KFALP
assessment scores served as dependent variables in each analysis. Management level, full-time work experience
(years), categorical education, categorical employer size, categorical industry, and categorical job function were
included as covariates. These served to isolate the effects of gender and ethnicity and avoid spurious findings
or non-findings based on related omitted variable bias and/or unequal cell sizes. A single simultaneous gender
and ethnicity analysis was conducted to isolate effects. Two post-hoc analyses were completed, one for gender
and one for ethnicity. Omnibus between-group main effects for gender and ethnicity on centroids, and omnibus
profile parallelism (gender x KFALP vector and ethnicity x KFALP vector interactions) are examined first. In
the case of significant omnibus tests, planned post-hoc pairwise contrasts of least squares adjusted means
are pursued and examine whether historically disadvantaged groups differ from gender (male) and ethnicity
(Caucasian) reference groups. Pooled two-sample z-tests (2SD) were employed to examine whether pairwise
contrasts were significant (p ≤ .05). We report standard deviation unit discrepancies (Cohen’s δ) from reference
groups and the 4/5ths “impact ratio” (IRA) consistent with EEOC guidelines.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved.46
KFALP resultsMANCOVA results can be examined in Table 13 and show effects for several control variables on the centroid.
Management level (F [1, 1386] = 97.179, p < .001), education (F [7, 1386] = 6.061, p < .001), company size
(F [4, 1386] = 5.849, p < .001), industry (F [17, 1386] = 1.683, p < .040), function (F [21, 1386] = 3.963, p < .001),
and years of experience (F [1, 1150] = 30.764, p < .001) are each significantly associated with centroid differences.
Gender showed an effect on centroids (F [1, 1386] = 7.661, p < .006), however, ethnicity did not impact centroids
significantly (p = .103). Omnibus tests of parallelism show that Management level (F [19, 1368] = 32.966, p < .001),
education (F [133, 9618] = 2.138, p < .001), company size (F [76, 5484] = 1.670, p < .001), industry
(F [323, 23528] = 1.158, p < .027), function (F [399, 26334] = 1.768, p < .001), and years of experience
(F [19, 1368] = 16.302, p < .001) all have nonparallel lines.
Gender and ethnicity profiles are also not parallel, (F [19, 1368] = 1.523, p < .001) and
(F [114, 8238] = 1.523, p < .001), respectively.
Table 13. MANCOVA sample sizes for group differences analyses for ethnicity and gender across KFALP measures.
Terms df F p Partial h2
Equal levels, between-groups main effect
Management level x Centroid 1 97.179 .000 .066
Ethnicity x Centroid 6 1.765 .103 .008
Gender x Centroid 1 7.661 .006 .005
Education x Centroid 7 6.061 .000 .030
Company size x Centroid 4 5.849 .000 .017
Industry x Centroid 17 1.683 .040 .020
Function x Centroid 21 3.963 .000 .057
Full-time employment x Centroid 1 30.764 .000 .022
Flatness, within-groups main effect 19 2.558 .000 .034
Parallelism, multivariate interactions
Management level x Profile 19 32.966 .000 .314
Ethnicity x Profile 114 1.523 .000 .021
Gender x Profile 19 6.981 .000 .088
Education x Profile 133 2.138 .000 .029
Company size x Profile 76 1.670 .000 .023
Industry x Profile 323 1.158 .027 .016
Function x Profile 399 1.768 .000 .026
Full-time employment x Profile 19 16.302 .000 .185
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved. 47
Post-hoc contrasts for gender reveal seven significant effect sizes/standardized mean differences, as shown
in Table 14. Whether significant or not, however, only one pairwise contrast exceeds the 4/5ths conventional
threshold beyond which differences are seen as practically problematic; all other impact ratios are > .85 for
females. On eight scales, the impact ratio favors females. All but one effect size is ≤ .25. Trait-based mean
and median effect sizes are -.03 and -.02 respectively, indicating very small gender differences in aggregate.
Experience mean and median effect sizes are -.09 and -.08 respectively, perhaps indicating small differences
between men and women in the opportunity to gain experiences.
Table 14. Gender contrasts on KFALP variables.
Females N = 526
Scale ES p IR 2SD
Advancement drive -0.13 .016 0.90 -1.99
Role preferences 0.02 .728 1.02 0.29
Core experience -0.08 .106 0.93 -1.40
Perspective -0.05 .353 0.96 -0.76
Key challenges -0.14 .006 0.88 -2.41
Self-awareness 0.02 .753 1.01 0.26
Situational self-awareness -0.17 .001 0.87 -2.68
Mental agility -0.11 .031 0.91 -1.76
People agility 0.08 .110 1.07 1.29
Change agility -0.19 .000 0.85 -2.95
Results agility -0.08 .138 0.94 -1.22
Focus 0.00 .984 1.00 0.02
Persistence 0.02 .736 1.01 0.28
Tolerance of ambiguity -0.18 .001 0.87 -2.70
Assertiveness -0.10 .061 0.92 -1.53
Optimism 0.07 .172 1.06 1.13
Raven’s APM -0.06 .277 0.95 -0.92
Volatile 0.29 .000 0.78 4.58
Micromanaging 0.00 .973 1.00 0.03
Closed -0.02 .671 1.02 0.35
Note. Male participants served as the reference group (n = 919). ES = Cohen’s δ effect size. IR = Impact ratio, 2SD = Pooled two-sample z-test.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved.48
We examine contrasts between Caucasians and all other ethnic groups on each scale in Table 15. Ten of the
60 pairwise contrasts had p < .05. Of these, two favored the traditionally disadvantaged group. Of the eight
remaining, only three suggested a potential impact ratio < .80 for a historically disadvantaged group. Raven’s
APM for African Americans had an effect size of -.80 and an impact ratio of .60, values consistent with the
literature on cognitive tests. Focus for Asians and Micromanaging for African Americans just exceeded the
impact ratio threshold at .79 and .78 respectively, representing effect sizes of -.26 and -.27. The mean and median
effect sizes for traditionally disadvantaged groups were -.05 and -.07 for trait-based scales and -.05 and .02 for
Experience scales. When evaluated in aggregate, as recommended, KFALP results have generally very low group
differences and are unlikely to substantially disadvantage any group.
Table 15. Ethnicity contrasts on KFALP variables.
African American N = 41 Hispanic-Latino N = 103 Asian N = 81
Scale ES p IR 2SD ES p IR 2SD ES p IR 2SD
Advancement drive -0.22 .144 0.91 -2.61 -0.16 .098 0.85 -2.66 -0.11 .324 0.92 -1.52
Role preferences -0.17 .264 0.89 -3.13 0.04 .717 1.03 0.51 0.04 .693 1.04 0.63
Core experience 0.12 .436 1.03 0.77 -0.02 .804 0.97 -0.45 -0.02 .854 0.98 -0.31
Perspective 0.03 .850 0.98 -0.60 -0.04 .681 0.97 -0.57 0.07 .540 1.06 1.10
Key challenges -0.28 .059 0.85 -4.32 -0.19 .054 0.82 -3.32 -0.14 .190 0.88 -2.21
Self-awareness 0.49 .001 1.15 3.12 0.13 .189 1.13 2.19 -0.10 .360 0.92 -1.43
Situational self-awareness
0.25 .102 1.11 2.41 0.08 .434 1.07 1.21 -0.05 .627 1.00 0.10
Mental agility -0.02 .869 0.98 -0.46 -0.24 .015 0.81 -3.34 0.01 .928 1.04 0.79
People agility 0.06 .686 1.11 2.48 -0.20 .044 0.83 -3.09 -0.04 .714 0.97 -0.63
Change agility -0.07 .621 0.93 -2.08 0.02 .831 1.02 0.34 -0.01 .944 0.99 -0.13
Results agility -0.24 .105 0.90 -2.93 0.08 .433 1.07 1.24 -0.01 .893 0.99 -0.23
Focus -0.16 .275 0.91 -2.56 -0.04 .685 0.97 -0.61 -0.26 .016 0.79 -3.67
Persistence 0.12 .413 1.03 0.71 0.16 .104 1.15 2.65 0.14 .198 1.12 2.07
Tolerance of ambiguity
-0.31 .038 0.85 -4.42 -0.21 .028 0.81 -3.45 -0.23 .034 0.82 -3.31
Assertiveness -0.09 .560 0.95 -1.37 0.13 .192 1.11 1.95 -0.17 .124 0.85 -2.67
Optimism -0.20 .188 0.89 -3.14 0.16 .096 1.14 2.53 -0.17 .122 0.85 -2.61
Raven’s APM -0.80 .000 0.60 -12.30 -0.12 .220 0.91 -1.67 -0.11 .321 0.92 -1.51
Volatile 0.44 .004 1.38 6.93 -0.22 .025 1.21 3.58 0.05 .670 1.04 0.70
Micromanaging -0.27 .071 0.78 -4.40 0.06 .507 0.94 -1.04 -0.19 .080 0.86 -2.53
Closed -0.03 .828 0.98 -0.49 -0.07 .454 1.07 1.28 0.07 .503 1.07 1.16
Note. Caucasian participants served as the reference group (n = 1172). ES = Cohen’s δ effect size. IR = Impact ratio, 2SD = Pooled two-sample z-test.
NormsKorn Ferry Assessment of Leadership Potential (KFALP) norms have expanded since the inception of the
assessment. There are now six target level norms used in scoring the KFALP. They are defined briefly in Table 16
below. Complete norm demographic descriptions are found in Appendix C “Norm descriptions.”
Current KFALP norms are global, with approximately 70% of persons included in the norms working outside of
North America. The approximate split is one-third North America, one-third Europe/Middle East, and one-third
Asia/Pacific. This varies somewhat by level and is detailed in Appendix C.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved. 49
Clients should carefully choose a target level, typically two to three levels up from the current level. The goal is
to provide differentiation of talent which requires a relatively “high bar.” There is little value in specifying First
level leader on down as a target level.
Individual contributors are not leaders, so no Individual contributor target level norm is available. Individual
contributor norms are used for scoring Experience for current Individual contributors. For most leaders and
experienced Professional/Individual contributors, the Team lead and First level leader target levels are probably
not appropriate. However, where entry level Individual contributors are the population, those norms may be
appropriate. However, a more demanding, higher level norm is recommended.
Table 16. Target level definitions.
Target level Definition Norm
Chief executive officer/top organizational executive
Serves as the senior most leader of a business or organization. In the public sector, includes the Director, Secretary, or Administrator of an agency or department.
Chief executive officer/top organizational executive
Top business or organizational group executive
Serves as the senior most leader of a large business line or group of business lines (e.g., service lines, product lines, divisions, regions), or as a C-suite executive (e.g., Chief Financial Officer, Chief Operating Officer). In the public sector, includes senior executives who lead large groups of service lines or products directly related to the mission, or top-tier senior executive roles (e.g., Deputy Director, Under Secretary, or Chief Operating Officer of an agency or division).
Top business or organizational group executive
Senior/top functional executive
Serves as the senior most leader of a business support or mission support function (e.g., Finance, HR, Facilities).
Senior/top functional executive
Business or organizational unit/division leader
Leads a line of business (e.g., service line, product line, division, region). In the public sector, includes senior executives who lead large service or product groups directly related to the agency mission.
Business or organizational unit/division leader
Functional leader Leads a function (e.g., Accounting, Compensation & Benefits, Marketing, IT), as opposed to a line of business (e.g., service line, product line, region). In the public sector, includes senior executives who play key leadership roles within mission support groups (e.g., Finance, HR, Procurement, Security).
Functional leader
Mid-level leader Supervises one or more first-line supervisors/managers. Mid-level leader
First level leader Guides the direction of a group or team of Individual contributors and serves as their formal manager/supervisor.
First level leader
Team lead Guides the direction of a team of Individual contributors, or serves as its key subject-matter expert, without having formal supervisory/managerial responsibilities for team members.
First level leader
Individual contributor/professional
Supports a business or organization without supervising others as part of his/her role.
Individual contributor/professional (Experience only)
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved.50
Appendix A. Frequently asked questions
1. What is the research behind Item Response Theory (IRT)?It is a complex technology. In Classical Test Theory (CTT), for self-report personality assessment, it is typical to
use Likert ratings (responding on an item or question on a 1-to-5 or 1-to-7 ordered scale, e.g., Strongly agree to
Strongly disagree) despite a number of limitations:
• Response styles (e.g., acquiescence bias, social desirability bias, preferring the center of the scale vs. the extremes of the scale).
• Scale anchors may be interpreted differently by respondents.
• Intentional faking in an effort to “game” cannot be controlled.
• Interpretation of scores is dependent on a particular norm and on the specific scale content.
• The estimated reliability and true score estimate error are the same for all participants.
• Item difficulty is not taken into account in scoring.
Tests based on IRT can overcome these issues by:
• Removing response styles and scale anchor issues by eliminating the rating scale.
• Thwarting faking by eliminating the transparency of the Likert rating scale.
• Freeing interpretation from specific scale content by estimating true trait level.
• Providing true score estimates of error at the individual level.
• Taking item content into consideration when estimating true score, thus improving fidelity of the score to the trait construct.
Methodologies developed at the University of Barcelona by Brown and Maydeu-Olivares (2011) have opened
up an IRT methodology for forced-choice items formats using a Thurstonian paired-comparison measurement
model. This model presents items in multi-item blocks (2 or more items) and asks respondents to rank the
items. Traditionally, such response formats in CTT produced results that were ipsative (items dependent on one
another) with clearly detrimental effects. The state-of-the-art FC-IRT overcomes these issues.
The FC-IRT method of scoring uses a two-stage approach. It begins with applying mathematical model for
preferences. If a person prefers or ranks item 1 higher than item 2, then that person’s preference for item 1 is
higher than item 2. For a given block of four items, there are six possible comparisons to make between the
items. After a series of blocks are administered, we can use the information provided by these comparisons to
fit an IRT model. We first use the IRT model to obtain information about how difficult the items are to endorse
(called item difficulty), and how well the items relate to the traits being measured (called item loadings).
Then, using these two pieces of information, we can accurately estimate a person’s true score for each of the
underlying traits measured by the assessment. By abstracting the ranking information into two separate yet
linked statistical models, this two-stage approach allows us to take forced-choice item responses and produce
normative scale scores free of the ipsative limitations of traditional forced-choice measurement. Using Forced-
Choice Item Response Theory, we can take advantage of the bias and faking resistance of the forced-choice item
response format while eliminating the psychometric limitations of the classical test theory scoring methods.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved. 51
Korn Ferry scientists completed extensive simulations and pilots to:
• Test the underlying assumptions of the model and methods.
• Ensure that the new model did indeed overcome the prior psychometric insufficiencies of forced-choice methods.
• Develop extensions of the model to test the effects of intra- and inter-block “item cross talk” to better understand when item parameters may not be robust.
• Ensure that error estimates are meaningful.
• The results from these statistical simulations and from live pilot data affirmed that the FC-IRT methodology is a superior technology in the ways described above.
2. Why do we select levels based on current vs. target vs. aspiration categorizations? How do each of these show up in reporting? Whose responsibility is it to select each categorization?
KFALP is an assessment of the capacity and interest to develop the qualities required for effective performance
in significantly more challenging leadership roles. Inherent in that is the assumption that this potential will play
out over many years or even decades.
A person’s Aspiration is their current idea of what they would like to be doing someday. The person provides
this as part of the assessment. It does not influence scoring, rather, it provides information useful in feedback.
Today’s aspirations can change, sometimes quickly, with regard to level aspirations, mobility, and areas of
professional interest. They are often linked to current situation such as children’s education needs or the needs
of other family members.
A person’s Current Level is simply where they are situated today. It is used only to provide norms for
Experience scales. This is because it would be unreasonable to expect persons at a First Level Leader level to
have experiences normally gained further along in a career. Current level defaults to the level provided by the
participant during the assessment, however, it can also be set for a group of participants by the client.
Target Level provides a common reference norm for clients to use to compare participants. It is important
to select a target level that is two or three levels above the person’s current level. This sets a high bar and
contemplates the possibility—even the likelihood—of that person advancing two or three or more levels beyond
current level over the course of several years.
3. Do the Key challenges remain the same in the instrument regardless of the target level selected by the client?
Yes, the Key challenges remain the same. It is the amount and depth of participation in such challenges that vary
with level.
4. Why are scores on viaEDGE® and KFALP different for Learning agility?By any standard, though different technologies, the two versions of Learning agility are quality measures of the
same traits, and they share a very large majority of their items.
There are several possible causes of observed differences that may apply to a specific circumstance:
1) Though both measures are quite solid measures of Learning agility, different response scales are used in viaEDGE® (Likert) and KFALP (IRT)—they are different technologies. In large samples, Likert and IRT scores are correlated very highly (> .75)—as highly as can be expected by the reliabilities of the two measures. Still, since they are not perfectly correlated, a noticeable number of scores will not align. Even with this high correlation, approximately 10% of persons in the top quartile will drop to the bottom half of scores, and vice versa. This is normal, even for first-class measures such as these.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved.52
2) Normal error of measurement. Normal error of measurement comes into play when comparing scores on two similar assessments, or on repeated use of the exact same assessment—even top-notch assessments like viaEDGE® and KFALP. You can expect to see:
• Exactly the same score is extraordinarily rare.
• Scores just a few points apart will actually be rare as well.
• In any large-sample administrations of the two instruments, a large majority of scores will be similar, that
is, within about a fourth of the full range of raw scores from each other. This is all that can be expected
from even state-of-the-art assessments.
• A smaller number will naturally vary to a degree greater than that.
• A few will be wildly different, just due to measurement error. This is a normal part of using assessments.
3) Percentiles can mislead by exaggerating the difference in raw score differences. Percentiles are very desirable to explain what an individual score means to the person (70 means you scored better than 70% of the people in the norm). However, the dark side of percentiles is that they very much exaggerate or stretch the raw score differences, especially toward the middle of the natural normal distribution—where most persons’ raw scores are. Percentiles are intended to help understanding of single scale scores for individuals, and it accomplishes that objective. However, it is a source of misunderstanding when comparing two or more scores.
4) Different norms are used in viaEDGE® (single norm) and KFALP (level-specific norms). Different norm groups will result in different reported scores.
The bottom line is that each score is an estimate of true score that contains error. No one can determine which
one is more “right” for any individual; however, reliability of KFALP scales is slightly higher than the viaEDGE®
scales. That is one reason we are switching to that technology.
5. Can client organizations choose to opt out of Capacity while using the KFALP?Yes. Thus, only six signposts will be reported in the reports, rather than seven.
6. What signposts are likely to stay constant over time for an individual? What are the biggest areas to see the most change?
None are constant, as all can change over time, but at different rates.
Experience is the most directly malleable through proactively seeking broader experiences.
Capacity, on the other hand, is probably the one that is least malleable. The Raven’s APM measures the raw core
of reasoning. However, as people move through their careers, they rely less and less on narrow raw reasoning
and more and more on domain knowledge and learned heuristics for judgment and success. Also, the people we
assess are already in the top third to top half of the cognitive ability distribution.
Drivers can change with career and life stage as leaders decide to pursue more or less challenge or try a new
path and discover a new passion.
The Awareness, Learning agility, Leadership traits, and Derailment risks signposts are trait-based. Traits change
within an individual as they mature, faster than Capacity, but perhaps more slowly than Experience or Drivers.
It is important to remember that personality is not fixed and does not directly determine behavior. It has a big
influence on behavior, as does work situation, learned self-regulation skills, fatigue, and other state factors.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved. 53
7. What, if any, differences in norms have we seen in regards to different regions, countries, cultures? Do results look similar across different regions, countries, or cultures?
Regions are in some sense arbitrary and we do not offer Region-based norms. We may offer geography-based
norms at some point in the future if sufficient sample sizes demonstrate meaningful difference may occur in
decisions made about participants.
We are keenly interested, though, in understanding any differences across cultures. We expect the FC-IRT
technology to limit differences due to cultural differences in use of a rating scale—eliminating culturally-based
response sets—which we know plague classical test theory assessments.
In general, based on very early analysis:
• Differences appear small, that is, there are far greater differences between persons within culture than
between averages for different cultures.
• Geographically-based differences are generally small relative to other important factors (management
level).
• Profiles are quite similar across geographies.
We await larger samples across more countries to adequately represent any differences found. We expect to
launch a research program on these issues as soon as sufficient data are available.
8. What qualifies signposts to be categorized as Strong (green), Needs development (yellow), or orange?In this case “Strong” is relative to the target level (except Experience)—usually a high bar.
Signposts are marked “green” based on the number of sub-dimensions on which an individual has reached a
green threshold:
• Drivers – any two of three sub-dimensions.
• Experience – any two of three sub-dimensions.
• Awareness – both sub-dimensions.
• Learning agility – any three of four sub-dimensions.
• Leadership traits – any four of five sub-dimensions.
• Capacity – one (of one) sub-dimension.
• Derailment risks – none of the three Derailment risks factors.
If these qualifications aren’t met, then you score Needs development (yellow or orange).
The green and yellow signpost markers are provided as an “at a glance” view of the participants’ results. They
are also used to assist in sorting the Talent Grid Report. No signpost appearing yellow should be considered on its own as a “knock out” factor for consideration of potential. The KFALP data must be considered as a whole-
person view and in conjunction with everything job-relevant that is known about the person.
Users should look further than the signposts. More interesting data are available at the scale level. At the scale
level, the meaning of yellow and green is whether or not the individual scored above or below the average of
persons who have already advanced to the target level. One should keep in mind that the scoring is against a
high target.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved.54
9. What is the calculation for the subscales that contribute to the ranking in the Talent Grid Report?1) Number of green signposts (7,6,5…), then ties broken by
2) Number of scales 50th percentile or above, then ties broken by
3) Number of scales 91st and above, except Derailment risks (that is the black square with the up arrow
icon), then ties broken by
4) Number of scales 50th to 90th, except Derailment risks, (that is the black square icon), then ties broken by
5) Number of scales 11th to 49th, except derailers (that is the clear square icon), then ties broken by
6) Number of scales 10th and below (that is the clear square with the down arrow icon), then ties broken by
7) Order in the report left to right.
The result is sort of an overall heat map with the generally higher scores rising.
10. Say more about the Capacity measure, why it is used, and its limitations.The Raven’s APM Version 2 was chosen because it is the only effectively language-free test available on the market, a feature important to global organizations.
The Raven’s APM predicts job performance when Problem solving is a main part of the job and when people must depend on their raw reasoning to do problem solving. In cases where persons have been pre-selected for problem solving or are already known to be of high mental ability, it does not predict so well. We have studies where it does predict and studies where it does not. The studies where it does not predict are with highly educated populations where success is based on what the leaders have learned more than raw reasoning. The studies where it does predict are in leadership populations less pre-selected and where leaders “worked their way up.”
Raven’s APM is a measure of raw reasoning ability. People rely less on this ability as they progress through their work life and more on heuristics, tacit knowledge, teams, etc., for problem solving. The degree to which raw reasoning is required will depend on the work a leader is expected to do.
As is true of all tests of cognitive ability, the Raven’s APM has the possibility of introducing adverse impact against historically disadvantaged groups. Adverse impact occurs only when employee selection procedures used in making employment decisions have the effect of selecting persons belonging to a historically disadvantaged group at a rate that is substantially lower than that of the group with the higher selection rate. Adverse impact may occur due to the characteristics of an assessment tool or other components included in the selection process, or, due to characteristics of the labor pool, recruitment practices, or other process factors.
Use Raven’s scores only as a starting point for consideration of a person’s problem-solving skills.
• Persons in the leadership pipeline typically are already among the top half of the cognitive ability distribution and the norms we use are a “high bar.”
• Persons may have developed extremely valuable tacit knowledge, judgment heuristics, or domain knowledge, or rely on their teams such that they perform at a higher problem-solving level than those who have not developed similar approaches—better even than those with higher Capacity scores.
• It is important to verify test scores through inquiring about what has been observed about the person’s practical problem-solving skills.
While an individual can over-depend on extremely high cognitive ability, there is really no evidence that such
high scores result in failure or derailment of leaders so long as other skills are present and used. And as always,
consider Problem solving just a part of each person’s tool kit. People are much more than their mental engine.
11. Does it pay to guess on the Raven’s APM?It does not hurt to guess, however, the odds of it helping are small, given that the likelihood of a correct pure guess
is only 1 in 8. For the entire test, guessing would, on average, get a score of 3—sometimes less, sometimes more. It
is assumed that any participant may narrow the options on an item and guess if they cannot resolve a final answer.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved. 55
Appendix B. Inter-method correlations
Table 17. Correlation of CTT and FC-IRT versions of KFALP trait scales.
N = 2022
Awareness
Self-awarenessSituational self-awareness
.62
.52
Learning agility
Mental agilityPeople agilityChange agilityResults agility
.69
.71
.80
.73
Leadership traits
FocusPersistenceTolerance of ambiguityAssertivenessOptimism
.76
.74
.78
.77
.71
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved.56
Appendix C. Norm descriptions
General norm descriptions for the Korn Ferry Assessment of Leadership Potential.
First level leader N = 2001
Mean age: 38.6, SD = 6.88
Gender: Males 65.6%, Females 34.4%
Highest level of education completed:
Secondary/High School Graduate 2.6%
Trade/Technical Education 1.9%
Associate’s Degree/Diploma 3.8%
Undergraduate Degree (Bachelor’s, etc.) 38.4%
Postgraduate Degree (Master’s, etc.) 36.7%
Doctorate/Professional (Ph.D., M.D., etc.) 5.9%
Primary industryAdvanced Technology 1.8%
Consumer Goods 9.7%
Distribution Services .6%
Education & Training 1.6%
Energy & Utilities 14.0%
Financial Services 9.9%
Government .3%
Healthcare & Biological Sciences 26.4%
Industrial & Manufacturing 16.8%
Media & Entertainment .9%
Nonprofit .0%
Professional & Business Services 3.1%
Real Estate & Property Management .2%
Research & Development .3%
Retail 1.9%
Telecommunications 10.0%
Travel, Hospitality & Leisure .7%
Multi-Industry Holding Companies 1.4%
Current functional areaAdministrative Services 1.2%
Communications 1.2%
Consulting 2.9%
Creative .1%
Customer Service 3.5%
Education & Training 1.4%
Executive & General Management .8%
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved. 57
Finance & Control 9.1%
Financial Services 2.5%
Healthcare 2.0%
Human Resources 10.6%
Information Technology 4.8%
Legal 1.5%
Manufacturing 2.9%
Marketing 6.9%
Operations 12.6%
Public Safety & Military .3%
Research, Development & Engineering 9.0%
Retail .2%
Sales 16.3%
Strategic Planning 3.5%
Multiple Functions 5.9%
Company typePublicly Traded 61.1%
Subsidiary of Publicly Traded 14.0%
Privately Held 16.8%
Government Organization-Agency 2.2%
Government Owned Enterprise 3.9%
Nonprofit 1.5%
Self-Employed .4%
Global regionAsia 19.1%
Caribbean .2%
Europe 26.6%
Latin America 7.3%
Middle East and North Africa 7.4%
North America 25.5%
Oceania 2.2%
Sub-Saharan Africa 1.7%
No Data 9.9%
Ethnicity (US only, optional)Hispanic or Latino 8.3%
White 71.1%
Black or African American 6.6%
Native Hawaiian or Other Pacific Islander .4%
Asian 9.3%
American Indian or Alaska Native .2%
Two or more races (more than one of the above) 4.0%
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved.58
Mid-level leader N = 1959
Mean age: 41.85, SD = 6.68
Gender: Males 66.8%, Females 33.2%
Highest level of education completed:
Some Secondary Education .5%
Secondary/High School Graduate 2.1%
Trade/Technical Education 1.6%
Associate’s Degree/Diploma 4.2%
Undergraduate Degree (Bachelor’s, etc.) 31.4%
Postgraduate Degree (Master’s, etc.) 40.4%
Doctorate/Professional (Ph.D., M.D., etc.) 5.6%
Primary industryAdvanced Technology 2.2%
Consumer Goods 11.5%
Distribution Services 2.1%
Education & Training 1.5%
Energy & Utilities 11.0%
Financial Services 8.8%
Government .3%
Healthcare & Biological Sciences 20.1%
Industrial & Manufacturing 14.3%
Media & Entertainment .9%
Nonprofit .2%
Professional & Business Services 3.4%
Real Estate & Property Management .6%
Research & Development .2%
Retail 4.4%
Telecommunications 16.6%
Travel, Hospitality & Leisure .6%
Multi-Industry Holding Companies 1.4%
Current functional areaAdministrative Services 1.0%
Communications 1.3%
Consulting 2.9%
Creative .5%
Customer Service 2.5%
Education & Training 1.0%
Executive & General Management 2.7%
Finance & Control 9.4%
Financial Services 1.7%
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved. 59
Healthcare 1.8%
Human Resources 13.1%
Information Technology 4.6%
Legal 1.9%
Manufacturing 3.6%
Marketing 8.6%
Operations 12.7%
Public Safety & Military .2%
Research, Development & Engineering 6.4%
Retail .9%
Sales 12.0%
Strategic Planning 3.9%
Multiple Functions 7.3%
Company typePublicly Traded 68.0%
Subsidiary of Publicly Traded 14.5%
Privately Held 13.4%
Government Organization-Agency 1.2%
Government Owned Enterprise 1.2%
Nonprofit 1.5%
Self-Employed .1%
Mean years in management: 10.47, SD = 6.44
Global regionAsia 15.0%
Caribbean .2%
Europe 22.9%
Latin America 10.5%
Middle East and North Africa 9.0%
North America 24.8%
Oceania 2.4%
Sub-Saharan Africa 1.5%
No Data 13.7
Ethnicity (US only, optional)Hispanic or Latino 6.5%
White 80.8%
Black or African American 4.0%
Native Hawaiian or Other Pacific Islander 0.0%
Asian 6.0%
American Indian or Alaska Native .4%
Two or more races (more than one of the above) 2.2%
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved.60
Functional leader N = 2218
Mean age: 43.89, SD = 6.97
Gender: Males 65.9%, Females 34.1%
Highest level of education completed:
Some Secondary Education .3%
Secondary/High School Graduate 1.2%
Trade/Technical Education 1.2%
Associate’s Degree/Diploma 4.4%
Undergraduate Degree (Bachelor’s, etc.) 28.1%
Postgraduate Degree (Master’s, etc.) 42.4%
Doctorate/Professional (Ph.D., M.D., etc.) 7.6%
Primary industryAdvanced Technology 2.7%
Consumer Goods 8.0%
Distribution Services .8%
Education & Training 2.2%
Energy & Utilities 9.4%
Financial Services 11.0%
Government .1%
Healthcare & Biological Sciences 23.7%
Industrial & Manufacturing 20.6%
Media & Entertainment 1.1%
Nonprofit .2%
Professional & Business Services 4.1%
Real Estate & Property Management 1.0%
Research & Development .2%
Retail 2.8%
Telecommunications 10.4%
Travel, Hospitality & Leisure .2%
Multi-Industry Holding Companies 1.5%
Current functional areaAdministrative Services .6%
Communications 1.4%
Consulting 2.2%
Creative .2%
Customer Service 2.0%
Education & Training .9%
Executive & General Management 2.5%
Finance & Control 11.1%
Financial Services 1.9%
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved. 61
Healthcare 1.4%
Human Resources 19.2%
Information Technology 5.4%
Legal 2.1%
Manufacturing 3.3%
Marketing 8.2%
Operations 12.1%
Public Safety & Military .2%
Research, Development & Engineering 6.2%
Retail .4%
Sales 5.9%
Strategic Planning 4.1%
Multiple Functions 8.6%
Company typePublicly Traded 68.1%
Subsidiary of Publicly Traded 12.6%
Privately Held 15.8%
Government Organization-Agency .7%
Government Owned Enterprise .4%
Nonprofit 2.3%
Self-Employed .1%
Mean years in management: 10.47, SD = 6.44
Global regionAsia 17.3%
Caribbean .2%
Europe 20.1%
Latin America 8.6%
Middle East and North Africa 4.1%
North America 28.9%
Oceania 5.1%
Sub-Saharan Africa 1.2%
No Data 14.5%
Ethnicity (US only, optional)Hispanic or Latino 6.0%
White 82.1%
Black or African American 3.7%
Native Hawaiian or Other Pacific Islander .2%
Asian 5.8%
American Indian or Alaska Native 0.0%
Two or more races (more than one of the above) 2.2%
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved.62
Business or organizational unit/division leader N = 2218
Mean age: 44.82, SD = 6.77
Gender: Males 73.2%, Females 26.8%
Highest level of education completed:
Some Secondary Education .4%
Secondary/High School Graduate 1.8%
Trade/Technical Education 1.7%
Associate’s Degree/Diploma 4.4%
Undergraduate Degree (Bachelor’s, etc.) 26.5%
Postgraduate Degree (Master’s, etc.) 41.7%
Doctorate/Professional (Ph.D., M.D., etc.) 7.4%
Primary industryAdvanced Technology 2.4%
Consumer Goods 6.3%
Distribution Services .8%
Education & Training 2.4%
Energy & Utilities 9.1%
Financial Services 12.9%
Government .4%
Healthcare & Biological Sciences 26.1%
Industrial & Manufacturing 13.6%
Media & Entertainment .6%
Nonprofit .6%
Professional & Business Services 4.6%
Real Estate & Property Management .6%
Research & Development .3%
Retail 3.4%
Telecommunications 13.6%
Travel, Hospitality & Leisure .9%
Multi-Industry Holding Companies 1.2%
Current functional areaAdministrative Services 1.2%
Communications .6%
Consulting 2.3%
Creative .4%
Customer Service 1.8%
Education & Training 1.2%
Executive & General Management 19.3%
Finance & Control 4.6%
Financial Services 3.3%
Healthcare 1.5%
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved. 63
Human Resources 4.8%
Information Technology 4.3%
Legal 1.3%
Manufacturing 3.1%
Marketing 6.6%
Operations 9.4%
Research, Development & Engineering 4.3%
Retail 1.6%
Sales 12.6%
Strategic Planning 3.4%
Multiple Functions 12.5%
Company typePublicly Traded 62.6%
Subsidiary of Publicly Traded 17.0%
Privately Held 15.1%
Government Organization-Agency .5%
Government Owned Enterprise .6%
Nonprofit 4.0%
Self-Employed .2%
Mean years in management: 13.77, SD = 7.03
Global regionAsia 13.4
Caribbean .3
Europe 23.2
Latin America 9.5
Middle East and North Africa 4.4
North America 24.4
Oceania 6.3
Sub-Saharan Africa 2.7
No Data 15.8
Ethnicity (US only, optional)Hispanic or Latino 6.5%
White 82.7%
Black or African American 2.2%
Asian 6.7%
Two or more races (more than one of the above) 1.9%
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved.64
Senior/top functional executive N = 1593
Mean age: 46.43, SD = 6.89
Gender: Males 70.1%, Females 29.9%
Highest level of education completed:
Some Secondary Education .3%
Secondary/High School Graduate .8%
Trade/Technical Education .9%
Associate’s Degree/Diploma 3.3%
Undergraduate Degree (Bachelor’s, etc.) 27.7%
Postgraduate Degree (Master’s, etc.) 51.0%
Doctorate/Professional (Ph.D., M.D., etc.) 7.8%
Primary industryAdvanced Technology 2.9%
Consumer Goods 6.5%
Distribution Services 1.4%
Education & Training 2.1%
Energy & Utilities 5.5%
Financial Services 11.7%
Government .4%
Healthcare & Biological Sciences 15.8%
Industrial & Manufacturing 14.8%
Media & Entertainment 2.4%
Nonprofit .8%
Professional & Business Services 3.1%
Real Estate & Property Management 1.3%
Research & Development .1%
Retail 3.4%
Telecommunications 25.2%
Travel, Hospitality & Leisure 1.6%
Multi-Industry Holding Companies 1.1%
Current functional areaAdministrative Services .6%
Communications 1.9%
Consulting 1.3%
Creative .4%
Customer Service 2.0%
Education & Training .9%
Executive & General Management 12.2%
Finance & Control 12.8%
Financial Services .9%
Healthcare .8%
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved. 65
Human Resources 12.7%
Information Technology 6.3%
Legal 3.8%
Manufacturing 1.9%
Marketing 5.9%
Operations 9.8%
Public Safety & Military .1%
Research, Development & Engineering 3.5%
Retail .2%
Sales 6.7%
Strategic Planning 4.3%
Multiple Functions 10.9%
Company typePublicly Traded 61.6%
Subsidiary of Publicly Traded 18.6%
Privately Held 15.3%
Government Organization-Agency .6%
Government Owned Enterprise .7%
Nonprofit 3.1%
Self-Employed .1%
Mean years in management: 15.77, SD = 7.3
Global regionAsia 8.3%
Caribbean 0.0%
Europe 17.1%
Latin America 24.0%
Middle East and North Africa 8.5%
North America 29.0%
Oceania 3.6%
Sub-Saharan Africa 1.3%
No Data 8.1%
Ethnicity (US only, optional)Hispanic or Latino 5.5%
White 86.5%
Black or African American 2.3%
Asian 4.0%
Two or more races (more than one of the above) 6.0%
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved.66
Top business or organizational group executive N = 595
Mean age: 46.97, SD = 6.57
Gender: Males 79%, Females 21%
Highest level of education completed:
Secondary/High School Graduate 1.3%
Trade/Technical Education .8%
Associate’s Degree/Diploma 1.8%
Undergraduate Degree (Bachelor’s, etc.) 26.4%
Postgraduate Degree (Master’s, etc.) 50.4%
Doctorate/Professional (Ph.D., M.D., etc.) 10.6%
Primary industryAdvanced Technology 6.1
Consumer Goods 5.9
Distribution Services 1.7
Education & Training 3.4
Energy & Utilities 5.0
Financial Services 12.6
Government 1.3
Healthcare & Biological Sciences 22.0
Industrial & Manufacturing 12.1
Media & Entertainment 1.8
Nonprofit .8
Professional & Business Services 2.5
Real Estate & Property Management 1.0
Research & Development .8
Retail 4.7
Telecommunications 15.6
Travel, Hospitality & Leisure 1.3
Multi-Industry Holding Companies 1.2
Current functional areaAdministrative Services .2
Communications 1.3
Consulting 1.2
Creative .3
Customer Service 1.2
Education & Training .5
Executive & General Management 38.8
Finance & Control 7.4
Financial Services 1.0
Healthcare 1.5
Human Resources 3.7
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved. 67
Information Technology 3.2
Legal 3.0
Manufacturing 1.5
Marketing 3.2
Operations 6.1
Public Safety & Military .2
Research, Development & Engineering 3.0
Retail 1.2
Sales 6.6
Strategic Planning 2.5
Multiple Functions 12.4
Company typePublicly Traded 50.6%
Subsidiary of Publicly Traded 22.5%
Privately Held 18.2%
Government Organization-Agency 1.2%
Government Owned Enterprise .8%
Nonprofit 6.2%
Self-Employed .5%
Mean years in management: 16.53, SD = 7.29
Global regionAsia 8.9%
Caribbean 0.0%
Europe 19.0%
Latin America 15.8%
Middle East and North Africa 11.3%
North America 31.9%
Oceania 3.7%
Sub-Saharan Africa 1.2%
No Data 8.2%
Ethnicity (US only, optional)Hispanic or Latino 3.1%
White 80.2%
Black or African American 4.9%
Native Hawaiian or Other Pacific Islander .6%
Asian 6.8%
Two or more races (more than one of the above) 4.3%
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved.68
Appendix D. Global Personality Inventory (GPI) definitions
Thinking Factor
ThoughT AgiliTyThis is a measure of the tendency to be open both to multiple ideas and to using alternative modes of thinking.
innovATion This is a measure of the tendency to produce unique and original things.
ThoughT FocusThis is a measure of the tendency to understand ambiguous information by analyzing and detecting the systematic themes in the data.
vision This is a measure of the tendency to have foresight in one’s thinking.
Planning and Execution Factor
ATTenTion To DeTAil This is a measure of the tendency to be exacting and precise.
Work Focus This is a measure of the tendency to be self-disciplined in one’s approach to work.
Facilitating Leadership Factor
TAking chArge This is a measure of the tendency to take a leadership role.
inFluence This is a measure of the tendency to get others to view and do things in a certain way.
Derailing Leadership Factor
ego cenTereD This is a measure of the tendency to be self-centered and appear egotistical.
MAnipulATion This is a measure of the tendency to be self-serving and sly.
Micro-MAnAgingThis is a measure of the tendency to over-manage once a person has advanced to higher levels of management.
inTiMiDATing This is a measure of the tendency to use power in a threatening way.
pAssive-AggressiveThis is a measure of the tendency to avoid confronting others, conveying acceptance or cooperation and yet appearing to behave in uncooperative and self-serving ways.
Interpersonal Factor
sociAbiliTy This is a measure of the tendency to be highly engaged by any social situation.
consiDerATion This is a measure of the tendency to express care about others’ well-being.
eMpAThyThis is a measure of the tendency to understand what others are experiencing and to convey that understanding to them.
TrusT This is a measure of the tendency to believe that most people are good and well intentioned.
sociAl AsTuTenessThis is a measure of the tendency to accurately perceive and understand the meaning of social cues and use that information to accomplish a desired goal.
energy level This is a measure of the tendency to be highly active and energetic.
iniTiATive This is a measure of the tendency to take action in a proactive rather than reactive manner.
Desire For AchieveMenT This is a measure of the tendency to have a strong drive to realize personally meaningful goals.
Self-Management Factor
ADApTAbiliTy This is a measure of the tendency to be open to change and considerable variety.
openness This is a measure of the tendency to accept and respect the individual differences of people.
negATive AFFecTiviTyThis is a measure of the tendency to be generally unsatisfied with many things, including but not limited to work.
opTiMisM This is a measure of the tendency to believe that good things are possible.
eMoTionAl conTrol This is a measure of the tendency to be even-tempered.
sTress TolerAnceThis is a measure of the tendency to endure typically stressful situations without undue physical or emotional reaction.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved. 69
selF-conFiDence This is a measure of the tendency to believe in one’s own abilities and skills.
iMpressing This is a measure of the tendency to try to make a good impression on others.
selF-AWAreness This is the tendency to be aware of one’s strengths and weaknesses.
inDepenDence This is a measure of the tendency to be autonomous.
coMpeTiTiveness This is a measure of the tendency to evaluate one’s own performance in comparison to others.
risk-TAking This is a measure of the tendency to take chances based on limited information.
Desire For ADvAnceMenTThis is a measure of the tendency to be ambitious in the advancement of one’s career or position in organizational hierarchy.
Collective Orientation Factor
inTerDepenDence This is a measure of the tendency to work well with others.
DuTiFulness This is a measure of the tendency to be filled with a sense of moral obligation.
responsibiliTy This is a measure of the tendency to be reliable and dependable.
Global Personality Inventory (GPI) definitions (continued)
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved.70
Appendix E. GPI and KFALP correlations
Ad
vanc
emen
t d
rive
Rol
e p
refe
renc
es
Situ
atio
nal s
elf-
awar
enes
s
Self-
awar
enes
s
Men
tal a
gili
ty
Peop
le a
gili
ty
Cha
nge
agili
ty
Res
ults
ag
ility
Op
tim
ism
Tole
ranc
e of
am
big
uity
Focu
s
Pers
iste
nce
Ass
erti
vene
ss
Mic
rom
anag
ing
Clo
sed
Vol
atile
Attention to Detail -.23 -.25 .03 .12 -.14 -.05 -.21 .08 -.10 -.26 .61 .17 -.02 .51 .21 -.06
Adaptability .37 .36 .08 -.01 .36 .04 .49 .16 .13 .49 -.24 .01 .13 -.34 -.52 -.09
Social Astuteness .17 .07 .10 .10 .20 .22 .13 .06 .05 .18 .02 .18 .10 -.17 -.20 -.13
Competitiveness .32 .10 -.01 .00 -.04 .01 .13 .18 .04 .10 .12 .17 .29 .14 -.03 -.08
Consideration .05 .05 .08 .02 .14 .25 -.02 .07 .03 .04 .04 -.03 -.01 -.20 -.08 -.02
Desire for Achievement .36 .22 .02 .07 .30 .05 .36 .36 .10 .35 .00 .16 .27 -.11 -.32 -.07
Desire for Advancement .29 .09 .01 -.06 .06 -.01 .14 .14 .02 .13 .15 .13 .20 .10 -.07 -.07
Dutifulness -.04 .01 .11 .05 .07 .01 .01 .10 .11 .03 .15 .12 -.01 .04 -.07 -.10
Emotional Control .11 .02 .23 .11 .21 .05 .17 .03 .19 .21 .04 .13 -.01 -.03 -.18 -.42
Ego Centered -.07 -.17 -.04 -.03 -.07 -.08 -.04 .07 -.15 -.10 .17 .04 .12 .20 .15 .04
Energy Level .35 .25 .04 .06 .18 .12 .32 .38 .18 .28 .01 .22 .23 -.16 -.31 -.07
Empathy .03 .01 .13 .05 .14 .32 -.04 .04 .00 .01 .07 .04 -.02 -.20 -.04 -.11
Impressing -.11 -.12 .06 .00 -.01 .08 -.16 .05 -.03 -.15 .35 .16 -.05 .18 .13 -.09
Independence -.16 -.07 -.01 -.10 -.03 -.20 -.13 -.02 -.12 -.09 .07 -.06 -.06 .21 .16 .01
Influence .27 .13 .14 -.01 .23 .37 .19 .17 .11 .19 .08 .21 .34 -.12 -.11 -.15
Initiative .21 .18 .06 .05 .24 .09 .28 .23 .08 .20 .12 .16 .29 -.06 -.25 -.05
Innovation .14 .12 .14 -.03 .40 .16 .24 .11 .01 .26 .01 .10 .14 -.10 -.21 -.13
Interdependence .08 .06 .07 .01 .13 .10 .15 .02 .06 .08 -.03 .06 .02 -.20 -.21 -.13
Intimidating -.11 -.06 -.03 -.01 -.14 -.07 -.03 -.04 -.09 -.05 .14 .10 .08 .19 .13 .00
Manipulation -.02 -.05 -.12 -.07 -.16 -.11 -.06 -.13 -.16 -.09 .03 -.13 .02 .17 .08 .08
Micro-Managing -.18 -.16 -.06 .04 -.16 -.11 -.24 .06 -.24 -.27 .34 .03 .09 .38 .29 .09
Negative Affectivity -.18 -.17 -.07 -.08 -.15 -.11 -.15 -.10 -.26 -.16 .08 -.15 -.06 .19 .19 .08
Openness .23 .17 .08 .04 .22 .19 .30 .15 .25 .20 -.04 .16 .14 -.21 -.32 -.16
Optimism .25 .18 .11 .03 .19 .18 .23 .22 .20 .22 .05 .22 .11 -.16 -.24 -.14
Passive-Aggressive -.28 -.25 -.10 -.05 -.20 -.14 -.29 -.15 -.18 -.27 .15 -.12 -.16 .26 .31 .13
Responsibility .08 .07 .14 .07 .14 .11 .10 .23 .13 .04 .12 .23 .04 .00 -.07 -.17
Risk-taking .30 .22 .05 -.03 .26 .06 .40 .15 .00 .32 -.07 .16 .18 -.16 -.31 -.07
Self-Awareness .09 .02 .22 .13 .10 .08 .08 .08 .12 .04 .09 .29 .13 .02 -.08 -.23
Self-Confidence .16 .12 .15 .01 .21 .05 .15 .24 .11 .12 .11 .17 .19 .05 -.16 -.15
Sociability .33 .18 .10 .06 .13 .34 .19 .15 .22 .14 .02 .16 .33 -.21 -.20 -.14
Stress Tolerance .17 .13 .26 .04 .11 .09 .24 .07 .30 .22 -.02 .19 .01 -.08 -.22 -.39
Thought Agility .04 -.01 .08 .06 .22 .11 .08 .04 .00 .11 .13 .19 .04 -.03 -.13 -.14
Taking Charge .31 .18 .11 .02 .15 .12 .16 .22 .07 .19 .14 .18 .53 -.01 -.11 -.13
Thought Focus .10 .05 .15 .08 .29 .10 .12 .16 -.01 .18 .07 .16 .12 -.04 -.08 -.17
Trust .10 .09 .10 .03 .14 .17 .02 .09 .18 .09 -.02 .10 .04 -.22 -.04 -.05
Vision .08 .04 .14 .04 .27 .15 .17 .13 -.05 .20 .07 .15 .11 -.05 -.13 -.18
Work Focus -.07 -.14 .21 .16 -.01 -.01 -.09 .15 .06 -.12 .35 .31 .04 .25 .05 -.29
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved. 71
References
Ackerman, P. L., & Heggestad, E. D. (1997). Intelligence, personality, and interests: Evidence for overlapping traits.
Psychological Bulletin, 121(2), 219-245. doi: 10.1037/0033-2909.121.2.219
Allen, T. J., & Katz, R. (1986). The dual ladder: Motivational solution or managerial delusion? R&D Management, 16(2),
185-197.
American Educational Research Association (AERA), American Psychological Association (APA), & National Council
on Measurement in Education (NCME). (2014). Standards for educational and psychological testing. Washington,
D.C.: AERA.
Amos, B., & Klimoski, R. J. (2014). Courage: Making teamwork work well. Group & Organization Management, 39(1),
110-128. doi: 10.1177/1059601113520407
Anderson, C., & Kilduff, G. J. (2009). Why do dominant personalities attain influence in face-to-face groups? The
competence-signaling effects of trait dominance. Journal of Personality and Social Psychology. 96(2), 491–503.
doi: 10.1037/a0014201
Atwater, L., Waldman, D., Ostroff, C., Robie, C., & Johnson, K. M. (2005). Self-other agreement: Comparing its
relationship with performance in the U.S. and Europe. International Journal of Selection and Assessment, 13(1),
25-40.
Baer, R. A. (2003). Mindfulness training as a clinical intervention: A conceptual and empirical review. Clinical
Psychology: Science and Practice, 10(2), 125.
Barrick, M. R., & Mount, M. K. (2005). Yes, personality matters: Moving on to more important matters. Human
Performance, 18(4), 359-372.
Barrick, M. R., Parks, L., & Mount, M. K. (2005). Self-monitoring as a moderator of the relationships between personality
traits and performance. Personnel Psychology, 58(3), 745-767. doi: 10.1111/j.1744-6570.2005.00716.x
Bartram, D. (2007). Increasing validity with forced-choice criterion measurement formats. International Journal of
Selection and Assessment, 15(3), 263-272. doi: 10.1111/j.1468-2389.2007.00386.x
Baum, J. R., & Locke, E. A. (2004). The relationship of entrepreneurial traits, skill, and motivation to subsequent venture
growth. Journal of Applied Psychology, 89(4), 587598. doi: 10.1037/0021-9010.89.4.587
Blinkhorn, S., Johnson, C., & Wood, R. (1988). Spuriouser and spuriouser: The use of ipsative personality tests. Journal
of Occupational Psychology, 61, 153-162.
Boselie, P., Dietz, G., & Boon, C. (2005). Commonalities and contradictions in HRM and performance research. Human
Resource Management Journal, 15, 67-94.
Boxall, P., & Purcell, J. (2008). Strategy and human resource management. Hampshire: Palgrave MacMillan.
Brousseau, K., Driver, M., Hourihan, G., & Larsson, R. (2006). The seasoned executive’s decision-making style. Harvard
Business Review, 84, 109-121.
Brown, A. (2016). Item response models for forced-choice questionnaires: A common framework. Psychometrika, 81(1):
135-60.
Brown, A., & Maydeu-Olivares, A. (2011). Item response modeling of forced-choice questionnaires. Educational and
Psychological Measurement, 71(3), 460-502.
Brown, A., & Maydeu-Olivares, A. (2012). Fitting a Thurstonian IRT model to forced-choice data using Mplus. Behavior
Research Methods, 44(4), 1135-1147.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved.72
Bunker, K. A., Kram, K. E., & Ting, S. (2002). The young and the clueless. Harvard Business Review, 80(12), 80-87.
Campbell, J. P. (2012). Behavior, performance, and effectiveness in the twenty-first century. In S. W. J. Kozlowski (Ed.),
The Oxford handbook of organizational psychology (Vol. 1, pp. 159-194). Oxford, England: Oxford University Press.
Campbell, J. P., & Pritchard, R. D. (1976). Motivation theory in industrial and organizational psychology. In M. D.
Dunnette (Ed.), Handbook of industrial and organizational psychology (pp. 63-130). Chicago, IL: Rand McNally.
Cardon, M. S., Wincent, J., Singh, J., & Drnovsek, M. (2009). The nature and experience of entrepreneurial passion. The
Academy of Management Review, 34(3), 511-532. doi: 10.5465/AMR.2009.40633190
Cardon, M. S., Zietsma, C., Saparito, P., Matherne, B., & Davis, C. (2005). A tale of passion: New insights into
entrepreneurship from a parenthood metaphor. Journal of Business Venturing, 20, 23-45.
CareerBuilder. (2014). Majority of workers don’t aspire to leadership roles.
Retrieved from http://www.careerbuilder.com/share/aboutus/pressreleasesdetail.
aspx?sd=9%2f10%2f2014&id=pr841&ed=12%2f31%2f2014
Carroll, J. B. (1993). Human cognitive abilities: A survey of factor-analytical studies. New York, NY: Cambridge
University Press.
Carver, C. S., & Scheier, M. F. (2014). Dispositional optimism. Trends in Cognitive Sciences, 18(6), 293-299.
Cesare, S. J., & Thornton, C. (1993). Human resource management and the specialist/generalist issue. Journal of
Managerial Psychology, 8(3), 31-40.
Chajut, E., & Algom, D. (2003). Selective attention improves under stress: Implications for theories of social cognition.
Journal of Personality and Social Psychology, 85(2), 231-248. doi: 10.1037/0022-3514.85.2.231
Charan, R., Drotter, S., & Noel, J. (2011). The leadership pipeline: How to build the leadership powered company. San
Francisco, CA: Jossey-Bass.
Cheung, M. W. L., & Chan, W. (2002). Reducing uniform response bias with ipsative measurement in multiple-group
confirmatory factor analysis. Structural Equation Modeling, 9(1), 55-77. doi: 10.1207/S15328007SEM0901_4
Christian, M. S., Garza, A. S., & Slaughter, J. E. (2011). Work engagement: A quantitative review and test of its relations
with task and contextual performance. Personnel Psychology, 64(1), 89-136.
Church, A. H., & Rotolo, C. T. (2013). How are top companies assessing their high potentials and senior executives? A
talent management benchmark study. Consulting Psychology Journal: Practice and Research, 65(3), 199-223.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum
Associates.
Collins, C. J., Hanges, P. J., & Locke, E. A. (2004). The relationship of achievement motivation to entrepreneurial
behavior: A meta-analysis. Human Performance, 17(1), 95-117.
Cone, J. (2013). Tolerating ambiguity in four simple steps. Interaction Associates. Retrieved from http://
interactionassociates.com/insights/blog/tolerating-ambiguity-four-simple-steps#.Vj_I2ZV0zug
Corporate Executive Board. (2001). “Global personality inventory.” Arlington, VA.
Corporate Leadership Council. (2005). Realizing the full potential of rising talent. Washington, DC: Corporate Executive
Board.
Costa, P. T. & McCrae, R. R. (1992). Revised NEO Personality Inventory Manual. Odessa, FL: Psychological Assessment
Resources Inc.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved. 73
Crandell, S., Hazucha, J., & Orr, J. E. (2014). Precision talent intelligence: The definitive four dimensions of leadership
and talent (Report). Los Angeles, CA: Korn Ferry Institute.
Crandell, S., Orr, J. E., & Urs, L. (2015). Formative experiences may be key for CEO readiness. Los Angeles, CA: Korn
Ferry Institute.
Crocker, L., & Algina, J. (1986). Introduction to classical and modern test theory. New York, NY: CBS College Publishing.
Cronbach L. (1951). Coefficient alpha and the internal structure of tests. Psychometrika, 16(3), 297-334.
Dai, G., De Meuse, K. P., & Tang, K. Y. (2013). The role of learning agility in executive career success: The results of two
field studies. Journal of Managerial Issues, 25(2), 108–131.
Dai, G., Tang, K. Y., & Feil, J. (2014, July). Fast-rising talent: Highly learning agile people get promoted at double speed
(Report). Los Angeles, CA: Korn Ferry Institute.
Dane, E. (2011). Paying attention to mindfulness and its effects on task performance in the workplace. Journal of
Management. 37(4), 997-1018.
Deluga, R. J. (1998). American presidential proactivity, charismatic leadership, and rated performance. The Leadership
Quarterly, 9(3), 265-291.
D’Mello, S. (2015). Keys to success. Los Angeles, CA: Korn Ferry Institute.
De Meuse, K. P., Dai, G., & Hallenbeck, G. S. (2010). Learning agility: A construct whose time has come. Consulting
Psychology Journal: Practice and Research 62(2), 119-30.
De Pater, I. E., Van Vianen, A. E. M., Bechtoldt, M. N., & Klehe, U. (2009). Employees’ challenging job experiences and
supervisors’ evaluations of promotability. Personnel Psychology, 62(2), 297-325.
Depue, R. A., & Collins, P. F. (1999). Neurobiology of the structure of personality: Dopamine, facilitation of incentive
motivation, and extraversion. Behavioral and Brain Sciences, 22, 491-569.
DeRue, D. S., & Wellman, N. (2009). Developing leaders via experience: The role of developmental challenge, learning
orientation, and feedback availability. Journal of Applied Psychology, 94(4), 859.
DeYoung, C. G. (2010). Mapping personality traits onto brain systems: BIS, BAS, FFFS, and beyond. European Journal
of Personality, 24, 404-407.
DeYoung, C. G., Hirsh, J. B., Shane, M. S., Papademetris, X., Rajeevan, N., & Gray, J. R. (2010). Testing predictions from
personality neuroscience: Brain structure and the Big Five. Psychological Science, 21, 820-828.
Dowell, B. E. (2010). Managing leadership talent pools. In R. F. Silzer & B. E. Dowell (Eds.), Strategy-driven talent
management: A leadership imperative. San Francisco, CA: Jossey Bass.
Dragoni, L., Tesluk, P. E., Russell, J. E. A., & Oh, I. S. (2009). Understanding managerial development: Integrating
developmental assignments, learning orientation, and access to developmental opportunities in predicting
managerial competencies. Academy of Management Journal, 52, 731–743.
Dragoni, L., Oh, I., VanKatwyk, P., & Tesluk, P. E. (2011). Developing executive leaders: The relative contribution
of cognitive ability, personality, and the accumulation of work experience in predicting strategic thinking
competency. Personnel Psychology, 64(4), 829-864.
Dragoni, L., Tesluk, P. E., & Oh, I. (2009). Understanding managerial development: Integrating developmental
assignments, learning orientation, and access to developmental opportunities in predicting managerial
competencies. Academy of Management Journal, 52(4), 731-743.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved.74
Drasgow, F., Chernyshenko, O. S., & Stark, S. (2010). 75 Years after Likert: Thurstone was right! Industrial and
Organizational Psychology: Perspectives on Science and Practice, 3(4), 465-476.
Drescher, M., Korsgaard, M. A., Welpe, I. M., Picot, A., & Wigand, R. T. (2014). The dynamics of shared leadership:
Building trust and enhancing performance. Journal of Applied Psychology, 99(5), 771-783.
Dries, N., & Pepermans, R. (2008). “Real” high-potential careers: An empirical study into the perspectives of
organizations and high potentials. Personnel Review, 37(1), 85-108.
Dries, N., & Pepermans, R. (2012). How to identify leadership potential: Development and testing of a consensus model.
Human Resource Management, 51(3), 361-385. doi: 10.1002/hrm.21473
Dries, N., Vantilborgh, T., & Pepermans, R. (2012). The role of learning agility and career variety in the identification and
development of high potential employees. Personnel Review, 41(3), 340-359.
Driskell, J. E., & Salas, E. (1992). Collective behaviour and team performance, Human Factors, 34, 277-288.
Duckworth, A. L., Peterson, C., Matthews, M. D., & Kelly, D. R. (2007). Grit: Perseverance and passion for long-term
goals. Journal of Personality and Social Psychology, 92(6), 10871101. doi: 10.1037/0022-3514.92.6.1087
Ericsson, K. A., & Charness, N. (1994). Expert performance: Its structure and acquisition. American Psychologist, 49(8),
725-747. doi: 10.1037/0003-066X.49.8.725
Eskreis-Winkler, L., Shulman, E. P., Beal, S. A., & Duckworth, A. L. (2014). The grit effect: Predicting retention in the
military, the workplace, school and marriage. Frontiers in Psychology, 5(36). doi: 10.3389/fpsyg.2014.00036
Feldman, G., Hayes, A., Kumar, S., Greeson, J., Laurenceau, J. P. (2007). Mindfulness and emotion regulation: The
development and initial validation of the Cognitive and Affective Mindfulness Scale-Revised (CAMS-R). Journal of
Psychopathology and Behavioral Assessment, 29(3), 177-190.
Galasso, A., & Simcoe, T. S. (2011). CEO overconfidence and innovation. Management Science, 57(8).
doi: 10.1287/mnsc.1110.1374
Gerstner, C., Hazucha, J., & Davies, S. (2012, April). Motivators: What is important to leaders at different levels. In S. E.
Davies (Chair), Understanding and supporting transitions up the leadership ladder. Symposium presented at the
annual conference of the Society for Industrial and Organizational Psychology, San Diego, CA.
Goffin, R. D., & Christiansen, N. D. (2003). Correcting personality tests for faking: A review of popular personality tests
and an initial survey of researchers. International Journal of Selection and Assessment, 11(4), 340-344.
Goleman, D. (1998). Working with emotional intelligence. New York, NY: Bantam Books.
Gouldner, A. W. (1957). Cosmopolitans and locals: Toward an analysis of latent social roles. Administrative Science
Quarterly, 2, 281-306.
Gratton, L. (2010). The future of work. Business Strategy Review. 21(3), 16-23.
Gratton, L., & Erickson, T. (2007). Eight ways to build collaborative teams. Harvard Business Review. Retrieved from
https://hbr.org/2007/11/eight-ways-to-build-collaborative-teams
Guterl, F. (1984). Spectrum/Harris poll: The Career. IEEE Spectrum, 21(6), 59-63.
Haigh, E. A. P., Moore, M. T., Kashdan, T. B., & Fresco, D. M. (2011). Examination of the factor structure and concurrent
validity of the Langer Mindfulness/Mindlessness scale. Assessment, 18(1), 11-26. doi: 10.1177/1073191110386342
Hammond, S., & Barrett, P. (1996). The psychometric and practical implications of the use of ipsative, forced-choice
format, Questionnaires. Proceedings of the British Psychological Society Occupational Psychology Conference,
January, 135-144. Leicester: BPS Press.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved. 75
Hayes, S., Bond, F., & Barnes-Holmes, D. (2006). Acceptance and mindfulness at work: Applying acceptance and
commitment therapy and relational frame theory to organizational behavior management. Binghamton, NY:
Haworth Press.
Heggestad, E. D., Morrison, M., Reeve, C. L., & McCloy, R. A. (2006). Forced-choice assessments of personality for
selection: Evaluating issues of normative assessment and faking resistance. Journal of Applied Psychology, 91(1),
9-24. doi: 10.1037/0021-9010.91.1.9
Hogan, R. (1983). A socioanalytic theory of personality. In M. M. Page (Ed.), 1982 Nebraska symposium on motivation
(pp. 55-89). Lincoln, NE: University of Nebraska Press.
Hogan, R., & Hogan, J. (2001). Assessing leadership: A view from the dark side. International Journal of Selection and
Assessment, 9(1/2), 40-51.
Hogan, R., & Hogan, J. (2009). Hogan developmental survey overview guide. Tulsa, OK: Hogan Assessment Systems.
Hogan, R., & Warrenfeltz, W. (2003). Educating the modern manager. Academy of Management Learning and
Education, 2, 74-84.
Horn, J. L. (1980). Concepts of intellect in relation to learning and adult development. Intelligence, 4, 285-317.
Hough, L. M., & Ones, D. S. (2001). The structure, measurement, validity, and use of personality variables in industrial,
work, and organizational psychology. In N. Anderson, D. S. Ones, H. K. Sinangil, & C. Viswesvaran (Eds.), Handbook
of industrial, work, and organizational psychology: Volume 1 Personnel psychology (pp. 233-277). London,
England: Sage Publications.
Hough, L. M., Oswald, F. L., & Ployhart, R. E. (2001). Determinants, detection and amelioration of adverse impact in
personnel selection procedures: Issues, evidence and lessons learned. International Journal of Selection and
Assessment, 9(1-2), 152-194.
Houlfort, N., Philippe, F.L., Vallerand, R.J., & Ménard, J. (2014). On passion and heavy work investment: Personal and
organizational outcomes. Journal of Managerial Psychology, 29(1), 25-45.
House, R. J., Spangler, W. D., & Woycke, J. (1991). Personality and charisma in the U.S. presidency: A psychological
theory of leader effectiveness. Administrative Science Quarterly, 36(3), 364-396.
Howard, A. (2009, April). Global leader SOS: Can multinational leadership skills be developed? Symposium presented
at the annual conference of the Society for Industrial and Organizational Psychology, New Orleans, LA.
Howe, M. J. A. (1999). Genius explained. New York, NY: Cambridge University Press.
Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus
new alternatives. Structural Equation Modeling, 6, 1-55.
Hughes, R. L., Ginnett, R. C., & Curphy, G. J. (2008). Leadership: Enhancing the lessons of experience (6th ed.). New
York, NY: McGraw-Hill.
John, O. P., Naumann, L. P., & Soto, C. J. (2008). Paradigm shift to the integrative Big-Five trait taxonomy: History,
measurement, and conceptual issues. In O. P. John, R. W. Robins, & L. A. Pervin (Eds.), Handbook of personality:
Theory and research (pp. 114-158). New York, NY: Guilford Press.
Johnson, N. (2007). Self report measures of mindfulness: A review of the literature. (Master’s thesis, Pacific University).
Retrieved from http://commons.pacificu.edu/cgi/viewcontent.cgi?article=1145&context=spp
Judge, T. A., Bono, J. E., Ilies, R., & Gerhardt, M. W. (2002). Personality and leadership: A qualitative and quantitative
review. Journal of Applied Psychology, 87(4), 765-780.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved.76
Judge, T. A., Colbert, A. E., & Ilies R. (2004). Intelligence and leadership: A quantitative review and test of theoretical
propositions. Journal of Applied Psychology, 89(3),542-552. doi: 10.1037/0021-9010.89.3.542
Kam, C., & Meyer, J. P. (2012). Do optimism and pessimism have different relationships with personality dimensions? A
re-examination. Personality and Individual Differences, 52, 123-127.
Kanfer, R. (1990). Motivation theory and industrial and organizational psychology. In M. D. Dunnette & L. M. Hough
(Eds.), Handbook of industrial and organizational psychology (pp. 75-170). Palo Alto, CA: Consulting Psychologists
Press, Inc.
Karaevli, A., & Hall, D. T. (2003). Growing leaders for turbulent times: Is succession planning up to the challenge?
Organizational Dynamics, 32, 62-79.
King, J. E., & Figueredo, A. J. (1997). The five-factor model plus dominance in chimpanzee personality. Journal of
Research in Personality, 31(2), 257-271.
Kirschkamp, A. (2007). A contingency-based view of chief executive officers’ early warning behavior: An empirical
analysis of German medium-sized companies. Deutscher Universitäts-Verlag | Gabler (GWV) Fachverlage GmbH,
Wiesbaden.
Kooij, D. T., De Lange, A. H., Jansen, P. G. W., Kanfer, R., & Dikkers, J. S. (2011). Age and work-related motives: Results of
a meta-analysis. Journal of Organizational Behavior, 32(2), 197-225.
Kornhauser, W. (1962). Scientists in industry: Conflict and accommodation. Berkeley, CA: University of California Press.
Kumar, S., Feldman, G., & Hayes, A. (2008). Changes in mindfulness and emotion regulation in an exposure-based
cognitive therapy for depression. Cognitive Therapy and Research, 32(6), 734-744. doi: 10.1007/s10608-008-9190-1
Langer, E. J. (2009). Counterclockwise: Mindful health and the power of possibility. New York, NY: Ballantine Books.
Lee, R. A. (2012). Accelerating the development and mitigating derailment of high potential through mindfulness
training. The Industrial-Organizational Psychologist, 49(3), 23-34.
Lewis, J. (2012). Model-based best-in-class profiling with Decision Styles: An overview and technical report. [Internal
technical document available by request.] Los Angeles, CA: Korn Ferry International.
Lewis, J. L., & Ream, R. K. (2012). The criterion-related validity of the Decision Styles assessment. Los Angeles, CA:
Korn Ferry International.
Liu, D., Chen, X., & Yao, X. (2011). From autonomy to creativity: A multilevel investigation of the mediating role of
harmonious passion. Journal of Applied Psychology, 96(2), 294-309. doi: http://dx.doi.org/10.1037/a0021294
Livingstone, H. A., & Day, A. L. (2005). Comparing the construct and criterion-related validity of ability-based and
mixed-model measures of emotional intelligence. Educational and Psychological Measurement, 65(5), 757-779.
Locke, E. A. (1978). The ubiquity of the techniques of goal setting in the theories and approaches to employee
motivation. Academy of Management Review, 3, 594-601.
Lombardo, M. M., & Eichinger, R. W. (1989). Preventing derailment: What to do before it’s too late. Greensboro, NC:
Center for Creative Leadership.
Lombardo, M. M., & Eichinger, R. W. (2000). High potentials as high learners. Human Resource Management, 39(4),
321-329.
Lombardo, M. M., & McCauley, C. D. (1988). The dynamics of management derailment. Greensboro, NC: Center for
Creative Leadership.
Mann, R. D. (1959). A review of the relationships between personality and performance in small groups. Psychological
Bulletin, 56(4), 241-270. doi: 10.1037/h0044587
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved. 77
Martin, J., & Schmidt, C. (2010). How to keep your top talent. Harvard Business Review, 88(5), 54-61.
Maslow, A. H. (1954). Motivation and personality. New York, NY: Harper & Row.
Maydeu-Olivares, A., Hernández, A., & McDonald, R. P. (2006). A multidimensional ideal point item response theory
model for binary data. Multivariate Behavioral Research, 41(4), 445-472.
McCall, M. W., Jr., Lombardo, M. M., & Morrison, A. M. (1988). The lessons of experience: How successful executives
develop on the job. Lexington, MA: Lexington Books.
McCauley, C. D., Ruderman, M. N., Ohlott, P. J., & Morrow, J. E. (1994). Assessing the developmental components of
managerial jobs. Journal of Applied Psychology, 79(4), 544.
McClelland, D. C. (1965). Toward a theory of motive acquisition. American Psychologist, 20, 321-333.
McClelland, D. C., & Boyatzis, R. (1982). Leadership motive pattern and long-term success in management. Journal of
Applied Psychology, 67(6), 737-743.
McClelland, D.C., & Burnham, D. H., (1976). Power is the great motivator. Harvard Business Review, 54, 100-110.
McCrae, R. R., & Costa, P. T. (1983). Social desirability scales: More substance than style. Journal of Consulting and
Clinical Psychology, 51(6), 882-888. doi: 10.1037/0022-006X.51.6.882
McCrae, R. R., & Costa, P. T. (1987). Validation of the 5-factor model of personality across instruments and observers.
Journal of Personality and Social Psychology, 52, 81-90.
Meade, A. W. (2004). Psychometric problems and issues involved with creating and using ipsative measures for
selection. Journal of Occupational and Organizational Psychology, 77, 531-552.
Moutafi, J., Furnham, A., & Paltiel, L. (2005). Can personality factors predict intelligence? Personality and Individual
Differences, 38(5), 1021-1033. doi: 10.1016/j.paid.2004.06.023
Muthén, L. K., & Muthén, B. O. (2010). Mplus user’s guide (6th ed.). Los Angeles, CA: Muthén & Muthén.
Nadkarni, S., & Barr, P. S. (2008). Environmental context, managerial cognition, and strategic action: An integrated
view. Strategic Management Journal, 29(13), 1395-1427. doi: http://dx.doi.org/10.1002/smj.717
NCS Pearson. (2007). Raven’s Advanced Progressive Matrices (APM): Evidence of reliability and validity. Bloomington,
MN: NCS Pearson.
Northouse, P. G. (1997). Leadership: Theory and practice. Thousand Oaks, CA: Sage Publications, Inc.
Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric theory (3rd ed.). New York, NY: McGraw-Hill.
Ones, D. S., Dilchert, S., Viswesvaran, C., & Judge, T. A. (2007). In support of personality assessment in organizational
settings. Personnel Psychology, 60(4), 995-1027. doi: 10.1111/j.1744-6570.2007.00099.x
Ones, D. S., Dilchert, S., Viswesvaran, C., & Salgado, J. F. (2010). Cognitive abilities. In J. L. Farr & N. T. Tippins (Eds.),
Handbook of employee selection (pp. 255-275). New York, NY: Routledge/Taylor & Francis.
Oreg, S., Vakola, M., & Armenakis, A., (2011). Change recipients’ reactions to organizational change: A 60-year review of
quantitative studies. Journal of Applied Behavioral Science, 47(4), 461–524.
Osburn, H. G. (2000). Coefficient alpha and related internal consistency reliability coefficients, Psychological Methods,
5(3), 343-355.
Passmore , J. (2007). An integrative model for executive coaching. Consulting Psychology Journal: Practice and
Research, 59(1), 68-78.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved.78
Passmore, J., & Marianetti, O. (2007, Dec). The role of mindfulness in coaching. The Coaching Psychologist, 3(3),
130-136.
Pepermans, R., Vloeberghs, D., & Perkisas, B. (2003). High potential identification policies: An empirical study among
Belgian companies. Journal of Management Development, 22(8), 660-678.
Peterson, D. B., & Erdahl, P. (2007, April). Early identification and development of senior leadership talent: The
secret insider’s guide. Workshop presented at the 22nd annual conference of the Society for Industrial and
Organizational Psychology, New York, NY.
Philippe, F. L., Vallerand, R. J., Houlfort, N., Lavigne, G. L., & Donahue, E. G. (2010). Passion for an activity and quality of
interpersonal relationships: The mediating role of emotions. Journal of Personality and Social Psychology, 98(6),
917-932. doi: 10.1037/a0018017
Pittman, T. S., & Zeigler, K. R. (2007). Basic human needs. In A. W. Kruglanski & E. T. Higgins (Eds.), Social psychology:
Handbook of basic principles (2nd ed., pp. 473-489). New York, NY: Guilford Press.
Quast, L. N., Wohkittel, J. M., Chung, C., Vue, B., Center, B. A., & Phillips, A. E. (2013, February). Behaviors associated
with managerial career derailment: An exploration of self-other agreement pattern groups in multisource feedback.
Paper presented at the 2013 AHRD Americas International Conference, Washington, DC.
Quińones, M. A., Ford, J. K., & Teachout, M. S. (1995). The relationship between work experience and job performance:
A conceptual and meta-analytic review. Personnel Psychology, 48(4), 887-910.
Rokeach, M. (1973). The nature of human values. New York, NY: Free Press.
Sackett, P., (2011). Faking in personality assessments: Where do we stand? In M. Ziegler, C. MacCann, & R. Roberts
(Eds.), New perspectives on faking in personality assessments, (pp. 330-344). Oxford University Press.
Sackett, P. R., & Lievens, F. (2008). Personnel selection. Annual Review of Psychology, 59, 419-450.
Sala, F. (2003). Executive blind spots: Discrepancies between self- and other-ratings. Consulting Psychology Journal:
Practice and Research, 55(4), 222-229.
Schein, E. H. (1987). Individuals and careers. In: J. W. Lorsch (Ed.), Handbook of organizational behavior (pp. 155-171).
Englewood Cliffs, NJ: Prentice-Hall.
Schmitt, D. P., Allik, J., McCrae, R. R., & Benet-Martínez, V. (2007). The geographic distribution of Big Five personality
traits: Patterns and profiles of human self-description across 56 nations. Journal of Cross-Cultural Psychology,
38(2), 173-212.
Schultz, W. (1999). The reward signal of midbrain dopamine neurons. News in Physiological Sciences, 14(6), 249255.
Segerstrom, S. C. (2007). Optimism and resources: Effects on each other and on health over 10 years. Journal of
Research in Personality, 41(4), 772-786.
Sharpe, J. P., Martin, N. R., & Roth, K. A. (2011). Optimism and the Big Five factors of personality: Beyond Neuroticism
and Extraversion. Personality and Individual Differences, 51, 946-951.
Sherrill, W. W. (2001). Tolerance of ambiguity among MD/MBA students: Implications for management potential.
Journal of Continuing Education in the Health Professions, 21, 117-122.
Silzer, R., & Church, A. H. (2009). The pearls and perils of identifying potential. Industrial and Organizational
Psychology, 2(4), 377-412.
Silzer, R., & Church, A. H. (2010). Identifying and assessing high potential talent: Current organizational practices. In
R. F. Silzer & B. E. Dowell (Eds.), Strategy-driven talent management: A leadership imperative (pp. 213-280). San
Francisco, CA: Jossey-Bass.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved. 79
Simonton, D. K. (1994). Greatness: Who makes history and why. New York, NY: Guilford Press.
Slan-Jerusalim, R., & Hausdorf, P. A. (2007). Managers’ justice perceptions of high potential identification practices.
Journal of Management Development, 26(10), 933-950.
Sliter, K. A., & Christiansen, N. D. (2012). Effects of targeted self-coaching on applicant distortion of personality
measures. Journal of Personnel Psychology, 11(4), 169-175.
Spangler, W. D. (1992). Validity of questionnaire and TAT measures of need for achievement: Two meta-analyses.
Psychological Bulletin, 112(1), 140-154.
Spearman, C. (1904) The proof and measurement of association between two things. The American Journal of
Psychology, 15(1), 72-101.
Stark, S., & Chernyshenko. O. S. (2007). Adaptive testing with the multi-unidimensional pairwise preference model. In
D. J. Weiss (Ed.), Proceedings of the 2007 GMAC Conference on Computerized Adaptive Testing. Retrieved from
www.psych.umn.edu/psylabs/CATCentral
Stark, S., Chernyshenko, O. S., Chan, K. Y., Lee, W. C., & Drasgow, F. (2001). Effects of the testing situation on item
responding: Cause for concern. Journal of Applied Psychology, 86(5), 943.
Stark, S., Chernyshenko, O. S., & Drasgow, F. (2005). An IRT approach to constructing and scoring pairwise preference
items involving stimuli on different dimensions: The multi-unidimensional pairwise-preference model. Applied
Psychological Measurement, 29(3), 184-203.
Stogdill, R. M. (1948). Personal factors associated with leadership: A survey of the literature. Journal of Psychology,
(25), 35-71.
Strosaker, G. (2010). Developing a tolerance for ambiguity. Constant Cogitation. Retrieved from
http://gregstrosaker.com/2010/01/developing-a-tolerance-for-ambiguity/
Sull, D. (2009, December 23). Competing through organizational agility. Forbes. Retrieved from
http://www.forbes.com/2009/12/23/strategy-innovation-agility-leadership-managing-mckinsey.html
Takane, Y., & De Leeuw, J. (1987). On the relationship between item response theory and factor analysis of discretized
variables. Psychometrika, 52(3), 393-408.
Tang, K. Y., & Dai, G. (2013). The Leadership Architect® 2013 global norms report II: Career stallers and stoppers norms
and analysis. Los Angeles, CA: Korn Ferry.
Tang, K. Y., Dai, G., & De Meuse, K. P. (2013). Assessing leadership derailment factors in 360° feedback: Differences
across position levels and self-other agreement. Leadership & Organization Development Journal, 34(4), 326-343.
Tannenbaum, S. I. (1997). Enhancing continuous learning: Diagnostic findings from multiple companies. Human
Resource Management, 36(4), 437-452.
Terman, L. M., & Oden, M. H. (1947). The gifted child grows up: Twenty-five years’ follow-up of a superior group (Vol. 4).
Stanford University Press.
Tesluk, P. E., & Jacobs, R. R. (1998). Toward an integrated model of work experience. Personnel Psychology, 51, 321-355.
Thompson, W. D. (1990). Kappa and attenuation of the odds ratio. Epidemiology, 1, 357-369.
Thoresen, C. J., Bradley, J. C., Bliese, P. D., & Thoresen, J. D. (2004). The Big Five personality traits and individual job
performance growth trajectories in maintenance and transitional job stages. Journal of Applied Psychology, 89(5),
835-853.
Thurstone, L. L. (1927). A law of comparative judgment. Psychological Review, 79, 281-299.
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
© Korn Ferry 2015–2016. All rights reserved.80
Venus, M., Mao, C., Lanaj, K., & Johnson, R. E. (2012). Collectivistic leadership requires a collective identity. Industrial and
Organizational Psychology: Perspectives on Science and Practice, 5(4), 432-436.
Weick, K. E., & Roberts, K. H. (1993). Collective mind in organizations: Heedful interrelating on flight decks.
Administrative Science Quarterly, 38(3), 357-381.
Wells, S. J. (2003). Who’s next: Creating a formal program for developing new leaders can pay huge dividends, but
many firms aren’t reaping those rewards. HR Magazine, 48(11), 44-64.
Wiggins, J. S. (Ed.). (1996). The five-factor model of personality: Theoretical perspectives. New York, NY: Guilford.
Winter, D. G. (1987). Leader appeal, leader performance, and the motive profiles of leaders and followers: A study of
American presidents and elections. Journal of Personality and Social Psychology, 52(1), 196-202.
Worley, C. G., & Lawler, E. E., III. (2010). Agility and organization design: A diagnostic framework. Organizational
Dynamics, 39(2), 194-204.
Yammarino, F. J., Salas, E., Serban, A., Shirreffs, K., & Shuffler, M. L. (2012). Collectivistic leadership approaches: Putting
the “we” in leadership science and practice. Industrial and Organizational Psychology: Perspective on Science and
Practice, 5(4), 382-402.
Yukl, G., & Mashud, R. (2010). Why flexible and adaptive leadership is essential. Consulting Psychology Journal: Practice
and Research, 62(2), 81-93.
Zaccaro, S. J. (2001). The nature of executive leadership: A conceptual and empirical analysis of success. Washington,
DC: American Psychological Association.
Zes, D. (2016). KFmlfcirt: FC-IRT Multi-loadings. R package version 0.6.7.
Zes, D., & Landis, D. (2013). A better return on self-awareness (Report). Los Angeles, CA: Korn Ferry Institute.
Zes, D., Lewis, J., & Landis, D. (2014). kcirt: k-Cube Thurstonian IRT Models [Computer Software]. URL https://cran.r-
project.org/web/packages/kcirt/index.html
Korn Ferry Assessment of Leadership Potential • Research guide and technical manual
About Korn FerryKorn Ferry is the preeminent global people and organizational advisory firm. We help leaders, organizations, and societies succeed by releasing the full power and potential of people. Our nearly 7,000 colleagues deliver services through our Executive Search, Hay Group and Futurestep divisions. Visit kornferry.com for more information.
Visit www.kornferry.com for more information on Korn Ferry, and www.kornferryinstitute.com for thought leadership, intellectual property and research.