Main TextbooksJones S., Wahba K., van der Heijden B., How to Write Your MBA Thesis, Maastricht School of Management Series in Intercultural and Global Management, Meyer & Meyer Media, 2007.
Sekaran, Uma, Research Methods For Business, 4th Edition, John Wiley & Sons Inc, 2003.
Other ReferencesJohn W. Creswell, Research Design Qualitative & Quantitative Approaches,McGraw-Hill International Editions, 1994.
Robert K. Yin, Case Study Research Design and Method, 2nd Edition, SAGE Publications Thousands Oaks, California, 1994.
John Gill and Phil Johnson, Research Methods for Managers, 2nd Edition, Paul Chapman Publishing, a SAGE Publications Company, 1997.
Naresh K-Malhotra, Marketing Research, 3rd edition, Prentice Hall 1999.
Babbie, Earl (1997), Survey Research Methods, wordsworth Publishing Company.
The manager of a multinational corporation is curious to know if the work ethic values of employees working in its subsidiary in India are different from those in its subsidiary in America.
Since very little is known about what work ethics values mean to people in other cultures, the manager had to conduct Exploratory Research, to satisfy his curiosity, via interviewingemployees in the Indian subsidiary.
Religion, political, economic, and social conditions were revealed to play a major role in how people view their work in different parts of the world.
It includes, not only describing the situation, but also the analysis
and the explanation (Why, How).
It aims to understand phenomena by discovering and measuring
causal relations among the different factors (i.e. variables).
How can we reduce the number of complaints made by customers? How can we improve the delivery times of our products?How can we expand the range of our services?
It relies on “hypothesis testing” to explain the variance in the
Primary DataPrimary Data Secondary DataSecondary Data
Collection purposeCollection purpose For the problem at handFor the problem at hand For other problemsFor other problemsCollection processCollection process Very involvedVery involved Rapid & easyRapid & easyCollection costCollection cost HighHigh Relatively lowRelatively lowCollection timeCollection time LongLong ShortShort
A Comparison of Primary & Secondary DataA Comparison of Primary & Secondary Data
Literature Review is the documentation of a comprehensive review of the published and unpublished work, from the secondary data sources, in the area of specific interest to the researcher.
It is one way of summarizing secondary data.
A start point is the Library: books, journals, magazines, conference proceedings, doctoral and master’s dissertations, government publications and reports.
The problem statement can be made more precise and clear.
LR insures that no important factor (variable) that has been found to have an impact on the problem is ignored.
It may happen that researcher spends considerable time and effort to “discover” something that has already been thoroughly researched. LR would prevent such a waste of resources, i.e. the risk of “Reinventing the Wheel”.
LR gives a good basis to proceed further with the research process, and provides the foundation for developing a comprehensive theoretical framework (What variables would be most important).
LR also facilitates the creative integration of the information gathered (via interviews, survey, etc.) with what is found in previous studies.
It involves identifying the various published and unpublished materials that are available on the topic of interest and gaining access to them.Today, almost every library has an online computer system to locate published information on various topics (by topic, author, keyword, etc.)Databases are also available for statistical information (Marketing, financial, etc.)Published articles in journals, newspapers, periodicals, and conference proceedings are now available on databases. 3 types of databases exists, these are:
The Bibliographic DatabasesThe Abstract DatabasesThe Full-text Databases
The abstracts provided by the databases, and library online systems that seem to be relevant to the topic in hand are then obtained as a full text.
After the collection of all the relevant materials (book chapters, articles, web pages, statistical info, etc.) researcher starts the reading and extracting process.
While reading, it is possible that certain other factors (variables) are found to be closely related to the problem in hand. Example: while researching the effectiveness of employee motivation, the researcher may find that the size of the company is also an important factor, thus wanting to read more about organization size categories.
Resources considered relevant are then listed as references using the appropriate referencing format.
The documentation of the relevant studies, citing the author, and the year of the study is called literature review or survey.It is a clear and logical presentation of the relevant research work done so far in the area of investigation.LR should bring together all relevant studies/information in a logical manner instead of presenting them in a chronological order.A Good LR should leads the reader logically to the problem statement. A suggested patterns is as follows:
Introduction/Background/Overview (History of the business /Industry/Sector)Overview about the ABC (Theoretical Part) Worldwide Cases (Best Practices) Local Case (Egypt)
Step 3: Problem DefinitionAfter the unstructured interviews and the literature review, theresearcher should narrow down the problem from its broad base and define the issues of concern more clearly.A “Problem” does not necessarily means that something is seriously going wrong with the current situation, and needs to b rectifiedimmediately. A “Problem” could simply indicate an interest in an issue where finding the right answers might help improving an existing situation.Thus, researcher should define the “Problem Statement” as any situation where a gap exists between the actual and the desired ideal states.Symptoms of the problem should not be defined as the real problem.
Example: a manager might have tried to increase productivity by increasing the piece rate, but with little success. Here, thelow productivity may be a symptom of the real problem which is the low morale and motivation of the employees.
A Theoretical Framework is a conceptual model of how one theorizes or makes logical sense of the relationships among the several factors (variables) that have been identified as important to the problem.
This theory flows logically from the LR (documentation of previous related researches).
TF is no more than identifying the network of relationships among the variables considered important to any given problem situation.
From TF, research questions and hypotheses can be developed and tested via primary data to examine whether the theory formulated is valid or not.
Thus, the entire research rests on the basis of the TF.
A variable is anything that can take on differing or varying values.
Production Units: One worker in a manufacturing department may produce one packet per minute, a second might produce two per minute, and a third might produce five per minute.
Motivation: The levels of motivation of employees to learn in a work team might take on varying values ranging from “very low” to “very high”.
A manager is concerned that the sales of a newly introduced product is not meeting his expectations.
Since sales of the product can vary (low, medium or high) and since sales is the main focus of interest to the manager, it is the dependent variable.
A vice president is concerned that the employees are not loyal to the organization. The DV in this case would be “organizational loyalty”.
The VP may want to know what accounts for the variance in the loyalty of employees, with the purpose of controlling it. If he finds that pay level affects loyalty, he can offer then pay raises to keep them in the organization.
It is the variable that influences the dependent variable in either a positive or a negative way (i.e. with each unit of increase in the independent variable, there is an increase or decrease in the dependent variable.Thus, the variance in the dependent variable is accounted for by the independent variable.This relationship is called: “Cause & Effect”.
By theory, the diversity of the workforce (comprising people of different ethnic origins, races and nationalities) contributes more to organizational effectiveness because each group brings its own special expertise and skills to the workplace.However, the effective utilization of different talents, perspectives, and capabilities of the various groups for enhanced organizational effectiveness is contingent on the skill of the managers to act as catalysts.
Distinction between Independent& Moderating Variable
Sometimes confusion occur between IV and MV.The decision whether to label them dependent, independent, or moderating depends on how they affect one another.
An independent variable of one research, might become a moderating variable in another different research.
It is the one that surfaces between the time independent variable start operating to influence dependent variable and the time its impact is felt on it.The creative synergy results from the diverse workforce interacting and bringing together their multifaceted expertise in problem solving.Creative synergy will not result from the diverse workforce unless the manager iscapable of harnessing that synergy by coordinating the different skills.If manager lacks the expertise to perform this role, no matter the workforce have different skills, synergy will not surface, and organization will not function effectively.
PROBLEMS AT ROADWAY HOSPITALITYPROBLEMS AT ROADWAY HOSPITALITYJohn Campbell, the CEO of Roadway Hospitality, was wondering howJohn Campbell, the CEO of Roadway Hospitality, was wondering how to to
differentiate among the three different types of facilities offedifferentiate among the three different types of facilities offered under the red under the Roadway Hospitality flagship so as to attract the right type of Roadway Hospitality flagship so as to attract the right type of clients to each of clients to each of the facilities. The the facilities. The Roadway Deluxe Roadway Deluxe was meant for business travelers, the was meant for business travelers, the Roadway Roadway Express Express was meant for those looking for the least expensive accommodatiowas meant for those looking for the least expensive accommodation, and n, and the the Roadway Royal Roadway Royal was meant to provide high quality services for big spenders. was meant to provide high quality services for big spenders. Campbell felt that revenues could be quadrupled if only clients Campbell felt that revenues could be quadrupled if only clients understand the understand the distinction among the three types of facilities offered.distinction among the three types of facilities offered.
Keen on developing a viable strategy to eliminate the brand confKeen on developing a viable strategy to eliminate the brand confusion and make usion and make clear the distinctions, John Campbell conducted a customer surveclear the distinctions, John Campbell conducted a customer survey of those who y of those who had used each type of facility and found the following. The conhad used each type of facility and found the following. The consumers were sumers were blissfully unaware of the differences among the three types of fblissfully unaware of the differences among the three types of facilities. Many acilities. Many complained about how old the buildings were and how poorly the fcomplained about how old the buildings were and how poorly the facilities were acilities were maintained. The quality of services was also rated as poor. maintained. The quality of services was also rated as poor.
Furthermore, when rumor seems to have spread that one of the ideFurthermore, when rumor seems to have spread that one of the ideas of as of Campbell was a name change to differentiate the facilities, franCampbell was a name change to differentiate the facilities, franchise chise owners became angry and the mixed messages they gave to the custowners became angry and the mixed messages they gave to the customers omers had not helped clients to understand the differences.had not helped clients to understand the differences.
Campbell thought that he first needed to understand how the diffCampbell thought that he first needed to understand how the different erent classifications would be important to the several classes of cliclassifications would be important to the several classes of clients, and then ents, and then he could develop a marketing strategy that would enhance revenuehe could develop a marketing strategy that would enhance revenues. s. Simultaneously, he recognized that unless the franchise owners fSimultaneously, he recognized that unless the franchise owners fully ully cooperated with him in all his plans, mere face lifting and imprcooperated with him in all his plans, mere face lifting and improvement of ovement of customer service would not bring in the added revenues he hoped customer service would not bring in the added revenues he hoped for.for.
a. Identify the problemb. Develop a theoretical frameworkc. Develop at least four hypotheses
Major R.Q. Major R.Q. the Broadest Question in the study.the Broadest Question in the study.
Major R.Q.= 1 and Minor R.Qs = No. of Variables.Major R.Q.= 1 and Minor R.Qs = No. of Variables.
Minor R.Qs Minor R.Qs help design Data Collection Instrument.help design Data Collection Instrument.
Begin with wordsBegin with words ““whatwhat””, , ““howhow””, or , or ““to what extentto what extent””..
Use nonUse non--directional wording and adirectional wording and avoid wording that suggests a void wording that suggests a relationship between variables, such asrelationship between variables, such as ““affectaffect", "", "influenceinfluence", ", ""impactimpact", "", "determinedetermine””..
Use openUse open--ended questions without reference to the literature or ended questions without reference to the literature or theory, unless otherwise dictated by the research design.theory, unless otherwise dictated by the research design.
A hypothesis is a testable statement of logically defined relationship between two or more variables/groups.
Hypotheses are developed based on the theoretical framework to test whether the relationships that have been theorized do in fact hold true.
By testing the hypotheses (using software like SPSS), and confirming/ or rejecting relationships, it is expected that solutions can be found to correct the problem.
Directional Hypotheses are those in which the direction of the relationship between 2 variables or 2 groups is indicated. They include terms such as positive, negative, more than, less than.
Examples:
- The greater the stress experienced in the job, the lower the jobsatisfaction of employees.
Non-Directional Hypotheses are those which postulate a relationship or difference (between 2 variables or 2 groups), but offer no indication of the direction of these relationships or differences.
They are formulated either because the relationships or differences have never been previously explored, thus no basis for indicating direction, or because there have been conflicting findings in previous studies on the variables.Examples:
- There is a relationship between age and job satisfaction.- There is a difference between the work ethic values of American and Asian
Null hypothesis imply that any differences found between 2 groups or any relationship found between 2 variables is due to random sample errors, and not due to any “true”differences between 2 sample groups (men & women), or relationships between 2 variables (sales & profits).
Thus, Null hypothesis is formulated to be tested for possible rejection.
If analysis lead us to rejecting the Null Hypothesis, then Alternative hypothesis could be supported.
Theory allows us to have faith in the Alternative hypothesis generated in research.
Surveys (Quantitative)Surveys (Quantitative)Sample from Population to make inferences about the populationSample from Population to make inferences about the populationSample selection is critical and statistical techniques neededSample selection is critical and statistical techniques needed
CrossCross--sectional Studies/Surveys (Quantitative)sectional Studies/Surveys (Quantitative)Take a snapshot of an ongoing situationTake a snapshot of an ongoing situationConducted when there is constraint of time and resourcesConducted when there is constraint of time and resourcesTo find correlation/association between variablesTo find correlation/association between variablesAre not expensive and are conducted simultaneouslyAre not expensive and are conducted simultaneously
Case Studies (Case Studies (QualitativeQualitative))It focuses on understanding the dynamic present within single seIt focuses on understanding the dynamic present within single settingttingIt is exploratory Research, needs inIt is exploratory Research, needs in--depth and long time research, to depth and long time research, to understand, and to collect dataunderstand, and to collect data
Types of Research Methodology
Ethnography (Ethnography (QualitativeQualitative))Ethno (folk) Ethno (folk) graphygraphy (description)(description)It is used to understand the observed patterns of human activitiIt is used to understand the observed patterns of human activitiesesData collection through observation (member in a team)Data collection through observation (member in a team)It takes long timeIt takes long time
Triangulation,Triangulation, use different research approachesuse different research approaches
Cross-Sectional Designs:• A type of research design involving the collection of information from any given
sample of population only once.• It is the most frequently used type of descriptive design in marketing research.• These designs may be either:
• Single Cross-Sectional, or• Multiple Cross-Sectional
Longitudinal Designs:• A type of research design involving a fixed sample of population which is
measured repeatedly. The sample remains the same over time (i.e. Same respondents are studied over time).
• In contrast to the cross-sectional design, which gives a snapshot of the variables of interest at a single point of time, longitudinal study provides a series of pictures that , when viewed together, offer an in-depth view of the situation and the changes that take place over time.
• The term Panel is often used to describe the sample of respondents who have agreed to provide information at specific intervals over an extended period.
Detecting ChangeDetecting ChangeLarge amount of data collectionLarge amount of data collectionAccuracyAccuracyRepresentative SamplingRepresentative SamplingResponse biasResponse biasTime & CostTime & Cost
------+++++
++++++------
Note: A Note: A ““++”” indicates a relative advantage over the other indicates a relative advantage over the other design, whereas a design, whereas a ““--”” indicates a relative disadvantage.indicates a relative disadvantage.
Relative Advantages and Disadvantages of Relative Advantages and Disadvantages of Longitudinal & CrossLongitudinal & Cross--Sectional DesignsSectional Designs
Cross-Sectional Designs:• A marketing manager is interested in measuring customer
demand towards a specific newly developed product that is to be launched by the company. Data are collected in mall interviews pattern from the main 6 retailers outlets.
Longitudinal Designs:• A marketing manager is interested in attracting the pattern of
sales of a particular product in 4 different regions of the country on a quarterly basis for the next 2 years. Since data are collected several times to answer the same issues, it is considered longitudinal.
• Nominal Scale: A scale whose numbers serve only as labels or tags for identifying and classifying objects with a strict one-to-one correspondence between the numbers and the objects.
• Ordinal Scale: A ranking scale in which numbers are assigned to objects to indicate the relative extent to which some characteristic is possessed. Thus, it is possible to determine whether an object has more or less of a characteristic than some other object.
• Interval Scale: A scale in which the numbers are used to rate objects such that numerically equal distances on the scale represent equal distances in the characteristic being measured.
• Ratio Scale: The scale. It allows the researcher to identify or classify objects, rank order the objects, and compare intervals or differences. It is also meaningful to compute ratios of scale values.
• A comparative scaling techniques in which respondent is presented with 2 objects at a time and asked to select one object in the pair according to some criterion.• This technique is frequently used when the stimulus objects are physical products.• The data obtained are ordinal in nature (e.g. a respondent may state that he/she likes Crest more than Colgate).•It is used when the number of brands is limited, However, with a large number of brands, the number of comparisons become unwieldy.• Coca-Cola is reported for conducting more than 190,000 paired comparisons before introducing new Coke.• This is the most widely used comparative scaling technique.
Instructions: We are going to present you with ten pairs of shampoo brands. For each pair, please indicate which one of the two brands of shampoo you would prefer for personal use.Recording Form:
Jhirmack Finesse Vidal Sassoon
Heads & Shoulders
Pert
Jhirmack 0 0 1 0 Finesse 1a 0 1 0 Vidal Sassoon 1 1 1 1 Head & Shoulders 0 0 0 0 Pert 1 1 0 1 Number of Times Preferredb
3 2 0 4 1
aA 1 in a particular box means that the brand in that column was preferred over the brand in the corresponding row. A 0 means that the row brand was preferred over the column brand. bThe number of times a brand was preferred is obtained by summing the 1s in each column.
Disadvantages of Paired Comparison ScalingDisadvantages of Paired Comparison Scaling
The most common method of taste testing is paired comparison. The consumer is asked to sample two different products and select the one with the most appealing taste. The test is done in private and a minimum of 1,000 responses is considered an adequate sample. A blind taste test for a soft drink, where imagery, self-perception and brand reputation are very important factors in the consumer’s purchasing decision, may not be a good indicator of performance in the marketplace. The introduction of New Coke illustrates this point. New Coke was heavily favored in blind paired comparison taste tests, but its introduction was less than successful, because image plays a major role in the purchase of Coke.
Disadvantages
• The order in which objects are presented may bias the results.
• Bear little resemblance to the marketplace situations which involves selection from multiple alternatives.
• Respondents may prefer one object over others, but they may not like it in the absolute sense.
Rank Order Scaling• A comparative scaling techniques in which respondent is presented with several objects simultaneously and asked to order or rank them according to some criterion.• Commonly used to measure preferences for brands. • Like Paired Comparison, this approach is also comparative in nature, and it is still possible that the respondent may dislike the brand ranked 1 in the absolute sense.• However, this type of scaling more closely resembles the shopping environment.• Also, most respondent understand the instructions for ranking.• The major disadvantage is that this technique produces only ordinal data.
Preference for Toothpaste Brands Using Rank Order Scaling
Instructions:Instructions: Rank the various brands of toothpaste in order of Rank the various brands of toothpaste in order of preference. Begin by picking out the one brand that you like mospreference. Begin by picking out the one brand that you like most t and assign it a number 1. Then find the second most preferred brand assign it a number 1. Then find the second most preferred brand and and assign it a number 2. Continue this procedure until you haveand assign it a number 2. Continue this procedure until you haveranked all the brands of toothpaste in order of preference. The ranked all the brands of toothpaste in order of preference. The least least preferred brand should be assigned a rank of 10. preferred brand should be assigned a rank of 10.
No two brands should receive the same rank number.No two brands should receive the same rank number.
The criterion of preference is entirely up to you. There is no rThe criterion of preference is entirely up to you. There is no right or ight or wrong answer. Just try to be consistent.wrong answer. Just try to be consistent.
Constant Sum Scaling• A comparative scaling techniques in which respondent is required to allocate a constant sum of units such as points or dollars among a set of stimulus objects with respect to some criteria.
• If attribute is not important, the respondent should assign Zero. If attribute is twice as important as some other attribute, it receives as many points.
• The sum of all points should be 100, hence, the name of the scale.
• The attributes are scaled by counting the points assigned to each one by all the respondents and dividing by the number of respondents.
• Results are limited to the context of object scaled, they can not be generalized to other stimuli not included in the study. Hence this method should be considered an ordinal scale.
• The major disadvantage is that respondents may allocate more or fewer units than specified (e.g. 108, or 94)
Importance of Toilet Soap Attributes Using Importance of Toilet Soap Attributes Using Constant Sum ScaleConstant Sum Scale
Instructions
On the next slide are eight attributes of bathing soaps. Please allocate 100 points among the attributes so that your allocation reflects the relative importance you attach to each attribute. The more points an attribute receives, the more important the attribute is. If an attribute is not at all important, assign it zero points. If an attribute is twice as important as some other attribute, it should receive twice as many points.
• Also referred to as graphic rating scale. In this techniques respondents rate objects by placing a mark at the appropriate position on a line that runs from one extreme of the criterion variable to the other.• The form of the continuous scale may vary considerably. For example, scale point in can be in the form of points or brief description, may be few or many.
• Once the respondent has provided the ratings, the researcher divides the line into as many categories as desired, and assigns scores based on the categories into which the rating falls.• They are easy to construct, however scoring is unreliable. Moreover, they provide little information, hence, their use in marketing research is limited.
Probably the worst ------------------------------------------------------------------ Probably the best
Probably the worst ------------------------------------------------------------------ Probably the best
• A measurement scale in which the respondents are provided with a scale that has a number or a brief description associated with each category.• The categories are ordered in terms of the scale position.• The respondents are required to select the specified category that best describes the object being rated.• Itemized scales are widely used in marketing research.• The commonly used itemized scales are: Likert, Semantic Differential, and Stapel scales.
• Named after its developer, Rensis Likert.• It is a widely used rating scale that requires the respondents to indicate a degree of agreement or disagreement with each of a series of statements about the stimulus objects.• Each scale item has 5 response categories, ranging from “strongly disagree”to “strongly agree”
I get a feeling of accomplishment from the work I am doing
Strongly Disagree
1
Disagree
2
Neutral
3
Agree
4
Strongly Agree
5
• Likert scale is easy to construct and administer. Respondents readily understand how to use the scale, making it suitable for mail questionnaire and telephone interviews.• The major disadvantage of this scale is that it takes longer to complete than other scales, because respondents have to read each statement.
• In this technique, respondents rate objects on a seven-point rating scale bounded at each end by one of two bipolar adjectives.• Respondents mark the blank that best indicates how they would describe the object being rated.• The –ve adjective or phrase sometimes appears at the left side of the scale and sometimes at the right. This controls the tendency of some respondents (particularly those with very +veor very –ve attitudes) to mark the right or left sides without reading the labels.• Items on the scale may be scored on a 1 to 7 scale. • This techniques is popular in marketing research, and widely used in comparing brand, product, and company images.
A Semantic Differential Scale for Measuring SelfA Semantic Differential Scale for Measuring Self-- Concepts, Concepts, Person Concepts, and Product ConceptsPerson Concepts, and Product Concepts
• It is a unipolar rating scale with 10 categories numbered from -5 to +5, without a neutral point (Zero). This scale is usually presented vertically.
• Respondents are asked to indicate how accurately or inaccurately each term describes the object by selecting an appropriate numerical response category.
+5+4+3+2+1
High Quality-1-2-3-4-5
• The more accurately, the larger the plus number you should choose, and the less accurately, the larger the minus number the respondent should choose.
• This scale produces results similar to the semantic differential. Moreover, Stapel scales do not require a pretest of the adjectives or the phrases to ensure true bipolarity. However,some researchers believe that this scale is confusing and difficult to apply. Of the 3 itemized rating scales, this is theleast used.
1) Number of Categories Although there is no single, optimal number, traditional guidelines suggest that there should be between five and nine categories
2) Balanced vs. unbalanced In general, the scale should be balanced to obtain objective data
3) Odd/ even no. of categories If a neutral or indifferent scale response is possible from at least some of the respondents, an odd number of categories should be used
4) Forced vs. non-forced In situations where the respondents are expected to have no opinion, the accuracy of the data may be improved by a non-forced scale
5) Verbal description An argument can be made for labeling all or some (Verbal or Numerical) scale categories. The category descriptions should
be located close to the response categories
6) Physical form6) Physical form A number of options should be tried and the best selected
Summary of Itemized Scale DecisionsSummary of Itemized Scale Decisions
Extremely goodExtremely good Extremely goodExtremely goodVery goodVery good Very goodVery goodGoodGood GoodGoodBad Bad Somewhat goodSomewhat goodVery badVery bad Bad Bad Extremely badExtremely bad Very badVery bad
Balanced and Unbalanced ScalesBalanced and Unbalanced Scales
A variety of scale configurations may be employed to measure the gentleness of Cheer detergent. Some examples include.Cheer detergent is:1) Very harsh --- --- --- --- --- --- --- Very gentle
2) Very harsh 1 2 3 4 5 6 7 Very gentle
3) . Very harsh. Neither harsh nor gentle . Very gentle
4) ____ ____ ____ ____ ____ ____ ____ Very Somewhat Neither harsh Somewhat Gentle Very
Thermometer ScaleThermometer ScaleInstructions:Instructions: Please indicate how much you like McDonald’s
hamburgers by coloring in the thermometer. Start at the bottom and color up to the temperature level that best indicates how strong your preference is.
Form:Form:
Smiling Face ScaleSmiling Face ScaleInstructions:Instructions: Please point to the face that shows how much you like the
Barbie Doll. If you do not like the Barbie Doll at all, you would point to Face 1. If you liked it very much, you would point to Face 5.
Form:Form:11 22 33 44 55
Some Unique Rating Scale ConfigurationsSome Unique Rating Scale Configurations
• Certain practical factors should be considered in selecting scaling techniques for a particular marketing research problem.• Theses include:
• The level of information (nominal, ordinal, interval, or ratio) desired.• The capabilities and knowledge of the respondents.• The characteristics of the stimulus objects.• Method of administration of data collection.
Accurate? Valid?
Generilizable?
• As a general rule, using the scaling technique that will yield the highest level of information in a given situation will permit the use of the greatest variety of statistical analysis tools.
A A Target PopulationTarget Population is the collection of elements that possess the information sougis the collection of elements that possess the information sought by the ht by the researcher and about which inferences are to be made.researcher and about which inferences are to be made.
A A SampleSample is made up of some of the members of the population.is made up of some of the members of the population.A A Representative SampleRepresentative Sample of any population should be so drawn that every member of that of any population should be so drawn that every member of that
population has a specified nonpopulation has a specified non--zero probability of being included in the sample. In other wordszero probability of being included in the sample. In other words, , Every member of the population has a statistically equal chance Every member of the population has a statistically equal chance of being selected.of being selected.
A A Sampling FrameSampling Frame is a list of the population from which all the sampling units ais a list of the population from which all the sampling units are drawn. For re drawn. For example, in a large company, a list of all employees will form texample, in a large company, a list of all employees will form the sampling frame from which you he sampling frame from which you can take a sample.can take a sample.
Sample SizeSample Size depends mainly on the size of your study, and the size of your depends mainly on the size of your study, and the size of your population. You may population. You may use the entire population.use the entire population.
A A ParameterParameter is the summary description of a given variable in a population.is the summary description of a given variable in a population. The mean income of The mean income of all families in a city and the age distribution of the cityall families in a city and the age distribution of the city’’s population are parameters.s population are parameters.
A A StatisticStatistic is the summary description of a given variable in the study samis the summary description of a given variable in the study sample. Thus, the mean ple. Thus, the mean income computed from a survey sample and the age distribution ofincome computed from a survey sample and the age distribution of that sample are statistics.that sample are statistics.
A good sample must be:A good sample must be:Chosen at random.Chosen at random.Large enough to satisfy the needs of the research.Large enough to satisfy the needs of the research.Unbiased (e.g. volunteers to participate in the study). Unbiased (e.g. volunteers to participate in the study).
Probability sampling provides an efficient method for selecting a sample that should adequately reflect the variation that exists in the population as a whole.
A basic principle of Sampling is the following: “A sample will be representative of the population from which it is selected if all members of the population have an equal chance of being selected in the sample.”
Probability Samples will never be perfectly representative of the population , but are typically more representative than other types of samples.
Probability Theory permits you to estimate the accuracy or representativeness of your sample.
1. Simple 1. Simple Random Sampling (for small & Large Sample):Random Sampling (for small & Large Sample):For small population, numbers might be chosen at random, where For small population, numbers might be chosen at random, where
every member of the population is given a number.every member of the population is given a number.For large populations, numbers are taken from For large populations, numbers are taken from ““Tables of Random Tables of Random
NumbersNumbers””..Bias will occur if sample is chosen by a person, as it will leadBias will occur if sample is chosen by a person, as it will lead to to
favoritism, or chosen haphazardly.favoritism, or chosen haphazardly.
Number the names from 1 to 7,500, and then use the Random NumberNumber the names from 1 to 7,500, and then use the Random Numbers Table to select a s Table to select a random sample of 300. The division of 300 into 7,500 is called trandom sample of 300. The division of 300 into 7,500 is called the he Sampling Fraction. Sampling Fraction. In this case In this case it isit is 25.25.That means, a random selection of one name in every 25. However,That means, a random selection of one name in every 25. However, the selection must be the selection must be truly truly Random.Random.If your Sampling Frame is in electronic version, SRS can be seleIf your Sampling Frame is in electronic version, SRS can be selected automatically through the cted automatically through the use of a use of a Sampling Computer ProgramSampling Computer Program. .
2. Stratified Sampling (for relatively small sample)2. Stratified Sampling (for relatively small sample)::Random sampling might result in some members of the Random sampling might result in some members of the
population being significantly under or over presented.population being significantly under or over presented.Stratified sampling overcomes this problem as each identifiable Stratified sampling overcomes this problem as each identifiable
““StrataStrata”” of the population is taken into account.of the population is taken into account.
Example:Example: Population: 850 (580 females and 270 males)Population: 850 (580 females and 270 males)Sample Size: 100Sample Size: 100
580/850*100=68 Females 580/850*100=68 Females AND AND
270/850*100=32 Males270/850*100=32 Males
Real Case:Real Case: If your sampling frame consists of all the employees in a partiIf your sampling frame consists of all the employees in a particular company, cular company, you may identify the following strata: Directors, Managers, admiyou may identify the following strata: Directors, Managers, administrators, and nistrators, and production workers. And if 25% of the whole employees were produproduction workers. And if 25% of the whole employees were production workers, you ction workers, you have to ensure that 25% of your sample are production workers.have to ensure that 25% of your sample are production workers.
3. Cluster Sampling (for reducing physical areas covered/intervi3. Cluster Sampling (for reducing physical areas covered/interviewing cost)ewing cost)::
Takes the advantage of the fact that most populations are strucTakes the advantage of the fact that most populations are structured in some way, or could tured in some way, or could be divided into subbe divided into sub--sections according to certain characteristics (i.e. they can be sections according to certain characteristics (i.e. they can be clustered).clustered).
Random selection is made from a sampling frame listing groups rRandom selection is made from a sampling frame listing groups rather than individuals. ather than individuals. Every individual in a selected group is then interviewed.Every individual in a selected group is then interviewed.
Example:Example: Population is clustered into 3 levels: State, region, and local Population is clustered into 3 levels: State, region, and local sectionsection13 State, each contains 2513 State, each contains 25--40 region, each contains 12040 region, each contains 120--260 Sections.260 Sections.
Work systematically from level to level with the table of randomWork systematically from level to level with the table of random numbers.numbers.First we choose, say, 4 out of the 13 state at random. First we choose, say, 4 out of the 13 state at random. (4 States)(4 States)From each of those four we choose, say, 6 regions at random. From each of those four we choose, say, 6 regions at random. (24 Regions)(24 Regions)From each of those 24 regions we choose, say, 8 sections at ranFrom each of those 24 regions we choose, say, 8 sections at random. dom. (192 Section)(192 Section)Finally, from each of those 192 sections, we choose 10 people aFinally, from each of those 192 sections, we choose 10 people at random for our sample.t random for our sample.Result: < 2,000 names/addresses, accurately representative & clResult: < 2,000 names/addresses, accurately representative & clustered geographically.ustered geographically.
4. 4. Systematic Sampling (Virtually Identical to SRS):Systematic Sampling (Virtually Identical to SRS):Occasionally, when the Sampling Frame is available, Occasionally, when the Sampling Frame is available,
researchers usually employ this method rather than the SRS.researchers usually employ this method rather than the SRS.In Systematic Sampling , every In Systematic Sampling , every kkth th element in the list is chosen element in the list is chosen
for inclusion in the sample.for inclusion in the sample.
First divide the population by the required sample size: 300/10First divide the population by the required sample size: 300/10=30=30Start by selecting a randomly chosen number between 1 and 30, saStart by selecting a randomly chosen number between 1 and 30, say 23. Then choose the y 23. Then choose the thirtieth one after that, until ten have been selected.thirtieth one after that, until ten have been selected.
23, 53, 83, 113, 143, 173, 203, 233, 263, 29323, 53, 83, 113, 143, 173, 203, 233, 263, 293This method involves one danger:This method involves one danger: If the list of elements is If the list of elements is
arranged in a cyclic pattern that coincides with the sampling arranged in a cyclic pattern that coincides with the sampling interval, a grossly biased sample might be drawn. This is calledinterval, a grossly biased sample might be drawn. This is called““PeriodicityPeriodicity””..
Despite the superiority of probability sampling methods, nonDespite the superiority of probability sampling methods, non--probability methods are probability methods are sometimes used instead, for those situations in which the probabsometimes used instead, for those situations in which the probability methods are expensive, ility methods are expensive, and/or when precise representativeness is not necessary.and/or when precise representativeness is not necessary.
Judgmental Sampling (for Pilot Work)Judgmental Sampling (for Pilot Work)To select your sample on the basis of your own knowledge of theTo select your sample on the basis of your own knowledge of the
population, and the nature of your research aims.population, and the nature of your research aims.Researcher chooses the elements to be included in the sampleResearcher chooses the elements to be included in the sample, as , as
he or she believes that they are representative of the populatiohe or she believes that they are representative of the population of n of interest.interest.
During the initial design of your questionnaire, you might wanDuring the initial design of your questionnaire, you might want to t to select the widest variety of respondents to test the broad appliselect the widest variety of respondents to test the broad applicability cability of your questions.of your questions.
Findings would NEVER represent any meaningful population, but Findings would NEVER represent any meaningful population, but might effectively uncover any particular defects in your researcmight effectively uncover any particular defects in your research h instrument.instrument.
Common examples include: test markets selected to determine the Common examples include: test markets selected to determine the potential of a new product.potential of a new product.
2. Quota Sampling (for Marketing Research)2. Quota Sampling (for Marketing Research)Can be viewed as 2 stage restricted judgmental sampling.Can be viewed as 2 stage restricted judgmental sampling.The first stage consists of developing control categories, or The first stage consists of developing control categories, or quotas, of population elements.quotas, of population elements.The relevant control characteristics may include: sex, age, raceThe relevant control characteristics may include: sex, age, race, , and are identified on the basis of judgment.and are identified on the basis of judgment.In the second stage, sample elements are selected based on In the second stage, sample elements are selected based on judgment.judgment.The only requirement is that the elements selected fit the contrThe only requirement is that the elements selected fit the control ol characteristics.characteristics.Often, the quotas are assigned so that the proportion of the Often, the quotas are assigned so that the proportion of the sample elements possessing the control characteristics is the sample elements possessing the control characteristics is the same as the proportion of population elements with theses same as the proportion of population elements with theses characteristics.characteristics.Quota sampling attempts to obtain representative samples at a Quota sampling attempts to obtain representative samples at a relatively low cost.relatively low cost.Under certain conditions, this method obtains results close to Under certain conditions, this method obtains results close to those of probability sampling.those of probability sampling.
A sampleA sample’’s Accuracy is more important than its size.s Accuracy is more important than its size.A properly drawn sample can give us more reliable estimates on A properly drawn sample can give us more reliable estimates on a population of a population of
many millions.many millions.Statistical tables exist which will show the Statistical tables exist which will show the Sampling Error Sampling Error which is theoretically which is theoretically
obtainable for samples of different sizes. Thus, the accuracy ofobtainable for samples of different sizes. Thus, the accuracy of a given result might be a given result might be plus or minus:plus or minus:
2% of a sample of 2,000, OR2% of a sample of 2,000, OR6.5% of a sample of 2006.5% of a sample of 200
Example:Example:A finding that 30% of our sample said A finding that 30% of our sample said ‘‘YESYES’’ to a particular question must therefore to a particular question must therefore
be interpreted to mean that the true figure for the population wbe interpreted to mean that the true figure for the population will most likely lie ill most likely lie between 28% between 28% -- 32%, or between 23.5% 32%, or between 23.5% -- 36.5%, depending on Sample size. 36.5%, depending on Sample size.
Theoretically, adequate sample should be 3%Theoretically, adequate sample should be 3%--6% of the total population. However, 6% of the total population. However, sample size will be determined by other factors such as: clustersample size will be determined by other factors such as: cluster size, required accuracy size, required accuracy of estimates, nature of the population, and of course by the conof estimates, nature of the population, and of course by the constraints of time and cost.straints of time and cost.
Several focus groups are used in any research project. There may be variation between groups, but an individual group must be comprised of similar types of people.
Recruitment is usually done by telephone. Participants are screened for verbal ability, experience and willingness to be candid about their opinions.
Setting is important. Many firms use central facilities with one-way mirrors and sophisticated electronic recording equipment. Hotel rooms and other meeting facilities can be quite acceptable venues for a focus group especially for the professional and business community.
Major issues are reviewed prior to the group. The moderator should prepare the agenda (also termed a "protocol") which will guide the group discussion.
Focus groups are usually Video/Audio-taped Under no circumstances should this be kept secret; participants must be informed both of the recording devices and that they are being observed. Anyone who is uncomfortable may leave.
Group dynamics are critical. The meeting must be positive and lead to easy interaction on the part of group members.
Homogenous group composition is the norm. It is usually unproductive to combine young and old, or junior and senior in the same group unless they are united by a strong common interest. The comfort level respondents have with each other is extremely important to gaining useful information.
Moderator skill is very important. It requires maturity, research experience, andsubstantive knowledge of the field.
A clear agenda for the group discussion is essential. The topic must be familiar to all participants with the agenda consisting of 5 – 10 main issues. A common mistake is to "load" the agenda with a detailed set of questions which need to be explored. When participants are rushed, they will not provide in-depth reactions. For example, in a two hour (120 minute) focus group of 12 participants with 20 issues, each has 30 seconds on average to react to each issue.
Tolerate silence. It is important to respect the need for individuals to collect their thoughts. If you have spent time explaining why their thoughtful opinions are important to the research, why rush them for information?
Enlist the aid of dominant participants. Ask another person to react to the dominant participant’s comment. In this way, others have the opportunity to enter the discussion. Drawing out those who are reluctant to speak is the mark of a skilled moderator (i.e. ensure that everyone speaks).
Conflict is an indicator of mixed opinions on the issue. Exploring these differences is important. Of course, personality conflicts among group members are unproductive and mediation skills are another indicator of a god moderator.
A minimum number of groups are needed if the population is low diverse. Fewer sessions may be conducted if the range of issues is slight and if variation within the population is expected to be low.
Structured InterviewsStructured interviews are those conducted when it is known what information is needed. The interviewer has a list of predetermined questions to be posed to the respondents. The questions are likely to focus on factors that had surfaces during the unstructured interviews.
The same questions are asked to everybody. Sometimes, however, the experienced researcher, based on the situation might take a lead from an answer and ask other relevant questions not on the interview protocol.
At this point of research process, the interviewer does not have the luxury of initiating new topics, adding new questions, or even following up interesting ideas.
You should be bound by the requirement of Stimulus Equivalence.
Interviewer should be able to distinguish three types of questions:
Factual Questions, where he/she is expected to read out the question as printed, but has some latitude to offer explanations or correct misunderstanding.
Attitude & Opinion Questions, in which interviewers are forbidden to explain or reword in any way.
A questionnaire is a formalized written set of questions for obtaining information from respondents. Questionnaires are a efficient data collection mechanism when the researcher knows exactly what is required and how to measure variables of interest.
In the present context, the term “Questionnaire” can be fairly used to cover postal questionnaires, self-administered questionnaires and structured interview script.
Advantages of the Survey Method:
1. Questionnaire is simple to administer.
2. Use of fixed-response questions reduces the variability in the responses which may be caused by interviewers.
3. Coding, editing and analysis of data are relatively simple
Disadvantages of the Survey Method:
1. Respondents may be unable or unwilling to respond if the information requested is sensitive or personal.
The Delphi Method is based on a structured process for collecting and distilling knowledge from a group of experts by means of a series of questionnaires interspersed with controlled opinion feedback
The objective of most Delphi applications is the reliable and creative exploration of ideas or the production of suitable information for decision making.
The outcome of a Delphi sequence is nothing but opinion (Qualitative).
Single experts sometimes suffer biases; group meetings suffer from “follow the leader” tendencies and reluctance to abandon previously stated opinions.
The main point behind the Delphi method is to overcome the disadvantages of Personal Interviews and Focus group discussions.
The team will be responsible for developing the entire exercise, choice of experts, development of questionnaires, analyzing the questionnaires and for feedback, as well as for monitoring the process and for preparing report.
The Delphi technique requires a coordinator to organize requests for information, information received, and to be responsible for communication with the participants.
The Delphi technique also requires an efficient communication channel to link the coordinator with each of the participants.
Once the list of nominees is formed, each person should be contacted individually. Experts from the required disciplines are first identified and asked to participate in the inquiry.
The results of a Delphi depend on the knowledge and cooperation of the panelists
Knowledgeable persons are usually identified through literature searches for who has published on the subject under study, recommendations from institutions.
One possibility that helps ensure that the required skills are represented is to form a matrix in which the required skills are listed.
The rows of the matrix consist of the names of the prospective participants.
The cells are checked to indicate the "coverage" that the nominees are expected to provide. One interesting method for choosing the experts is the so-called Lock-and-Key Approach.
In this approach, the researcher attempt to match the capabilities of participants with the requirements of the questions, and questions are addressed only to those respondents whose profiles match.
Weighted sum method is used to evaluate the experts based on thesame parameters, which will be used to evaluate the questions.
Steps of the Delphi Method3. Development of the first questionnaire (1st Round)The next step is to formulate the questions. The questions must
be sharp and answerable. Phrasing of questions is important. For this reason, it is desirable to conduct a test of the instrument.
In-depth interviews with experts have been used with great success as an alternative to questionnaires. In this approach, the same kinds of experts are first identified, invited to participate, assured of their anonymity. Appointments are made at the convenience of the interviewees. Interview protocols are prepared and tested to elicit judgments. Feedback can be introduced if two rounds of interviews are employed; however, single-round studies are used more frequently. An advantage of one-on-one interviews is that they provide flexibility, which is absent in questionnaires.
4. Transmission of the first round questionnaireOnce qualified in this way, the questionnaire would be sent to the
participants. The cover letter would remind the participants about the objectives of the study, establish the schedule for the response, and include the return address.
In most cases, the first questionnaire poses the problem in broad terms and invites answers and comments. For example, what actions should be taken to improve the quality of renting services?
The first questionnaire asks each participant to engage in individual brainstorming so as to generate as many ideas as possible for dealing with the issue. Essay-type answers could be permitted at this stage. Each participant returns the list anonymously to the coordinator.
5. Analysis of the first round responses and feedback iterationSome studies indicate that a response rate from 40 to 75 percent of the
participants can be anticipated.
Responses obtained from the panel are collated by a central coordinator and fed back to the respondents in a synthesized form. A feedback round would be used to present the results, thereasons for the extreme positions, and a call for reassessment.
The reasons, in Delphi style, would be placed in front of the participants who answered the first round. They would be then asked to reconsider their former answers in view of the reasons for the extreme opinions.
6. Preparation and transmission of the second round questionnaireResponses to the first questionnaire are used to construct the second
questionnaire. The coordinator prepares and sends a second questionnaire to participants that contains all of the ideas sent in response to the first questionnaire and provides space for participants to refine each idea, clarify issues, identify areas of agreement or disagreement, to comment on each idea’s strengths and weaknesses for addressing the issue, to develop priorities and to identify new ideas.
This interactive process can be repeated as many times as appropriate. The aim of each round or iteration is for responses to become stable.
Participants anonymously record their responses to the second questionnaire and return them to the coordinator.
The coordinator analyses the responses, summarizes them, and asks for additional clarifications, strengths, weaknesses, and new ideas. Preparation, transmission, and analysis are reiterated aslong as desired or necessary to achieve stability in the results.
If desired, the coordinator performs iterations of the preceding process until it becomes clear that no new ideas are emerging and that all strengths, weakness, and opinions have been identified.
Steps of the Delphi Method8. Reporting on the exerciseThe outcome of a Delphi sequence is nothing but opinion. The
results of the sequence are only as valid as the opinions of theexperts who made up the panel.
The data from a Delphi can be displayed in several ways. The group judgment should be based on the median rather than the mean, since single extreme answers can "pull" the mean unrealistically. Furthermore, it is incumbent on the analyst to show the spread of opinion, which can be done by showing the inter-quartile range (the range that contains the answers of 50 percent of the respondents).
Delphi studies are difficult to perform well. A great deal of attention must be given to the choice of participants; the questionnaires must be meticulously prepared and tested to avoid ambiguity.
Multi-round studies require a great deal of time; inevitably, some participants will drop out during the process.
Finally, a weakness of the Delphi method is the time that it takes. A single round can easily require three weeks; a three-round Delphi is at least a three to four month affair, including preparation and analysis time.
The general principle is that the first “layer” of a questionnaire that needs to be dealt with is the questionnaire’s objectives.
You can not begin to formulate questions and worry about wording and formatting unless you know what you want to accomplish with your questions and words.
Begin with an introduction which includes the questionnaire's purpose, who is conducting it, to what use the information will go, and confidentiality.
Open-ended questions are those where respondents provide their own answers to the question, without any previously provided options. These are fill-in-the-blank responses. They allow respondents to express their own thoughts and comments but are more demanding of both the respondent and the person doing the analysis.
Close-ended questions have answer options provided and respondents must select either one answer or multiple answers from what is given. These questions have greater uniformity in responses but depend on your knowing and including all relevant responses. Responses for close-ended questions must be exhaustive [i.e. include all possible answers] and also mutually exclusive in providing for the selecting of a single response [without the choice seeming to belong to more than one option].
1. Multiple-Choice Questions: Respondents are asked to select one or more of the alternatives given.
How frequently is a systematic monitoring of work progressing?Every monthEvery quarterTwice a yearOnce a year
The response alternatives should include the set of all possiblechoices.The general guideline is to list all alternatives that may be ofimportance and include an alternative labeled “Other (Please Specify)”
Instructions:Instructions: Rank the various brands of toothpaste in order of Rank the various brands of toothpaste in order of preference. Begin by picking out the one brand that you like mospreference. Begin by picking out the one brand that you like most t and assign it a number 1. Then find the second most preferred brand assign it a number 1. Then find the second most preferred brand and and assign it a number 2. Continue this procedure until you haveand assign it a number 2. Continue this procedure until you haveranked all the brands of toothpaste in order of preference. The ranked all the brands of toothpaste in order of preference. The least least preferred brand should be assigned a rank of 10. preferred brand should be assigned a rank of 10.
No two brands should receive the same rank number.No two brands should receive the same rank number.
The criterion of preference is entirely up to you.The criterion of preference is entirely up to you.
There is no right or wrong answer. Just try to be consistent.There is no right or wrong answer. Just try to be consistent.
3. Dichotomous Questions: Respondents have to choose only one out of 2 alternatives; Yes/No, approve/disapprove, and so on. Often a neutral alternative is supplemented.
Underlying decision-making process may reflect uncertainty, which can be best captured by multiple-choice respondents.
Sometimes, when the neutral response is included, respondents are can avoid taking a decision, thereby biasing the results.
Do you intend to buy a new car within the next 6 months?
4. Likert Scales Questions: It is a widely used rating scale that requires the respondents to indicate a degree of agreement or disagreement with each of a series of statements about the stimulus objects. Each scale item has 5 response categories, ranging from “strongly disagree” to “strongly agree”.
I get a feeling of accomplishment from the work I am doing
Strongly Disagree
1
Disagree
2
Neutral
3
Agree
4
Strongly Agree
5
• Likert scale is easy to construct and administer. Respondents readily understand how to use the scale, making it suitable for mail questionnaire and telephone interviews.
• The major disadvantage of this scale is that it takes longer to complete than other scales, because respondents have to read each statement.
5. A Semantic Differential Scale 5. A Semantic Differential Scale for Measuring Selffor Measuring Self--Concepts, Person Concepts, and Product ConceptsConcepts, Person Concepts, and Product Concepts
The Funnel Approach is a strategy of ordering questions in a questionnaire in which the sequence starts with the general questions, followed by progressively specific questions.
A “Filter Question” is used to exclude some respondents from a particular question sequence if those questions are irrelevant to them.
Start by asking few multiple-choice questions about the respondent’s own habits, and backgrounds. Further questions follow naturally about wider issues. However, don’t leave important questions till the end.
Always use the suggestions of the “Pilot Work”.
A simple way is to attain logical order is to prepare a flow chart for the logical possibilities and then develop branching questions and instructions based on it.
WhatWhat’’s wrong with the following question?s wrong with the following question?Do you approve or disapprove of people that go to football matchDo you approve or disapprove of people that go to football matcheses
1.1. YesYes
2.2. NoNo
Please comment on Why:Please comment on Why:
Wording of Questions
Well, quite a lot. It is an example of an unanswerable question Well, quite a lot. It is an example of an unanswerable question because:because:
The answer categories (Yes or No) are not appropriate to the forThe answer categories (Yes or No) are not appropriate to the form of question. m of question. Approve/Disapprove might be better.Approve/Disapprove might be better.
Also, a leading question, since the Also, a leading question, since the ““approveapprove”” precedes the precedes the ““disapprovedisapprove””..
Fails to offer any kind of middle position, such as: Fails to offer any kind of middle position, such as: ““uncertainuncertain”” OR OR ““it dependsit depends””..
Does not have a Does not have a ““NoNo”” answer or answer or ““Not ApplicableNot Applicable”” to cater for people who have to cater for people who have no experience of football matches.no experience of football matches.
Probe is grammatically incorrect. Poor grammar may irritate respProbe is grammatically incorrect. Poor grammar may irritate respondents.ondents.
WhatWhat’’s wrong with the following question? (Cont.)s wrong with the following question? (Cont.)Do you approve or disapprove of people that go to football matchDo you approve or disapprove of people that go to football matcheses
1.1. YesYes
2.2. NoNo
Please comment on Why:Please comment on Why:
Wording of Questions
Now let us look at the layout of the question:Now let us look at the layout of the question:
No question mark after No question mark after ““matchesmatches””..
More importantly, No instructions have been given to the respondMore importantly, No instructions have been given to the respondent (Tick, ring, or underline).ent (Tick, ring, or underline).
Now come the most serious deficiency in this question that is noNow come the most serious deficiency in this question that is not immediately apparent:t immediately apparent:
What is the PURPOSE of this question?What is the PURPOSE of this question?
No way does it qualify as a factual question, nor will it allow No way does it qualify as a factual question, nor will it allow us to infer how often the respondent go us to infer how often the respondent go to football matches.to football matches.
Length:Length: Not more than twenty words. If longer is needed, break up into Not more than twenty words. If longer is needed, break up into several shorter several shorter sentences. sentences. Always add Always add ““Introductory SentencesIntroductory Sentences”” before each section, and never forget the before each section, and never forget the answering procedures on selfanswering procedures on self--administered questionnaire.administered questionnaire.
Avoid DoubleAvoid Double--Barreled Questions: Barreled Questions: ““What is your evaluation of the price and convenience What is your evaluation of the price and convenience offered by BMWoffered by BMW’’s showroom?s showroom?””. Would a positive answer refer to price, to convenience, or to . Would a positive answer refer to price, to convenience, or to both?both?
Avoid Double Negatives: Avoid Double Negatives: ““Do you know if the deadline has expired yet?Do you know if the deadline has expired yet?””. A negative answer . A negative answer might mean that the respondent does not know, or knows that the might mean that the respondent does not know, or knows that the deadline has not yet expired. deadline has not yet expired. Always be Positive.Always be Positive.
DonDon’’t Know & Not Applicable: t Know & Not Applicable: In case of an opinion question such as In case of an opinion question such as ““Do you think that the Do you think that the government is doing a good job?government is doing a good job?”” a a ““DonDon’’t Knowt Know”” response makes good sense.response makes good sense.
All closed questions should start their lives as open ones All closed questions should start their lives as open ones (Answer categories based on Pilot (Answer categories based on Pilot Work).Work).
DonDon’’t Overt Over--tax the respondentstax the respondents’’ memories.memories.
A “Loaded” word or phrase is one which is emotionally colored, and suggests an
automatic feeling of approval or disapproval. For instance: Nazi, Jewish, intelligent, socialist.
Respondents react not so much to the issue posed by the question, as to the loaded phrase itself.
Trials will show differences in distribution of answers to a question using the phrase “bosses” compared to the one using “Leaders”. Both are loaded phrases,
Here are some items from a questionnaire dealing with different Here are some items from a questionnaire dealing with different aspects ofaspects of
””Job ChoiceJob Choice””
Each item consisted of a pair of choices, of which only one coulEach item consisted of a pair of choices, of which only one could be selected by the respondent. d be selected by the respondent. This technique proved to help overcome the problems of loading.This technique proved to help overcome the problems of loading.
Wording of Questions
A job in which you do a lot of hard thinking.
Where the pay is not very high, but you get pension when you retire.
OROR A job where, once you have learned it, you always know how to do it.
Where the pay is high, but you don’t get a pension when you retire.
Items are fairly balanced and respondents will express their preference without really reacting to such loaded words such as: “Routine Jobs” Or “Security”.
Thus, Much leading comes from a failure to state alternatives.
4. Social Desirability BiasWe should know that many questions are loaded with “Prestige”.People claim many things: making donations, visiting museums.People are reluctant to admit lack of knowledge, and do not like to think of themselves as fools.There is no simple answer to this problem, but there are two general measures that may help:
1. Use filter questions, or word the main question in such a way that a low-prestige answer is equally possible. Example: Instead of “Have you read any of the following magazines at all within the past week?”, use “Have you had time to read any magazines within the past seven days”.
2. Tell the respondents that Accuracy is the prime requirement and that a negative response is as good as positive one.
In this case, the respondent is used as “repository of facts”. An intermediary between the researcher and the required information.
To ascertain Reliability, you should plan to have a number of internal checks. Examples: (1) Introduction of non existent brand name or TV program, endorsement of such items would suggest guessing or carelessness on the part of the respondent. (2) Can you tell me if you have visited any of the following within the past two weeks? Followed by a question about frequency.
To ascertain Validity, a variety of technique, usually known as external checks, where a second, independent source of information is required.
In some cases, quality checks are made by highly trained group of senior interviewers, who can often track down a discrepancy.
In this case, questions are more sensitive to changes in wording, context, emphasis, and so on. Thus, it is almost impossible to assess reliability by asking the same question in another form.
When measuring non-factual questions, we should have a set of questions (Attitude Scales). A set of questions are MORE RELIABLE than a single question, and they give more consistent results.
By using SETS of questions, provided they all relate to the same attitude, we maximize the reliability by reducing the instability due to particular moderating factors such as: mood changes and so on.
Non of the previously mentioned Internal Checks can be applied. Instead, Factor Analysis (Analytic Statistical Tool) can enable you to find out what are the main underlying dimensions of the SET of any Attribute/Variable.
The chief difficulty in assessing the Validity of those questions is the lack of criteria.
Begin with an introduction which includes the questionnaire's purpose, who is conducting it, to what use the information will go, and confidentiality. In mailed questionnaires, use cover letter.Make the first questions non-provoking and interesting.The beginning questions should not be open-ended or questions with a long list of answer choices.Put the more important questions at the beginning.Arrange the order of questions to achieve continuity and a natural flow.Try to keep all questions on one subject together. Put the more general questions first, followed by a more specific question. Place demographic questions (age, gender, race/ethnicity, etc.) at the end of the questionnaire.Keep the whole question and its answers on the same page. Pre-code items and response categories as much as possible, to help in data tabulation and analysis.Use instructions for “self-Administered” Questionnaires.Use transitional statements to build continuity.
Pre-testing the QuestionnaireThis means a careful examination of the individual questions and the
questionnaire as a whole. Allow enough time to incorporate any revisions. Unfortunately, too many people consider pilot testing as a superficial task
they can avoid, but find later that the intended respondents did not understand well enough to deal with most questions.
Is each question measuring what it is intended to measure?Are all the words understood?Are questions interpreted similarly by all respondents?Does each close-ended question have an answer that applies to each respondent?Does the questionnaire create a positive impression, one that motivates people to answer it?Are questions answered correctly? Are some missed? Do some elicit uninterpretable answers?
Does any part of the questionnaire suggest bias on the part of the researcher?
Questionnaires do not emerge suddenly, they have to be created, tested, adapted and developed to maturity before being used in a real field work.
Questionnaire have to be composed and tried out, improved and then tried out again, often several times until we become sure that they can do the assigned job.
This whole lengthy process of designing and trying out questions is usually referred to as: “Pilot Work”.
On whom should our questions be tried out? In principle, respondents in pilot work should be as similar as possible to those in the main enquiry, which means they should be a “Judgmental Sample” to test questions for relevance and sensitivity.
Can Experts help? This temptation should be firmly resisted, mainly because it is based on illusions.
What should be piloted? Every thing.
At this stage, you are not trying to obtain ‘true measures’ from each respondent, you are trying to obtain ideas, concepts, forms of words and material for future questions.
Advance Warning: A letter informing the respondent of the study in advance, and inviting participation.
Explanation of Selection: Explaining sampling method, and WHY ME?
Sponsorship: Research organization pamphlet, Covering letter of introduction from someone expected to be influential.
Envelope: Addressed personally, looks professional rather that a “Junk Mail”.
Incentives: Small may help, larger may not.
Confidentiality: In the sense that only the researcher will have access to them. Steps must be taken to ensure that information will never be published about identifiable persons or organizations. Explicit statement or promise should be made.
Chapter 2Chapter 2 is the reference for everything in the thesis. It is the is the reference for everything in the thesis. It is the database/informationdatabase/information--base/knowledgebase/knowledge--base/eventbase/event--base, factbase, fact--based, based, EvidencesEvidences--Based that you will rely on, in your design (chapter 3) and Based that you will rely on, in your design (chapter 3) and your arguments (chapter 4). your arguments (chapter 4).
Structure of Chapter 2:Structure of Chapter 2:2.1 Introduction/Background/Overview (History of The 2.1 Introduction/Background/Overview (History of The
Chapter 4: Data Analysis, Findings and Discussions
4.2 Inferential 4.2 Inferential
This section is about Hypotheses testing (in case of This section is about Hypotheses testing (in case of quantitativequantitativeresearch paradigm). All research paradigm). All pp--valuesvalues have to be mentioned and have to be mentioned and reported for each test. You should here comment and justify why reported for each test. You should here comment and justify why you rejected or failyou rejected or fail--toto--rejected this particular hypothesis. rejected this particular hypothesis.
Do not be sad or disappointed that a lot of the hypotheses have Do not be sad or disappointed that a lot of the hypotheses have been been failedfailed--toto--reject. Yes, it means that the researcher failed to prove reject. Yes, it means that the researcher failed to prove his/her point, but it can be because of the following options:his/her point, but it can be because of the following options:
The Questionnaire questions were not well designed, formulated oThe Questionnaire questions were not well designed, formulated or understood from the r understood from the respondents, OR the Respondents were careless/notrespondents, OR the Respondents were careless/not--serious in their answering the serious in their answering the questions, OR the sample size (collected questionnaires) were noquestions, OR the sample size (collected questionnaires) were not enough,t enough,
Chapter 4: Data Analysis, Findings and Discussions
4.3 Discussion of Findings4.3 Discussion of Findings
Here you have to link/compile all results with the objective to Here you have to link/compile all results with the objective to answer the minor research questions. You have to refer to the answer the minor research questions. You have to refer to the literature in literature in chapter 2chapter 2 to support your findings. to support your findings.
Do not forget that this chapter is your own ground, i.e., you Do not forget that this chapter is your own ground, i.e., you should use all your should use all your languagelanguage and and opinionopinion in the interpretation in the interpretation of the results, of course with the of the results, of course with the supportsupport of the of the literature literature (it is (it is about about benchmarkingbenchmarking your results with respect to the results your results with respect to the results and the finding from the other and the finding from the other literaturesliteratures) and the results that ) and the results that you got from your analysis of the data.you got from your analysis of the data.
Chapter 5: Conclusion, Recommendations and Future Research
5.1 Conclusion 5.1 Conclusion Remind the reader with the problem and the objective, methodologRemind the reader with the problem and the objective, methodology, y,
models/theories, sampling etc. Then you summarize the most models/theories, sampling etc. Then you summarize the most important findings (from important findings (from chapterchapter 4, section 4.34, section 4.3).).
5.2 Recommendations 5.2 Recommendations (Based on the findings)(Based on the findings)•• Business, Industry, Customers, Government Business, Industry, Customers, Government …….etc. .etc. •• Make sure that all recommendations HAVE to come from the Make sure that all recommendations HAVE to come from the
research that has been done by youresearch that has been done by you•• Be close to the tactic level, i.e. be away from the macro levelBe close to the tactic level, i.e. be away from the macro level•• Be realistic in your recommendationsBe realistic in your recommendations
5.3 Future Work5.3 Future WorkSuggestions, new Ideas Suggestions, new Ideas …… as a results from the research experience.as a results from the research experience.Can come from the assumptions and limitations that you have put Can come from the assumptions and limitations that you have put in in
BiographyBiography (it is the last page in your thesis)(it is the last page in your thesis)
Here you write a brief biography about yourself, which includes Here you write a brief biography about yourself, which includes your birth date, your education path and your career path to your birth date, your education path and your career path to date. This takes around one paragraph (10date. This takes around one paragraph (10--12 lines)12 lines)
