QUANTITATIVE RESEARCH METHODOLOGIES

(Experimental Research)

EXPERIMENTAL RESEARCH

Best research methodology to establish cause-and-effect relatonships among variables

EXPERIMENTAL RESEARCH

Tests hypotheses about cause-and-effect relationships.

EXPERIMENTAL RESEARCH

Dependent VariableIndependent Variable

Experimental or Treatment Variable

Criterion or Outcome Variable

Essential Characteristics of Experimental Research

A. Comparison of groups

• 2 groups (experimental & control grps)

• experimental group receives treatment

• control group does not receive treatment

Essential Characteristics of Experimental Research

B. Manipulation of the independent variable

• The independent variable may be established as one form of the variable versus another

Example: A study comparing the learning by doing method with lecture method in TLE.

Essential Characteristics of Experimental Research

B. Manipulation of the independent variable

• The independent variable may be established as the presence versus absence of a particular form

Example: A study comparing the use of DLP versus no DLP in teaching history.

Essential Characteristics of Experimental Research

B. Manipulation of the independent variable

• The independent variable may be established as varying degrees of the same form

Example: A study comparing the effects of different contact hours on students interest in Mathematics.

Essential Characteristics of Experimental Research

C. Randomization

• Subjects are randomly assigned to groups.• Random assignment:

- individual participants have the chance of being assigned to any of the two groups

- eliminates extraneous variables

3 things to remember about randomization

Randomization

1) takes place before the experiment begins

2) is a process of assigning or distributing individuals to groups not a result of such distribution

3) forms groups that are equivalent and differ only by chance

GROUP DESIGNS IN EXPERIMENTAL RESEARCH

Weak Experimental Designs

The One-Shot Case Study

A single group is exposed to treatment and a dependent variable is observed (measured) to assess the effect of the treatment

Weak Experimental Designs

The One-Shot Case Study Design

X OTreatment Observation

(Dependent Variable)

Example: suppose you wish to see if a new textbook increasues student interest in your course (history, science, statistics, etc.)

Weakness: Absence if any control - researcher has no way of knowing if the results obtained at O are due to treatment X

Weak Experimental Designs

The One-Group Pretest-Posttest Design

A single group is measured or observed before and after the treatment.

Weak Experimental Designs

The One-Group Pretest-Posttest Design

O X OObservation

(Pretest)Treatment Observation

(Posttest)

Example: suppose you want to assess the effects of weekly counseling sessions on the attitudes of identified bullies in school

Weakness: Uncontrolled-for threats to internal validity that might explain the results of the posttest (Example: maturation, data collector characteristics, etc.)

Weak Experimental Designs

The Static-Group Comparison Design

Two existing intact (static) groups are used

Also called nonequivalent control group design

Weak Experimental Designs

The Static-Group Comparison Design

X O

O

Example: Suppose you want to test the effect of a new textbook in increasing interest of students in class. Find two intact groups, assign the new textbook to one class and have the other class use the regular textbook, then measure the degree of interest in both classes at the same time.

Weakness: Uncontrolled-for threats to internal validity (Vulnerable to maturity and location)

X - treatmentBlank space - control (with no treatment or with a different treatmen)O - observedDash - indicated that the two groups are not randomly assigned

Weak Experimental Designs

The Static-Group Pretest-Posttest Design

Two existing intact (static) groups are used

Pretest is given to both groups

Pretest score is subtracted from posttest score, permitting analysis of "gain or change"

Weak Experimental Designs

The Static-Group Pretest-Posttest Design

O X O

O OWeakness: amount of gain often depends on initial performance

True Experimental Designs

Subjects are randomly assigned to treatment and control groups.

True Experimental Designs

The Randomized Posttest-Only Control Group Design

Involves two groups formed by random assignment

One group receives treatment whike the other does not, then both groups are posttested

True Experimental Designs

The Randomized Posttest-Only Control Group DesignTreatment

Group R X OControl Group R C O

R - random assignment of individals to groupsX - treatment/exposure to treatmentC - control groupO - observation (measurement) of the dependent variable

Example

The Randomized Posttest-Only Control Group Design

100 randomly selected teachers

R50 teachers

XDance for

Peace Training-

Workshop

OPosttest:

Faculty Self-Awareness

Questionnaire

R50 teachers

CNo training

OPosttest:

Faculty Self-Awareness

Questionnaire

True Experimental Designs

The Randomized Pretest-Posttest Control Group Design

Involves two groups formed by random assignment

Both groups are pretested and posttested

Measurements or observations are collected at the same time for both groups

True Experimental Designs

The Randomized Pretest-Posttest Control Group DesignTreatment

Group R O X OControl Group R O C O

R - random assignment of individals to groupsX - treatment/exposure to treatmentC - control groupO - observation (measurement) of the dependent variable

Example

The Randomized Posttest-Only Control Group Design

100 randomly selected teachers

R50 teachers

OPretest:

Faculty Self-Awareness

Questionnaire

XDance for

Peace Training-

Workshop

OPosttest:

Faculty Self-Awareness

Questionnaire

R50 teachers

OPretest:

Faculty Self-Awareness

Questionnaire

CNo training

OPosttest:

Faculty Self-Awareness

Questionnaire

True Experimental Designs

The Randomized Solomon Four-Group Design

Involves random assignment of participants to four groups with two groups being pretested and two not

One of the pretested groups and one of the unpretested groups is exposed to treatment.

All four groups are posttested

True Experimental Designs

The Randomized Solomon Four-Group DesignTreatment

Group R O X OControl Group R O C O

TreatmentGroup R X O

Control Group R C O

The first two groups represent the pretest-posttest control group design, while the last two groups represent the posttest-only control group design.

ExampleThe Solomon Four- Group Design

100 randomly selected teachers

R25

teachers

OPretest:

Faculty Self-Awareness Questionnaire

XDance for

Peace Training- Workshop

OPosttest:

Faculty Self-Awareness Questionnaire

R25

teachers

OPretest:

Faculty Self-Awareness Questionnaire

CNo training

OPosttest:

Faculty Self-Awareness Questionnaire

R25

teachers

XDance for

Peace Training- Workshop

OPosttest:

Faculty Self-Awareness Questionnaire

R25

teachers

CNo training

OPosttest:

Faculty Self-Awareness Questionnaire

True Experimental Designs

Random Assignment with Matching

Pairs of individuals are matched on certain variables (choice of variables to match is based on previous research, theory and/or experience of researcher)

Members of each matched pair are assigned to the experimental and control groups at random.

True Experimental Designs

The Randomized Posttest-Only Control Group Design, Using Matched Subjects

TreatmentGroup Mr X O

Control Group Mr C O

Mr - Memhers of each matched pair are randomlyassigned to experimental and control groups

True Experimental Designs

The Randomized Pretest-Posttest Control Group Design, Using Matched Subjects

TreatmentGroup Mr O X O

Control Group Mr O C O

Quasi-Experimental Designs

No random assignment of participants or subjects

Quasi-Experimental Designs

The Matching-Only Design

Participants (subjects) from intact groups are matched.

Quasi-Experimental Designs

The Matching-Only Posttest-Only Control Group DesignTreatment

Group M X OControl Group M C O

M - Memhers have been matched but not randomly assigned

Quasi-Experimental Designs

The Matching-Only Pretest-Posttest Control Group DesignTreatment

Group M O X OControl Group M O C O

M - Memhers have been matched but not randomly assigned

Quasi-Experimental Designs

Counterbalanced Designs

Each group is exposed to ALL treatments in different order. Order is determined randomly.

Any number of treatments may be involved.

Example

Three-Treatment Counterbalanced DesignGroup I X1 O X2 O X3 OGroup II X2 O X3 O X1 OGroup III X3 O X1 O X2 O

Quasi-Experimental Designs

Time-Series Designs

Involve repeated observations or measuremetns over a period of time both before and after treatment.

01 O2 O3 O4 O5 X O6 O7 O8 O9 O10Basic Time-Series Design

Factorial Designs

with or without random assignment of participants or subjects

Extend the number of relationships that may be examined in an experimental study.

The interaction of an independent variable with one or more moderator variables can be studied.

Moderator variables may be either treatment variables or subject characteristic variables.

Example

Factorial Design (2 x 2 factorial design)

Treatment R O X Y1 O

Control R O C Y1 O

Treatment R O X Y2 O

Control R O C Y2 O

Y1 and Y2 - moderator variables with two levels

Example

Factorial Design (2 x 2 factorial design)

X CY1

Y2

Example

Factorial Design (2 x 2 factorial design)

X Learning by Doing

CLecture

Y1(Male)

Y2(Female)