Transcript Chapter 10: Experimental Research Design

Chapter 10: Experimental
Research Design
This multimedia product and its contents are protected under copyright law. The following are prohibited by law:

Any public performance or display, including transmission of any image over a network;

Preparation of any derivative work, including the extraction, in whole or in part, of any images;

Any rental, lease, or lending of the program
Introduction


Research Design—the outline, plan, or
strategy used to answer the research
question
Purpose of research design
– Control for unwanted variation
– Suggests how data will be statistically
analyzed
Designs with Threats to
Internal Validity

One-Group Posttest-Only Design
Treatment Response
X
Y
– Rarely useful because no pretest or control
group
– Almost all threats to internal validity apply
– Is useful only when specific background
information exists on the DV

One-Group Pretest-Posttest Design
Pretest
treatment
Posttest
Y
X
Y
– Most threats to internal validity exist
– To infer causality must identify and
demonstrate that internal validity threats do not
exist

Nonequivalent Posttest-Only Design
Treatment
Posttest
Exper. Gp.
X
Y
Control Gp.
Y
– No assurance of equality of groups because
they were not randomly assigned
– May confound selection with treatment effect
Requirements of Experimental
Research Designs


Answer the research question or test the
hypothesis
Control for the effect of extraneous variables
– With control techniques (Randomization)
– With a control group—group that does not get the
IV or gets some standard value



Serves as source of comparison to experimental group
Controls for rival hypothesis
Should allow generalizability of the results
Pretesting Participants





To increase sensitivity through matching
To test for a ceiling effect
To test for initial position
To insure initial comparability
To obtain evidence of change
Experimental Research
Designs




Posttest-Only Design
Treatment Response
Exper. Gp
X
Y
Control Gp
Y
– Represents basic components of most designs
Types of Posttest-Only Designs

Between-participants
– Can also match participants prior to random
assignment to insure equivalence on the
matched variable

Within-participants
Types of Between-Participants
Posttest-Only Designs

Simple-Randomized
– Used with more than 2 levels of one 1 IV
– Participants randomly assigned to group

Factorial
– Used to test
The effect of more than one IV
 The interaction of several IV’s

– Participants randomly assigned to groups
Simple-Randomized
Posttest-Only Design
Treatment Response





Group I
Group II
Group III
Group IV
X1
X2
X3
Y
Y
Y
Y
Factorial Design





Independent Variable
A1 A2 A3
Independent B1
Variable B
B2
B3
Components of a Factorial
Design



Cell—a treatment combination of two or
more IV’s
Main effect—the effect of one IV
Interaction effect—when the effect of one
IV depends on the level of another IV
Classic Interaction Effect
Treatment
A1
A2






B1
B2
Column
Mean
15
30
45
30
30
30
Row
A3 Mean
30
45
15
30
30

Difficulties with factorial designs
– Increases the number of research participants
– Difficulty in interpreting higher-order interactions.

Advantages of factorial designs
– Can manipulate more than one IV
– Can control potential extraneous variable by building
it into the design
– Provides greater precision when you add more than
one IV
– Can test the effect of interactions
Within-Participants
Posttest-Only Design






Treatment
A1 A2 A3
Same 20
P1 P1 P1
Participants in
. . .
Each Treatment
. . .
Condition
P20 P20 P20

Advantages of within-participants design
– Equivalence of research participants
– Requires fewer participants than betweenparticipants design

Disadvantage of within-participants design
– Sequencing effects
Pretest-Posttest Design
Primary unique characteristic is that it includes a
pretest in addition to a posttest
Pre- test Treatment

Exper. Gp
Control Gp
Y
Y
Post-test
X
Y
Y