Chapter 9
Experimental Designs
© 2009 John Wiley & Sons Ltd.
www.wileyeurope.com/college/sekaran
1
Causal Research
 Research conducted to identify cause-and-effect
relationships among variables when the research has
already been narrowly defined
© 2009 John Wiley & Sons Ltd.
www.wileyeurope.com/college/sekaran
2
Evidence for Causality

Covariation

Time order of occurrence of variable

Elimination of other possible causal factors

A logical explanation
–
Evidence of the extent to which X and Y occur together or
vary together in the way predicted by the hypothesis
–
Evidence that shows X occurs before Y
–
Evidence that allows the elimination of factors other than X
as the cause of Y
–
About why the independent variable affects the dependent
variable.
© 2009 John Wiley & Sons Ltd.
www.wileyeurope.com/college/sekaran
3
Experiment

Data collection method in which one or more IVs are
manipulated in order to measure their effect on a DV,
while controlling for exogenous variables in order to
test a hypothesis

Cause and effect relationship is established by
–
–
Manipulation of independent variable
Controlling for exogenous factors
© 2009 John Wiley & Sons Ltd.
www.wileyeurope.com/college/sekaran
4
Manipulation of IV

Manipulation
Creation of different levels of the IV to assess the impact on
the DV

Treatment levels
The arbitrary or natural groups a researcher makes within the
IV

Evidence for causality
–
–
Covariation (difference between groups)
Time order control
© 2009 John Wiley & Sons Ltd.
www.wileyeurope.com/college/sekaran
5
Exogenous Variables

Controlling for exogenous/confounding variables
–
–
Eliminating other possible causal factors
Eliminating alternative explanations

Experimental designs available

Two types of exogenous variables
–
–
Related to participants
Related other, environmental factors
© 2009 John Wiley & Sons Ltd.
www.wileyeurope.com/college/sekaran
6
Related to Participants
 Selection bias: improper assignment of participants to the experimental groups
– Matched groups: Match the different groups as closely as possible in terms of age,
interest, expertise etc.
– Random assignment: Randomly assign members to different treatment groups. The
differences will be randomly distributed. Systematic bias will reduce.
–Statistical control: Measuring the external variables and adjusting for their effect
through statistical methods
 Mortality: Loss of participants during the experiment
© 2009 John Wiley & Sons Ltd.
www.wileyeurope.com/college/sekaran
7
Related to other actors
 History effects: External events occurring at the same time that may
affect the DV
 Maturation effects: Changes in the participants as a passage of time
that may affect the DV
 Testing effects: The experiment itself affect the responses
– Main testing effect: prior responses affect later responses
– Interactive testing effect: prior responses affect perception of IV
 Instrumentation effects: Changes in measuring instrument
© 2009 John Wiley & Sons Ltd.
www.wileyeurope.com/college/sekaran
8
Experimental Design
© 2009 John Wiley & Sons Ltd.
www.wileyeurope.com/college/sekaran
9
Experimental Design
© 2009 John Wiley & Sons Ltd.
www.wileyeurope.com/college/sekaran
10
Experimental Design
© 2009 John Wiley & Sons Ltd.
www.wileyeurope.com/college/sekaran
11
Experimental Design
© 2009 John Wiley & Sons Ltd.
www.wileyeurope.com/college/sekaran
12
Experimental Design
© 2009 John Wiley & Sons Ltd.
www.wileyeurope.com/college/sekaran
13
Experimental Design
© 2009 John Wiley & Sons Ltd.
www.wileyeurope.com/college/sekaran
14
Experimental Design
© 2009 John Wiley & Sons Ltd.
www.wileyeurope.com/college/sekaran
15
Exercise
An organization would like to introduce one of two types of new manufacturing processes
to increase the productivity of workers. Both involve heavy investment in expensive
technology. The company wants to test the efficacy of each process in one of its small
plants.
Propose an experiment, using:
- Pretest posttest control group design
- Posttest control group design
And calculate for each design the specific effect of each new process on the productivity.
© 2009 John Wiley & Sons Ltd.
www.wileyeurope.com/college/sekaran
16
Validity
 Internal validity
– Determination of whether the effect is actually caused by the manipulation
of treatments and not by other, exogenous variables
 External validity
– Determination of whether the cause-and-effect relationships found in the
experiment can be generalized
© 2009 John Wiley & Sons Ltd.
www.wileyeurope.com/college/sekaran
17
Compared to field experiments,
lab experiments have the following
 Advantages
– High degree of control
– High internal validity / replication
– Less costly and less expensive
 Disadvantages
– Artificiality => reactive error
– Demand artifacts
– Lower external validity
© 2009 John Wiley & Sons Ltd.
www.wileyeurope.com/college/sekaran
18