Chapter 8: Experiments
Experiments probe causal relationships. Experiment is a mode of observation, in which
researchers take action and examine the consequences of such action.
What topics are appropriate to experiments?
Experiments are well suited to research projects involving relatively limited and well defined
concepts and propositions. Experiment models are appropriate for hypothesis testing and
explanatory purpose. Example 1: learning about the contribution of aboriginal Canadians to
Canadian history will reduce prejudice. Example 2: study of small group interaction
Define “Natural experiment”.
Experiments that occur in the regular course of social events.
What are the three components of a classical experiment?
1) Independent & dependent variables:
an experiment examines the effect of an independent variable on a dependent
variable. Typically, the independent variable takes the form of an experimental
stimulus. The experimenter compares what happens when the stimulus is present
to what happens when it is not.
2) Pretesting and posttesting:
Pretesting occurs first, whereby subjects are measured in terms of a dependent
variable. Then the subject is exposed to a stimulus representing an independent
variable. Finally in posttesting, they are remeasured in terms of dependent
variable. Any differences between the first and last measurements on the
dependent variable are then attributed to the independent variable.
3) Experimental and control groups:
Experiments involve both an experimental group that administers a stimulus and a
control group that does not receive the experimental stimulus.
Define “dichotomous variable”.
A variable that has only two attributes. It is also called a binominal variable. The stimulus is a
dichotomous variable, having two attributes, present or not present.
The measurement of a dependent variable among subjects before they are exposed to a stimulus
representing an independent variable.
The remeasurement of a dependent variable among subjects after they have been exposed to a
stimulus representing an independent variable.
Define “control group”. In experimentation, a group of subjects to whom no experimental stimulus is administered and
who should resemble the experimental group in all other respects. The comparison of the control
group and the experimental group at the end of the experiment points to the effect of the
Define “Hawthrone effect”.
A term coined in reference to a series of productivity studies at the Hawthorne plant of the
Western Electric Company in Chicago, Illinois. The researchers discovered that their presence
affected the behaviour of the workers being studied. The term now refers to any impact of
research on the subject of study.
Define “double blind experiment”.
An experimental design in which neither the subjects nor the experiments know which is the
experimental group and which is the control.
What are 3 ways to make the experimental and control group similar?
1) Probability sampling: Beginning with a sampling frame composed of all the people in
the population under study, the researcher might select two probability samples. If these
samples each resemble the total population from which they’re selected, they’ll also
resemble each other.
2) Randomization: The procedure of randomly assigning experimental subjects to
experimental and control groups. Not to be confused with probability sampling
techniques such as simple random sample.
3) Matching : In connection with experiments, the procedure whereby pairs of subjects are
matched on the basis of their similarities on one or more variables; then one member of
the pair is assigned to the experimental group and the other to the control group.
What are the 2 reasons why randomization is better than matching?
1) You may not be in a position to know in advance which variables are relevant for the
2) Most of the statistics used to analyze the results of experiments assume randomization.
Failure to design your experiment that way, make your later use of those statistics less
What are the 4 “pre-experimental” designs that Campbell and Stanley discussed?
1) One shot case study: a single group of subjects is measured on a dependent variable
following the administration of some experimental stimulus.
2) One group pretest-posttest design: notes the possibility that some factor other than the
independent variable might cause a change between the pretest and posttest results.
3) Static group comparison: some research is based on experimental and control groups
but has no pretests.
Posttest only control group: It consists of the second half (Group 3 & 4) o