Experiments
Why would a double-blind experiment be used?
Discuss how the following factors may threaten internal
validity: history, maturation, testing, instrumentation,
statistical regression, selection biases, experimental
mortality, causal time-order, diffusion or imitation of
treatments, compensation, compensatory rivalry, and
demoralization.
Does a true experimental design address the problem of
external validity?
Identify strengths and weaknesses of the experimental
method.
Experimental Research
Experiments
Begin with a Hypothesis
Modify Something in a Situation
Compare Outcomes
Cases or People are Termed
“Subjects”
Random Assignment
Probability of Equal Selection
Allows Accurate Prediction
An Alternative to Random
Assignment is Matching
Parts of the Classic
Experiment
Treatment or Independent Variable
Dependent Variable
Pretest
Posttest
Experimental Group
Control Group
Random Assignment
Variations on Experimental
Design
Pre-experimental Design
One-shot Case Study
One-group Pretest-Posttest Design
Static Group Comparison
Quasi-Experimental and Special Designs
Types of Validity
External Validity
Do the results apply to the broader
population?
Internal Validity
Is the independent variable responsible for
the observed changes in the dependent
variable?
Confounding Variables That
Threaten Internal Validity
Maturation
Changes due to normal growth or
predictable changes
History
Changes due to an event that occurs
during the study, which might have
affects the results
Confounding Variables That
Threaten Internal Validity
Instrumentation
Any change in the calibration of the measuring
instrument over the course of the study
Regression to the Mean
Tendency for participants selected because of
extreme scores to be less extreme on a retest
Selection
Any factor that creates groups that are not equal at
the start of the study
Confounding Variables That
Threaten Internal Validity
Attrition
Loss of participants during a study; are
the participants who drop out different
from those who continue?
Diffusion of treatment
Changes in participants” behavior in one
condition because of information they
obtained about the procedures in other
conditions
Subject Effects
Participants are not passive
They try to understand the study to help them
to know what they “should do”
This behavior termed “subject effects”
Participants respond to subtle cues about what
is expected (termed “demand characteristics”)
Placebo effect: treatment effect that is
due to expectations that the treatment
will work
Experimenter Effects
Any preconceived idea of the
researcher about how the
experiment should turn out
Compensatory effects
Types of Control
Procedures
General control procedures (applicable to
virtually all research)
Control over subject and experimenter
effects
Control through the selection and
assignment of participants
Control through specific experimental design
Principles of Experimental
Design
Control the effects of lurking
variables on the response, most
simply by comparing two or more
treatments
Randomize
Replicate
Randomization
The use of chance to divide
experimental units into groups is called
randomization.
Comparison of effects of several
treatments is valid only when all
treatments are applied to similar groups
of experimental units.
How to randomize?
Flip a coin or draw numbers out of a hat
Use a random number table
Use a statistical software package or
program
Minitab
www.whfreeman.com/ips
Statistical Significance
An observed effect so large that it
would rarely occur by chance is
called statistically significant.
A few more things…
Double-blind: neither the subjects nor
the person administering the treatment
knew which treatment any subject had
received
Lack of realism is a major weakness of
experiments. Is it possible to duplicate
the conditions that we want?
