Chapter 9 Between Subjects Design Notes 1
Chapter 9 Between Subjects Design Notes 1
Chapter 9 Between Subjects Design Notes 1
the general structure of an experiment (but not its specific probability of getting a true treatment effect
content) probability of rejecting the null hypothesis if your research
hypothesis (alternative hypothesis) is true
Three aspects of the experiment play the biggest part in determining
the design: Power is affected by sample size.
1.Number of Independent Variable Larger sample size= Higher Power
2.Number of Treatment Conditions You can use software or power charts in determining sample size.
3.Using the same or different subjects in each treatment condition
Between - Subjects Design
a design in which different subjects take part in each condition
of the experiment each subject is tested in only one condition
Selecting and Recruiting Subjects
"The more the sample resembles the whole population, the more likely
it is that the behavior of the sample mirrors that of the population."
Random selection is the ideal way to recruit subjects, but it is
rarely achieved Two-Group Design
Some Methods to Increase Subject Participation the simplest experimental design, used when only two
treatment conditions are needed
Make your appeal interesting, nonthreatening, and meaningful.
Pay them if you can or give token gifts. Two-Independent Groups Design
How Many Subjects? an experimental design in which subjects are placed in each of
two different treatment conditions through random assignment
There is no simple answer to this question.
Consider the effect size—a statistical estimate of the size or Random Assignment
magnitude of the treatment effect
every subject has an equal chance of being placed in any of the
Large - 10 to 20 subjects per group treatment conditions
Moderate - 20 to 30 subjects per group
Small - larger sample size
Experimental group- control group design
Experimental condition – apply a value of our independent variable Two-matched-groups Design
to the subjects and measure the dependent variable. Also called an
Randomization does not guarantee that treatment groups will
experimental group
be comparable on all the relevant extraneous subject variables.
Therefore, Researchers sometimes use another two group
Control condition – determine the value of the dependent variable
procedure the two-matched-groups design.
without an experimental manipulation of the independent variable.
Also called a control group. We simply measure subjects’
responses without trying to alter them in any way.
Matching before and after an Experiment
To form matched groups, subjects must be measured on the
Two-experimental- groups design extraneous variables that will be used for the matching.
Used to look at behavioral differences that occur when subjects are
exposed to two different values or levels of the IV.
Precision Matching
Example: We could test whether a highly violent music video
produces aggressiveness than a music video with a low level of In which we insist that the members of the matched pairs have
violence. identical scores.
Social facilitation- The presence of an audience, the performance In which we require that the members of a pair fall within a
of some behaviors improves. previously specified range of scores.
Simple randomization – a technique that maintains complete
randomness of the assignment of a subject to a particular group
such as flipping a coin, rolling a dice or drawing cards from a deck. Rank Ordered Matching
The subjects are simply rank ordered by their scores on the
matching variable, and subjects with adjacent scores then
When to use two Independent matched groups become a matched pair.
We begin by looking at the hypothesis.
If there is only one independent variable, the two Independent
groups approach is a good choice if the hypothesis can tested When to use two matched groups
with two treatment condition.
By matching on a variable that is likely to have a strong effect variable before conducting the actual
on the dependent variable , we can eliminate one possible experiment.
source of confounding .
Multiple Groups
Multiple group design- a design in which there are more than
two groups of subjects and each group is run through a
different treatment condition.
Multiple-independent-groups design- is the commonly used
multiple-groups design in which the subjects are assigned to
the different treatments condition at random.
Assigning Subjects
Block Randomization- a method for assigning subjects to
treatment conditions in which individuals are divided to subsets
or blocks and then some random process is used to place
individuals from those blocks into the different conditions.
Psychologist use block randomization as a sophisticated
procedure to obtain equal numbers of subjects in all treatment
conditions.
Choosing Treatments:
An experimenter will select treatment conditions
that are proportional.
Always think in terms you are testing.
General Rule: Select the simplest design that
will make an adequate test of hypothesis.
Practical Limits:
As you set experiments, you will make
decisions about which comparisons will provide
the most appropriate test of the hypothesis.
Some researchers conduct a pilot study to
pretest selected levels of an independent