Within Subjects and Matched-subjects Approaches

- most often if lines are not parallel there is an interaction
-Significant interactions are more important than main effects

Comparing participants with themselves
Error variance is reduced
More power to detect differences

Might have carryover effect from on experience to the other
To control for this use intrasubject counterbalancing,

All groups are equal on every factor at the beginning
Its more sensitive to changes in the treatment effect
Total number of participants can be smaller

When the treatment has a lasting effect or the purpose of the study is to test for lasting effects
When the experiment is sensitive to time-related effects
- order effects are brought about through continued repitision of the tasks
- fatigue effects: a decline in performance
- practice effects: an improvement in performance

Each condition has to occur equally as often
And in a different order where each precedes every other one equal number of times

Every possible sequence appears at each presentation of the treatment

Each condition occurs equally as often
but does not precede and follow the other conditions an equal number of times

this is used when complete counterbalancing is difficult

Differential order effect
Presence of treatment-by-order effect
The treatment effects depend on the order of presentation

To study psychological process over time
Used lots in developmental research
Used for studying learning processes and determining confound of practice effects

Include both within and between subject components
Ex. if we want to know if peer victimization changes for girls and boys across kids
- Between factor is sex
- Wtihin factor is age, collect data over 5 years

Have advantages of both within-subjects and the random assignment of between subjects

Pairs are matched on some measure that is highly correlated with the dependent variable
One member is randomly assigned to either group

Randomized block design is the resulting procedure when the matching factor is analyzed
Advantages are:
ensures groups in study are equal
reduces the within groups variance

When matching takes place on a nominal scale factor
ex sex