Research Methods II: Repeated Measures Designs

Within Subjects Design

Testing same group in which the group experiences all conditions including the control group.

-Lower number of subjects meaning easier to test on vulnerable populations and rare conditions
-More efficient
-Higher sensitivity

-Ability to detect effect
-Ideally: participants respond similarly to manipulation
-Realistically: Error Variance (variance in both subject and procedure)

Increases, Decreases

-Could be an outlier in the group
-Goal is to obtain generalizable results

-Repeated Measures has a control condition that surveys do not
-Longitudinal surveys desire to see change over time

Practice Effect: the more times a subject is told to take the measures results may change

Interference Effect: Cognitive fatigue / boredom

Carry-Over Effect: Analysis of first stimuli may effect how one analyzes the second or following tests

Order Effect: what is presented first may change the results of the next (Dating and Quality of life study)

An increased likelihood participants may catch on to the hypothesis

Vary the order in which participants complete each condition

Complete: All participants encounter all the conditions multiple times

Incomplete: Participants experience each condition once

Block Randomization: used to order conditions for participants where size of block is equal to number of conditions. Reduces practice effect

Present conditions in one sequence (A->B) then present in opposite sequence (B->A) in order to see if practice effect is linear. MAY CREATE ANTICIPATION EFFECT

Selected Orders: use SOME of the practical orders

Latin Square:Each condition appears at ordinal position only once (constructed by taking # of conditions and using number as rows and columns)

2 conditions: use paired sample t test

2+ conditions: One way repeated measures ANOVA