Chapter 10: Within-Subjects Design (PSYC 305)

attrition (mortality)

a threat to internal validity that occurs when participants drop out of a study before it’s complete

baseline measurement

behavioral diary

a self-report data collection strategy where individuals record their behaviors and associated feelings as they occur

carryover effect

counterbalancing

dependent means t-test

fatigue effect

history

a threat to internal validity that occurs when an outside event, unrelated to the independent variable, occurs during the study and influences the dependent variable

instrumentation (instrumentation problems)

a threat to internal validity that occurs when the measurement instrument or method of measuring the dependent variable changes over the course of a study

latin square design

maturation

mortality

order effect

paired-samples t-test

practice effect

pretest-posttest design

a within-subjects design where participants are measured on a dependent variable before and after exposure to a treatment or intervention

repeated-measures analysis of variance (repeated-measures ANOVA)

repeated-measures design

research participant pool

sensitization effect

testing effect

a threat to internal validity where repeated exposure to the same measurement makes participants more likely to figure out the study’s purpose, which can change their responses

t-test for dependent means

within-subjects design

a research design in which the same participants are measured or exposed to more than one level of the independent variable

• Participants serve as their own control group

• Reduces random error from individual differences between groups

  • whatever unique characteristics participants bring stay constant across the study because we’re measuring the SAME people

• Require fewer total participants

  • because each person contributes data at multiple points

• Saves time and effort

  • 
    

What are the benefits of a within-subjects design?

retrospective bias

• don’t always remember our past behaviors and feelings accurately when recording them after the fact

What is a pitfall of behavioral diaries?

keep an ongoing journal in the moment instead of recalling from memory later

What is a solution to retrospective bias when using behavioral diaries?

doesn’t have an empty control group

• can’t rule out alternative explanations for any change we see

What is a limitation for a pretest-posttest design?

longitudinal design

a within-subjects design that tracks the same participants’ development or behavior over an exended period of time, often years

• especially vulnerable to attrition

  • participants may drop out, move away, or lose contact with researchers over long time spans

What is a drawback with longitudinal designs?

repeated-measures design

a within-subjects design where participants are exposed to each level of the independent variable and measured on the dependent variable after each exposure

• repeated-measures design

  • participants experience multiple conditions in some sequence

Which design type is most closely tied to order effects?

• potential external validity concerns

  • effects observed could be the cumulative result of multiple exposures rather than a single exposure to the independent variable

• Not realistic

  • uncommon in everyday life to experience multiple levels of the same variable in close succession

• More vulnerabe to order effects

  • sequence of conditions can itself influence the dependent variable

Wha are the disadvantages of within-subjects designs?

• If participants drop out because the program wasn’t working:

  • the remaining sample looks better than reality, and the program looks more effective than it is

• If participants drop out because they felt they no longer need the program:

  • we may underestimate its effectiveness for people who would have stuck with it

Why is attrition a threat to internal validity?

• distractor items

• varying the order of scale items each time

What are some potential solutions to prevent the testing effect?

• keep measurement instrument and method consistent across every point of data collection

What is a solution to instrumentation problems?