4
Experimental Research Methods (Subject) / L11 (Lesson)
After this lecture and studying the materials, students are able to:
• Explain that high internal and external validity and power are needed to achieve the goals of
experiments
• Explain why the commonly used convenience samples within psychology are not ideal
• Explain how to control for confounding variables in a between-subjects and within-subjects design
• Motivate when an experiment should be a between-subjects and when a within-subjects design
This lesson was created by tklmn.
This lesson is not released for learning.
- Goal experiments Determine whether independent variable (IV) has an effect on the dependent variable (DV):
- Goal experiments Determine whether independent variable (IV) has an effect on the dependent variable (DV): 1. External validity → the degree to which the conclusions of a study can be generalized to otherpeople, geographical location, context, time, and duration of effect2. Internal validity → the degree to which the relationship between the IV and DV indeed (only)reflects the relationship between IV and DV3. Power → maximizing the probability of finding an effect between IV and DV, so making an effort to decrease MSw
- How to control for confounding variables in BS design? 1. Random assignment of subjects to conditions• When it is unclear what the confounding variables exactly are, randomization and a large n per cellresults in balancing of all these variables2. Systematically balance confounding variable• Balance groups with regard to confounding variable → groups are guaranteed to be balanced (in contrastto randomization)• Include confounding variable in the analysis → statistical control3. Matched group design• Balance groups with regard to confounding variable → one person in one group (exactly) matches oneother person in all other groups with regard to the score on the confounding variable• Stronger than balancing because exact match• Include confounding variable in the analysis → statistical control
- How to control for confounding variables in WS design? • Internal validity in a within-subjects design can be decreased by order effects• Practice effects → getting better at a task when progressing in the experiment• Fatigue effects → if a subject becomes tired• Carry-over effects → the experience obtained with the first condition affects the scores in the secondcondition• Response sets → developing a habitual response for subsequent tasks• Order effects can be controlled by balancing the order of conditions:• Partial balance → Latin square design• Randomly balance → each participant gets a random order of conditions• Full balance → all orders occur and subjects are randomly assigned to orders of conditions
Advertisement