I. Finding existing research
A. Start with a textbook summary
B. Consider PsycLit and ERIC
C. Look for research performed by students
II. Improving internal validity
A. Determining which variable came first (could a "cause" be an effect?)
1. Self-esteem & teen pregnancy
2. Country music and depression
3. Wearing black and being violent
B. Can you rule out other causes?
1. Women getting prettier at closing time (is it the time of day, alcohol consumption?)
2. Attendance and grades
C. Can you change a correlational study into an experiment?
II. Improving construct validity
A. Different manipulations
1. More closely related to theory's definition
2. Use of placebo in no-treatment condition
B. Different measures
1. More in-depth (e.g., interviews or behavioral observation rather than rating scale responses)
2. More objective measures
C. Double-blind study to reduce the effects of participant and researcher expectancies
III. External validity
A. Different group of participants
B. More realistic amounts of the treatment variable
C. Look for functional relationships
D. Long term effects
E. More realistic stimulus materials
F. Take lab study to the field
G. Look for exceptions to the rule by adding potential moderator variables
IV. Replicate studies
A. "I don't believe it!" The value of skepticism
B. "But other studies find contradictory results!"
1. Value of taking notes
2. Mood dependent learning
3. Women's spatial abilities changing during the week
4. Overjustification effectCameron, J. &Pierce, W. D. (1994). Reinforcement, reward, and intrinsic motivation: A meta-analysis. Review of Educational Research, 64, 363-423.
C. "But they didn't find anything": Don't accept the null hypothesis!
V. Extend studies
A. Try to pinpoint the exact effect of the variable
1. Rather than saying that intelligence declines with age, ask "On what types of tasks (and what types of processes) are there age-related declines?"
B. Look for other effects of the treatment
1. Cognitive or physiological effects of manipulations proven to have behavioral effects.
2. Behavioral effects of variables proven to have cognitive or physiological effects.
3. Long-term effects
C. See if related treatments have similar effects
1. Replicate temperature studies using humidity as a factor
D. Flip findings around
1. Could effects also be causes?
a. Does being violent cause you to wear black?
b. Does being nice cause you to be perceived as more attractive?
2. Could some characteristic or treatment that is associated with positive events also have some negative side effects (and vice-versa)?
a. Are there negative aspects to trying to achieve the American Dream?
b. Are there advantages to being perceived as unintelligent?