Psych 225 Lecture 16 Outline of Last Lecture 1) Schmitt Article 2) Bray and Sugarman ArticleOutline of Current Lecture 1) Green Article2) Internal Validitya. Type 1 Errorsb. Type 2 ErrorsCurrent LectureSlide: Green-role of initial atts in separating CD (cognitive dissonance) and SP (self perception)-CD: more extreme the initial attitude, more need for attitude change if behavior is inconsistent-SP: initial attitudes are unimportant, just look at behavior and circumstances under which it occurs Slide: Green-Design2 (extremity of initial attitude: initial thirst) x2 (incentive for commitment to water deprivation)Between Slide: GreenThese notes represent a detailed interpretation of the professor’s lecture. GradeBuddy is best used as a supplement to your own notes, not as a substitute.-Cover story:“How biological drives affect psych variables…relation between thirst and perceptual judgments”“Important that everyone experiences the same degree of thirst”Slide: GreenExtremity of initial attitude (thirst) -low: saltine and ½ tsp peanut butter-high: saltine & 1 tsp hot sauce~after 2 minutes, 1st thirst rating Slide: Green-P’s made judgments of famous paintings (“perceptual task”)-Experiment asked if P’s would participate in series of experiments on drives Slide: Green-need to avoid liquids for 24 hours…wear litmus bandage that will detect water-incentive: offered $5 or $20 to engage in further thirst deprivation -virtually everyone signed commitment to engage in 24 hour liquid deprivation Slide: Green-2nd thirst rating (before they could take a drink) when counter attitudinal behavior is strongest (most dissonance) -predictions re: attitude change?Slide: Green-Festinger-Festinger’s Cognitive Dissonance Attitude -greatest attitude change should occur when there was high initial thirst with incentive of only $5 greatest amount of conflict so should be greatest attempt to resolve the dissonance Slide: Green-Bem-low incentive: infer attitude from behavior (change)-high incentive: do not infer attitude from behavior (no change) Slide: Green IncentiveInitial Thirst $5 $20Low No change No changeHigh Large decrease in thirst Small decrease in thirstNot good results for BemChapter 7: Testing the HypothesisThreats to Internal validity -internal validity: that IV is what has an effect on DVSlide: Type 1 and Type II Errors-Chart on CP pg. 131 H null is true H null is false Decision: Do not reject H null Decide (based on data) IV does not affect DV and IV does not affect DV (1-a) Decide IV does not affect DV and IV affects DVB Type IIDecision: Reject H null Decide IV affects DV and IV does not affect DVa Type IDecide IV affects DV and IV affects DV1-B (power) Why Type II Errors Occur (B)-few degrees of freedom-to few levels of IV -restricted range-too much variability Slide: Sample of Cook & Campbell, Campbell & Stanley’s threats to internal validity -can increase the likelihood of committing a type 1 or type II errorMaturation: changes in scores due to changes in Ps over the course of the studyP’s add to encode words in one of two ways EncodingRote ContextH: Higher Recall Very Long Lists8.50a 9.00aIf too many words and everyone does really badly, then we might miss the relation between variablesTesting: changes in scores during measurement -tries to respond how they think experimenter wants them to respond Pre-test of attitudes toward outgroup 5.00a 5.25a Partner in “get acquainted” sessionIngroup Outgroup Post-test of attitudes toward outgroup 8.25b 7.75One thing to make it more likely to get a difference in the post-test Slide: Threats to Internal Validity Statistical Regression: changes in scores due to regression to the mean Pre-test GRE The Princeton Review Post-test GRE%ile rank=25 %tile rank=35-took mini course-tend to shift toward overall average or mean score -shift from extreme toward normal Selection-non equivalent groups prior to introduction of IV-confounding variable due to selection of PpAloud (week 1-8) Silently (week 9-16)11.0b 6.00aH: expect those who read aloud (poem) will remember more because there are rhythms, sounds, beats, etc. -worked..BUT: could have committed a Type 1 error Might have different people in first group (motivated early on) than those who are in the secondgroup; minds might be fresher earlier on in the semester Aloud (week 9-16) Silently (week 1-8)7.00a 6.50a* would be a type 2 error if we committed one was an effect but said there wasn’t Slide: highly unlikely potential selection threats in IE and VP?-have to do with kinds of people that experience different levels of the variables IE: did criminal justice experience covary with judge’s instruction re: critical evidence?Random assignment (used to minimize chance of getting a selection threat) VP: did English fluency covary with relation between paired words?Random assignment (could not covary with valence because that was a within subjects variable) Slide: Threats to internal validity Mortality: differential attrition (drop out)lack of equivalence of groups beyond IVEx. Does density of rearing environment affect learning? Rearing EnvironmentSpacious Crowded9/10 survived 1 yr 5/10 survived 1 yr 50 s to complete maze 30 s to complete maze (learned more quickly) *could have been survival of the fittest: dumber mice die Similar findings for human study on depression for those on a waiting list and those who are active; those on waiting list dropped out more and thus had same level of depression, those who were in active therapy group did see improvement *in waiting list group: those lower on depression scale dropped out after waiting weeks, whereas the ones who stayed in this group were more depressed (could influence our supposedresults) (possible Type 1 error)Diffusion or imitation of treatments (specific to Type II errors) -dampens effect of IVincrease in power (Type II error) Encoding TaskSelf (referential) Structural (form-how it appears)9.75a 9.25a-if effect there, we failed to detect it: Type II error-with between subjects design: could be chatter amongst P’s talked about the encoding methodthey
View Full Document