PSYCH291 Study Guide - Final Guide: Center For Open Science, False Positive Rate, Hindsight Bias
Document Summary
Incentives to publish requires p < 0. 05: publication needed for profs to receive grants. Incentives of fame/fortune: dangers of p-hacking, sharpshooter fallacy: bulls eye around data that supports hypothesis, p-curve: p-values range of p=. 02 to . 05 are very unlikely. Possible confounds (popularity of the movie: results, between relational aggression and no aggression: d = Any video game could cause aggression: can be design confounds in a control condition. Individual differences (differences in tendency to view behaviour as intentional, differences in familiarity with football: random measurement error (eg. the instruction were vague or the measure was imprecise, situation noise (distracted during study, solutions: Increase sample size = more precise estimates of group means: use within-subject design variability is not an issue. P = . 22, d=. 38: d is cohen"s d, measures effect size strength, difference of means, within-subject design, participants in both groups, benefits, less participants needed, don"t have to worry about imbalances.