Psychology 3800F/G Study Guide - Final Guide: Family-Wise Error Rate, Type I And Type Ii Errors, Null Hypothesis
31 Dec 2017
School
Department
Course
Professor
Document Summary
When performing t-test on two independent groups w/ a=0. 05, chance of concluding that there is difference in means b/w two groups when difference is really due to chance is. False negative; represented by b, which occurs when investigator incorrectly fails to reject null hypothesis (ex: b of . 20 indicates that likelihood of concluding that there is no difference in means b/w two groups when one really exists is. Larger sample sizes increases power, large effect sizes increase power, smaller sample variance increases power, increasing significance level (a > . 05) will increase power. To determine if null result in thesis was likely due to an underpowered sample, to estimate # of participants you will need to un based upon predicted effect size. Chapter 6 repeated measures: experimental design: men and women in three sections of 3800 each complete 10 assignments throughout year, and one final exam. Orthogonal and oblique rotations will yield approx. same results.
