PSYC 2002 Lecture Notes - Lecture 22: Linear Regression, Multiple Correlation, Multivariate Analysis
Document Summary
If s or s information for each group is provided, ss values for anova can be calculated; this strategy is more efficient than calculating ss values from raw scores. For ss-between you need to calculate the grand mean, do not take the average of the means: *exam question about strategy. Ss values can be incorporates into the calculations for ss-within. (cid:2870)= (cid:3041) (cid:2869: =(cid:4666) (cid:883)(cid:4667)(cid:2870, (cid:1875)=(cid:1875)(cid:883)+(cid:1875)(cid:884)+(cid:1875)(cid:885) , =(cid:4666)(cid:883)(cid:4667)(cid:883)+(cid:4666)(cid:884)(cid:4667)(cid:884)+(cid:4666)(cid:885)(cid:4667)(cid:885) , = (cid:3042) (cid:3041, (cid:1854)= (cid:883)(cid:4666) (cid:4667)(cid:2870)+ (cid:884)(cid:4666) (cid:4667)(cid:2870)+ (cid:885)(cid:4666) (cid:4667)(cid:2870) , sum all the n"s across the groups, this provides the total n. Multiple correlation: overall degree of association between a dependent variable and predictor variables, multiple correlation coefficient = r. Multiple regression and correlation: you can determine the stat significance in multiple regression and correlation. If the overall r is not significant, the individual betas aren"t evaluated. Sequential examination of the effect of predictor variables on the dependent variable without direction from theory.
