Statistical Sciences 2035 Study Guide - Quiz Guide: Analysis Of Variance, Covariate, Explained Variation
16 Oct 2018
School
Department
Professor
Document Summary
Check whether another predictor can increase the explained variation. As seen on page 40, the model summary shows that 60% of the variation in sbd scores can be explained by age differences (r square) Can r square be increased by adding a 2nd independent variable (the one with the 2nd highest correlation with y check initial correlations table) Here: include bmi/quetelet as the 2nd predictor we then have a multiple regression model, because we have multiple predictors/independent variables. Expectations of adding another predictor depends on whether: the existing predictor does not correlate with the predictor which is to be added: We expect the non-standardized regression coefficients to not change, as there was no confounding/overlap in the first place. Value of additionally explained variation square the correlation of x2 on y (seen in correlations table) = value that is added to r square in the model summary: e. g.
