research methods advanced- week 5= hierarchal regression

hierarchal regression

hierarchal regression overview

seperates the predictors into step 1 and step 2, enables us to examine the contributions of additional variables to the models.

dummy coding (andy field)

count number of groups to recode and subtract 1. create new variables dummy values. choose one og your groups as a baseline. assign the baseline group values of 0. include all dummy variables in predictors in regression analysis

spss: durbin-watson

test statistic between 0 and 4, value of 2 means the residuals are uncorrelated, a value greater than 2 indicates a positive correlation , a value lower than 2 indicatess a negative correlation. values tend to suggest no issue

spss: model fit

gives us an indication of the proportion of variance account for in our outcome variable by our predictors.

spss: r2

the change statistics tell us whether adding another variable into the analysis at step 2 imporves the model by accounting the proportion of variance accounted for . if the change is significant , the f statistic is 0.05 or less this tells us that the increase in variance is accounted for is statistically significant

SUMMARY OF THE MODEL SUMMARY TABLE DURBIN WATSON

Durbin watson- assumption of independent errors if its near 1 assumption is met