1.You Are now trying to find good predictors for who uses the computer after the Training (variable: compuse2). You heard the time 2 variables were more correlated with your dependent variable, so focus on them as predictors (companx2, compcon2, complik2).
a.First, What is the correlation between each of those predictors and your criterion (i.E., dependent variable)?
The compax2 and compuse2 were significantly Correlated, r(38) = .76, p < .001.
The compcon2 and compuse2 were significantly Correlated, r(38) = .73, p < .001.
The complik2 and compuse2 were significantly Correlated, r(38) = .41, p = .008.
a.How about the correlation just between the Predictors? Is that hinting at multicollinearity?
The companx2 and compcon2 were significantly Correlated, r(38) = .46, p = .003.
The compcon2 and complik2 were significantly Correlated, r(38) = .38, p = .016.
The companx2 and complik2 were significantly Correlated, r(38) = .42, p = .006.
Because none of the correlations Were higher than .8, this does not hint at multicollinearity across the Independent variables.
a.When You input all the time 2 variables simultaneously In your model, is that a good model? Make sure you mention the multicollinearity Results.
Companx2, complik2, and compcon2 explained a significant proportion of variance in compuse2, R2 = .76, F(3,36) = 38.11, p < .001. Also, This model this does not seem to have any multicollinearity issues (Tolerance > .1; VIF < 10)
Report in APA style the results of your multiple regression.
Based on the multiple regression analysis, Companx2 significantly predicted compuse2, b = .53, T(36) = 5.49, p < .001, compcon2 significantly predicted compuse2, B = .49, t(36) = 5.20, p < .001, but complik2 did not Significantly predict compuse2, b = .004, T(36) =.04, p = .966.