Hello Sarah,

How did you set up the GLM model to determine whether group membership influences the relationship between the two variables of interest?

Ordinarily, I would presume you would set Variable 1, Group, and Variable 1 * Group as three IVs, and use Variable 2 as the DV. The interaction term would be the one to suggest that group membership influences the relationship.

Good luck with your work.

GLM using : Univariate regression analysis

to analyse whether chronological age predict memory in a similar way across group A and group B.

the results of the univariate regression analysis : there was NO significant effect of group and chronological age to predict memory .

on the other hand, Fisher-z transformation which test the difference between the two correlations in group A and group B found a Significant difference between the two correlations chronological age * memory.

Hi. Yes, it is necessary to use regression model. Your article results will become more valid. Because the regression model is higher than Pearson et al.

Hello Sarah, did you test for group x age interaction as a predictor (as suggested in the first comment)?

Hi David

yes group x age interaction as predictors .

my question is : i found it weird to have significant correlation, but non significant for the GLM model

Hi Mahsima,

my question is I am havinf different results in correlation vs. GLM model