im using a help-seeking attitudes measure that generates three distinct sub-scales. Two of the three subscales captures different dimensions of help-seeking attitudes while the third measures a persons intention to seek psychological help. In my analyses, I was using the intention subscale as my DV or outcome And the other two sub-scales as potential mediators. therefore, my question is, is it wrong or wise to use subscales from the same measure as outcome and predictor variables? The author of the scale recommended that researchers employ the three sub-scales instead of the total score, given that the total score represents two constructs (attitudes and intention).

thank you