My question is about using Z-scores as independent and categorical variables. So, it is possible to create Z-scores from the scale variables and transform these Z-scores into categorical variables (1 30% or lower, 2 30-70%, 3 over 70% of the values).
Thank you for your time and interest!
I'm going to study the effects of the intervention on the quality of adult's interaction. This z-score (categorical) variable would serve as an independent variable to measure how will children's temperament characteristics effect on the interaction quality. Statistical analysing method is ANCOVA (SPSS).
Hello Sivi,
If you believe that the relationship of each continuous IV is likely linear with the DV, it would make more sense to use multiple linear regression rather than ancova. If you believe the relationship is non-linear, then you might have a more sensitive study by converting the z-score variates to categories, as you suggest.
Most statisticians would be leery of abandoning information by converting from a continuous variable to a categorical one; ordinarily, it's not a choice I would recommend.
Good luck with your work!
Thank you for your answer David Morse ! This is very helpful. According to studies in educational science it seems quite common to perform these transformations from a continuous to a categorical variable. I understand that this may not be the ideal way, but still quite usual one. My concern would be that is this a proper way to deal with z-score variables.
- Badges
- Science topic