Error: "There are aliased coefficients in the model" for an ANOVA
Hello,
I'm attempting to run an ANOVA examining the effects of education and income on aggression (called sum sqrt) but JASP is returning an error message of "There are aliased coefficients in the model". I saw a previous thread that said it was due to different missing participant data but the descriptive table doesn't show any missing values. The values do work for education and income individually as well as another interaction with seperate variables (angry & intentionality). I've attached the jasp file as well as the results panel below. Any advice would be greatly appreciated!
Comments
Hi Hfac,
I looked at our dataset, and because you included the interaction term education x income, there need to be at least 2 observations per cell. If you look at "Descriptives", then select one variable, select the other variable as the "split by" variable, and select "frequency tables", you will see that for certain combinations of these 2 factors, there are only 1 or 0 observations:
This will make it impossible to estimate the variance for those cells, and therefore the F statistic cannot be calculated. The new version of JASP (hopefully released this week!) will have a more informative error message for this.
Kind regards,
Johnny