I have been asked by a journal editor to combine two separate measures of child disruptive behaviour into one variable. One meaure is parent rated and is from the 36-item Eyberg Child Behaviour Checklist. It is measured on a 7-point likert scale and gives a total score. The other measure is teacher-rated and is from the Total Difficulties score from the Strengths and Difficulties Questionnaire. It is based on a 3-point likert scale. They want me to combine these two measures into one variable, however I do not know how I should do this as they are separate measures from separate rating scales. I wondering about standardising these two variables but then did not know whether an average of these two standardised variables would suffice. Your help would be very appreciated.
I'm not sure it's really sensible to combine these two measures as they are from different raters and different existing (and presumably validated) questionnaires. If the journal editor is concerned about the number of measures you're presenting, I think it would be more sensible to present only one set in the results, with a note that "results using the [other] measure were similar" (assuming they were). The other set of results could be included in an online supplement if the journal allows that.
Standardise the two measures into z-scores, and then simply add the values together into one single scale (you can either average or not the two z-scores). Don't need to worry about the original scores, standardisation means they would be measured in the same units! The two measures do seem to assess different aspects of a similar construct.