I have a dataset that includes a number of variables that need to be reverse coded. I have already completed my missing data analysis, however, I did not reverse code the items prior to replacing the missing data (using the EM algorithm). I am also about to complete an exploratory factor analysis (EFA) on the data, followed by a confirmatory factor analysis (CFA) using SPSS.

Two questions:

  1. Should I have reverse coded all items prior to replacing the missing data? Will this have influenced which numbers were used to replace the missing data, and if I reverse code items now, will it be wrong?

  2. Should I have reverse coded the items prior to the EFA and CFA? I have heard that it can be okay to leave the reverse-coded items in EFA, as they will still load on the correct factors, but just be negatively loaded. However, I am wondering if this is correct/best practice?


1 Answer 1

  1. I don't think it should matter because you're just changing the scale of the variables, not the relationship between a given variable and other variables.

  2. Yes, you're right, it should just affect whether loadings are negative or positive. If you're ultimately planning on average items together, then I would just reverse code them anyways because it's easier to interpret.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.