A A A

Following the previous post, I still get a further question about missing data. For example, suppose the respondent has too many questions not answered. Then, can we still assign a mean score on the missing items?