For example,you may be interested in studying risky sexual behavior among school adolescents.In the survey you get data on only those who were found to be sexually active at the time of the survey.Hence,those who were not sexually active will be system missing on your software of choice.How do we manage such issues during analysis?