Data analysis of social networks is often impeded by the problem of missing data. Recent studies highlight the negative effects of this problem mainly regarding querying process. The analysis of data social networks would be severely distorted when limited to filled fields (i.e., not null valued fields) whereas missing data are ignored. To overcome the missing data problem, we provide in this paper an extension of classical Drilldown and Rollup operators in order to support analyses on multidimensional datasets containing missing values of dimension members.