I. Myrtveit et al., Assessing the benefits of imputing ERP projects with missing data, SEVENTH INTERNATIONAL SOFTWARE METRICS SYMPOSIUM - METRICS 2001, PROCEEDINGS, 2000, pp. 78-84
Incomplete, or missing, data is likely to be encountered in empirical softw
are engineering data sets. fn this paper we evaluate some methods for handl
ing missing data. The methods are presented and discussed in general and th
ereafter applied to effort estimation of ERP projects. We found that two sa
mpling-based methods, mean imputation (MI) and similar response pattern imp
utation (SRPI), waste less information than listwise deletion (LD). However
, MI may introduce more bias than the SRPI method. Compared to sampling-bas
ed methods, likelihood-based imputation methods require too large data sets
to be realistic to use in empirical software engineering. None of the samp
ling-based methods, such as MI and SRPI, seem able to correct bias. So, tho
ugh imputation is an attractive idea, the available methods still have seve
re limitations.