people said most used methods are chained equations and multivariate normal imputation. But in SPSS I can only found MCMC and monotone. Maybe they use different names?
Also, how to evaluate the imputation is good or not?
in a paper (Shrive FM, et al, 2006, BMC medical research methodology, 6:57 http://www.ncbi.nlm.nih.gov/pubmed/17166270
) they used Kappa statistics, but did not describe how Kappa was calculated.