search for: a_handy_trick_f

Displaying 1 result from an estimated 1 matches for "a_handy_trick_f".

2012 Jul 04
1
How do you impute missing data using Latent Class Model (poLCA package)
...columns? How would you use Latent Class Model specifically poLCA to impute the missing data? http://www.sscnet.ucla.edu/polisci/faculty/lewis/pdf/poLCA-JSS-final.pdf The reason why I chose not to use Multiple Imputation(MI) is because according to [http://blogs.iq.harvard.edu/sss/archives/2008/09/a_handy_trick_f.shtml] "MI packages assume the Multivariate Normal Distribution which may not hold for certain types of categorical and binary data. Yucel Recai, Yulei He, and Alan Zaslavsky point out in their May 2008 article in The American Statistician, naive rounding MI imputations can bias estimates, pa...