Missing data and maximum likelihood estimation |
| |
Authors: | Cheng Hsiao |
| |
Institution: | Princeton University, Princeton, NJ 08544, USA;University of Toronto, Toronto, Canada M5S 1A1 |
| |
Abstract: | The proper use of the maximum likelihood principle in the context of missing data is explored. The relationship between the false procedure of treating the predicted values as if they were observable and the E–M procedure suggested by Dempster, Laird, and Rubin (1977) is also discussed. |
| |
Keywords: | |
本文献已被 ScienceDirect 等数据库收录! |
|