Type I Error Rates For A One Factor Within-Subjects Design With Missing Values
{{output}}
Missing data are a common problem in educational research. A promising technique, that can be implemented in SAS PROC MIXED and is therefore widely available, is to use maximum likelihood to estimate model parameters and base hypothesis tests on these estimate... ...
