NotesFAQContact Us
Collection
Advanced
Search Tips
ERIC Number: ED445022
Record Type: Non-Journal
Publication Date: 2000-Apr
Pages: 34
Abstractor: N/A
Reference Count: N/A
ISBN: N/A
ISSN: N/A
Four Methods of Handling Missing Data in Predicting Educational Achievement.
Witta, E. Lea
Four methods of handling missing data were applied to missing values for variables selected from the National Education Longitudinal Study of 1988. Variables used were those selected by K. Singh and M. Ozturk (1999) for a study concerning high school students' academic achievement and work. Samples selected consisted of 100 cases, 300 cases, and 500 cases. The proportion of incomplete cases was manipulated to represent 30%, 50%, and 70% for each sample. In addition, composite variables were created and tested. Results indicate the expectation maximization (EM) algorithm and regression procedures provide accurate estimates under all conditions. Listwise and pairwise deletion were effective with small proportions of missing data and when composites were created. (Contains 1 figure, 8 tables, and 19 references.) (Author/SLD)
Publication Type: Reports - Research; Speeches/Meeting Papers
Education Level: N/A
Audience: N/A
Language: English
Sponsor: N/A
Authoring Institution: N/A