NotesFAQContact Us
Search Tips
Peer reviewed Peer reviewed
PDF on ERIC Download full text
ERIC Number: EJ1111310
Record Type: Journal
Publication Date: 2008-Apr
Pages: 29
Abstractor: As Provided
ISSN: EISSN-2330-8516
Model-Based Weighting and Comparisons: Research Report. ETS RR-08-17
Qian, Jiahe
ETS Research Report Series, Apr 2008
In survey research, sometimes the formation of groupings, or aggregations of cases on which to make an inference, are of importance. Of particular interest are the situations where the cases aggregated carry useful information that has been transferred from a sample employed in a previous study. For example, a school to be included in the sample of the High School Effectiveness (HSES) study must contain one or more cases transferred from the National Educational Longitudinal Study of 1988 (NELS:88). To calculate the aggregation inclusion probabilities, this study investigated three statistical models and, based on these models, derived the school weights for the HSES study. This study also assessed the effects of weighting by comparing the statistics yielded from different sets of weights: (a) those from an empirical population database and (b) those from data generated from simulation based on the principles of a superpopulation. Both categorical data and continuous variables were analyzed in the comparison.
Educational Testing Service. Rosedale Road, MS19-R Princeton, NJ 08541. Tel: 609-921-9000; Fax: 609-734-5410; e-mail:; Web site:
Publication Type: Journal Articles; Reports - Research
Education Level: High Schools; Secondary Education; Grade 10; Grade 8; Junior High Schools; Middle Schools; Elementary Education
Audience: N/A
Language: English
Sponsor: Office of Educational Research and Improvement (ED)
Authoring Institution: N/A
Grant or Contract Numbers: R999G50001; 84999G