ERIC Number: ED221553
Record Type: Non-Journal
Publication Date: 1981
Reference Count: N/A
The Use of Within-Group Slopes as Indices of Group Outcomes.
Burstein, Leigh; And Others
The possibilities and problems in the use of within-group slopes of outcomes on inputs as indicators of substantive educational group effects are considered. Slopes are proposed as outcome measures which may reflect within-group processes in between-group analyses of multilevel data. Research on aptitude x treatment interactions, contextual effects and school effects provide a theoretical rationale for the proposed methodology. Science data (Comber and Keeves, 1973) on 14-year-olds is used as an empirical example to illustrate a group-level analysis with slopes. In this example, a school-level analysis (N=107 schools) of factors affecting science test performance is considered. The findings of steeper within-group slopes related to greater opportunities for exposure to science instruction and with greater emphasis on individual exploration are consistent with results from research on informal/open, individually guided/less structured instruction. Further study is needed for schools with specific combinations of discovery approach to instruction, amount of science instruction and verbal ability. Despite their theoretical and empirical appeal, the use of within-group slopes as outcomes is fraught with problems. The statistical, empirical, substantive, and communication problems that arise from the use of slopes as outcomes are discussed. (Author/PN)
Publication Type: Guides - Non-Classroom; Reports - Research
Education Level: N/A
Sponsor: National Inst. of Education (ED), Washington, DC.
Authoring Institution: California Univ., Los Angeles. Center for the Study of Evaluation.