ERIC Number: ED213803
Record Type: RIE
Publication Date: 1981-Dec
Reference Count: 0
The Causal Analysis of Cognitive Outcomes in the Coleman Report.
Goldberger, Arthur S.; Cain, Glen G.
In the study "Public and Private Schools," the conclusions by James Coleman and others that private schools produce higher test scores and do so by employing certain school policies are not valid because the methods and interpretations used in the analysis fall below the standards for social-scientific research. Moreover, the conclusions are not warranted by their evidence. In implementing the research, Coleman and his associates used inadequate measures of cognitive outcomes, employed erroneous sampling procedures, and based conclusions on biased methods of analysis. One phase of analysis used a multiple regression approach with 17 background variables that was selectively biased against public schools. A second analysis, which attempted to show that Catholic schools are more egalitarian because they produce similar test scores among students with diverse backgrounds, was biased because it excluded relevant variables and because the range of student cognitive abilities was narrower in the Catholic sector than in the public sector. The third analysis which studied sophomore to senior change in scores did not control for background. Finally, the analysis which attributed test score gains in the private sector to selected school policies was based on choosing certain variables as "policy" without considering student ability and background. (Author/MJL)
Publication Type: Reports - Evaluative
Education Level: N/A
Sponsor: Wisconsin Univ., Madison. Graduate School.; Department of Health and Human Services, Washington, DC.; National Inst. of Education (ED), Washington, DC.; National Science Foundation, Washington, DC.
Authoring Institution: Wisconsin Univ., Madison. Inst. for Research on Poverty.
Identifiers: Public and Private Schools (Coleman et al)
Note: Also supported by funds from the William F. Vilas Trust Estate.