ERIC Number: ED294902
Record Type: RIE
Publication Date: 1988-Apr
Reference Count: 0
Differing Policy Implications of Alternate Multiple Regressions Using the Same Set of Student Background Variables to Predict Academic Achievement.
Levine, Daniel U.; Stephenson, Robert S.
Actual data sets were used to illustrate how substantially different conclusions and implications can be drawn from alternate multiple regressions predicting academic achievement from the same set of variables measuring student background. Variables assessing students' socioeconomic status and geographic mobility, used to predict reading and math achievement at 100 elementary schools, were interpreted for divergent conclusions about changes needed in schools. A second example showed differing interpretations of high school success, using 29 high schools. A third example considered dropout rates in over 300 elementary schools in a big city school district to support differing conclusions. Among the implications of the study are that statistics and their interpretation must take curvilinear relationships into account; that the influence of atypical schools should be considered; and that care should be exercised in selecting dependent variables. (SLD)
Publication Type: Speeches/Meeting Papers; Reports - Evaluative
Education Level: N/A
Authoring Institution: N/A
Note: Paper presented at the Annual Meeting of the American Educational Research Association (New Orleans, LA, April 5-9, 1988).