You are not currently logged in.
Access your personal account or get JSTOR access through your library or other institution:
Teaching about Influence in Simple Regression
Frederick O. Lorenz
Vol. 15, No. 2, Teaching Research Methods and Statistics (Apr., 1987), pp. 173-177
Published by: American Sociological Association
Stable URL: http://www.jstor.org/stable/1318032
Page Count: 5
Preview not available
Most introductory social science statistics courses include an abbreviated discussion of regression assumptions and the problem of outliers. An important idea in residual analysis is that all outliers are not equal; some have greater "influence" than others on the intercept and slope of the prediction equation. An example first proposed by Anscombe (1973) is used and then extended to illustrate how the idea of influence, as developed by Cook (1977; 1979), can be incorporated into introductory lectures on simple regression.
Teaching Sociology © 1987 American Sociological Association