Converting Data into Evidence by Alfred DeMaris & Steven H. Selman

Author:Alfred DeMaris & Steven H. Selman
Language: eng
Format: epub
Publisher: Springer New York, New York, NY

We see here that, instead of the probability having a linear relationship to x—as would be true of the linear regression function—its curve is S shaped, always remaining within the bounds of 0 and 1. The solid line is for the probit function, which is used in probit regression, and the dashed line is for the logit function, which is used in logistic regression. We focus only on the logit function in this primer, as it is the preferred technique in medical research for a binary study endpoint. The probit function is used extensively in other fields, such as economics and other social sciences. However, as we shall see, the logit function lends itself to the interpretation of explanatory variable effects in terms of “odds ratios,” which is intuitively appealing. The probit function does not have this property. Substantively, however, both modeling techniques result in the same conclusions about the sign and significance of explanatory variable effects on the study endpoint (DeMaris 2004).

The logistic regression model for a probability, as a function of two regressors, is

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.