Virginia Polytechnic Institute
The logistic model in the analysis of binomial response data
Abstract
dc:description.abstractThis paper is concerned with the experimental situation in which a response, y<sub>i</sub>, can be written as r<sub>i</sub>/n<sub>1</sub? where r<sub>i</sub> is binomially distributed with mean n<sub>i</sub>P <sub>i</sub>. The logistic model is one of three models most commonly used for this situation, and, for one independent variable, is of the form P<sub>i</sub> = 1 / 1 + e<sup>-a-Bx<sub>i</sub></sup> Reasons are given in Chapter I for preferring this model over the other two, the Integrated Normal and Angular models. Previously published literature on the logistic model, along with some new concepts are coined in this thesis in an effort to provide the experimenter with 2. detailed and comprehensive presentation of the theory and techniques involved in employing the logistic model. Three methods for estimating the parameters in the logistic model are discussed: maximum likelihood, minimum X<sup>2</sup> , and minimum logit X<sup>2</sup>; and the techniques involved in these estimation procedures are demonstrated in two completely worked examples.
Degree
thesis:*- Name thesis:degree_name
- M.S.
- Level thesis:degree_level
- masters
- Discipline thesis:degree_discipline
- Statistics
- Department dc:contributor.department
- Statistics
- Grantor dc:publisher
- Virginia Polytechnic Institute
- Year dc:date.issued
- 1969
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- DeRouen, Timothy Allen
- Chair dc:contributor.committeechair
-
- Myers, Raymond E.
- Committee members dc:contributor.committeemember
-
- Hinkelmann, Klauss
- Harshbarger, Boyd
Rights
dc:rights- Statement dc:rights
-
- In Copyright
- Licence dc:rights.uri
- Language dc:language.iso
- en
Identifiers
dc:identifier.*- Handle dc:identifier.uri
- https://hdl.handle.net/10919/141379