Discriminant Analysis and Merger Theory

The use of linear-discriminant analysis has gained wide acceptance in applied business research. The question of whether different linear-discriminant models constructed to classify acquired firms as to their merger status are robust when multiple random samples are used and whether the models conta...

Full description

Saved in:
Bibliographic Details
Published inRBER, review of business and economic research Vol. 20; no. 1; p. 76
Main Author Wansley, James W
Format Journal Article
LanguageEnglish
Published New Orleans, La University of New Orleans, Division of Business and Economic Research 01.10.1984
University of New Orleans, Business and Economic Research
Subjects
Acquisitions & mergers
Discriminant analysis
Economic theory
Mathematical models
Studies
Variables
Online AccessGet full text

Cover

More Information
Summary:The use of linear-discriminant analysis has gained wide acceptance in applied business research. The question of whether different linear-discriminant models constructed to classify acquired firms as to their merger status are robust when multiple random samples are used and whether the models contain the same entering variables of linear-discriminant functions was examined. Variables were selected based on one of 2 reasons: 1. They were found important in earlier merger studies. 2. A sound theoretical basis for the variable's inclusion exists. The considerable variation found between the lowest and highest error rates in the sample indicates considerable sensitivity in error rates with respect to the particular firms chosen for matching. The results also indicate that past research using discriminant analysis in a similar context may be sample sensitive. Those who seek to identify merger candidates should expand their research beyond the limited number of variables identified in past research.
ISSN:1058-3300
0362-7985
1873-5924