Title: Measuring type I and type II errors in an estimation model: an empirical analysis; the case of bankruptcy
Authors: Musa Darayseh, Elaine Waples
Addresses: School of Management, American University of Sharjah, P.O. Box 26666, Sharjah, UAE. ' School of Management, Purdue University Calumet, Hammond, Indiana 46323, USA
Abstract: The objective of this study is to generate a model that will measure the cost of type I and type II errors in a logistic estimation model. The paper|s objective also is to find out how cost behaves when the cut-off point changed in an estimation model which leads to different decisions that should be made by investors and decision-makers. In other words, how will the investors and decision-makers behave when the expectation about the company|s future changed which will lead to different costs in regard to making an investment decisions? The results in this study indicate that the best possible cut-off score and the percentage of type I errors fall as the relative cost of type I error increases, reflecting the changing trade-off between type I and type II errors. The results also show that changing in cut-off point lead to different decisions from investors and decision-makers.
Keywords: misclassification; type I error; type II errors; bankruptcy; measurement; logistic estimation models; investment decisions; decision making; behavioural finance.
International Journal of Behavioural Accounting and Finance, 2010 Vol.1 No.3, pp.268 - 275
Published online: 31 Jan 2010 *Full-text access for editors Access for subscribers Purchase this article Comment on this article