Impact of corrections for dynamic selection bias on forecasts of retention behavior

Yang Li and Walter J. Mayer
Additional contact information
Yang Li: Department of Economics and Business, Kalamazoo College, Kalamazoo, Michigan, USA, Postal: Department of Economics and Business, Kalamazoo College, Kalamazoo, Michigan, USA
Walter J. Mayer: Economics Department, University of Mississippi, University, Mississippi, USA, Postal: Economics Department, University of Mississippi, University, Mississippi, USA

Journal of Forecasting, 2007, vol. 26, issue 8, 571-582

Abstract: We investigate the impact of corrections for dynamic selection bias on forecasting accuracy in a multi-period stay|leave model. While corrections for selection bias are needed for consistent coefficient estimates, they do not necessarily produce more accurate forecasts than uncorrected techniques. Theorem 1 shows that, apart from estimation errors, a shrinkage principle applies: the heterogeneity restriction imposed by uncorrected and combination techniques improves accuracy for forecasting individuals that leave, and hurts accuracy for forecasting individuals that stay. This has important implications for decision making because of the potential for asymmetric losses. We also present an illustrative empirical application and results from Monte Carlo experiments. We find that differences in relative accuracy vary directly with the degree of selection bias and inversely with the percentage of the initial population that stays. Copyright © 2007 John Wiley & Sons, Ltd.

Date: 2007
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
http://hdl.handle.net/10.1002/for.1028 Link to full text; subscription required (text/html)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:jof:jforec:v:26:y:2007:i:8:p:571-582

DOI: 10.1002/for.1028

Access Statistics for this article

Journal of Forecasting is currently edited by Derek W. Bunn

More articles in Journal of Forecasting from John Wiley & Sons, Ltd.
Bibliographic data for series maintained by Wiley-Blackwell Digital Licensing () and Christopher F. Baum ().