Outliers and Robust Inference in Archival Accounting Research

57 Pages Posted: 21 Jan 2021

See all articles by Joachim Gassen

Joachim Gassen

Humboldt University of Berlin - School of Business and Economics; TRR 266 Accounting for Transparency

David Veenman

University of Amsterdam - Amsterdam Business School (ABS)

Date Written: December 23, 2020

Abstract

This study examines the nature of outliers in archival accounting research and evaluates the merits and limitations of robust estimators in identifying and downweighing their influence. Using simulated and actual data samples, we demonstrate how outliers can result from the data generating process, research design choices such as scaling, and model misspecification. Given the nonrandom nature of outliers, we show that inferences based on robust estimators can be biased when the relation of interest varies with variables correlated with the likelihood of outliers (e.g., firm size). We also find that a failure to account for nonlinear relations can induce biases in robust estimators that are more severe than with OLS. Overall, we advise researchers to acknowledge the nonrandom nature of outliers in their samples, to be cautious in implementing and interpreting robust estimation methods, and to evaluate and report the sensitivity of these methods to critical design choices. We also highlight the usefulness of combining data visualizations with robust estimation techniques to assess both the nature of outliers and their impact on inference.

Keywords: Outliers, Robust Regression, MM-Estimation, OLS, Nonlinearities, Selection Bias, Model Specification, Scaling

JEL Classification: C31, C52, G10, M41

Suggested Citation

Gassen, Joachim and Veenman, David, Outliers and Robust Inference in Archival Accounting Research (December 23, 2020). Available at SSRN: https://ssrn.com/abstract=3754202 or http://dx.doi.org/10.2139/ssrn.3754202

Joachim Gassen

Humboldt University of Berlin - School of Business and Economics ( email )

Spandauer Str. 1
Berlin, D-10099
Germany
+49 30 2093 5764 (Phone)
+49 30 2093 5670 (Fax)

TRR 266 Accounting for Transparency

Warburger Stra├če 100
Paderborn, 33098
Germany

David Veenman (Contact Author)

University of Amsterdam - Amsterdam Business School (ABS) ( email )

Roetersstraat 18
Amsterdam, 1018WB
Netherlands

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Downloads
190
Abstract Views
620
rank
189,049
PlumX Metrics