An Examination of Convergent Validity between In-Lab and Out-of-Lab Internet-Based Experimental Accounting Research

Abstract

Although behavioral research is increasingly relying upon the Internet for data collection, few Internet-based experimental accounting studies have been published. Bryant et al. (2004) discuss the advantages and disadvantages of Internet-based research and conclude that experimental validity poses the largest potential concern. One explanation for the scarcity of internet-based accounting studies may be that accounting researchers find experimental validity concerns outweigh the benefits of Internet data collection. Krantz and Dalal (2000) report that Internet psychology experiments produce results similar to laboratory findings. However, experimental economics researchers find that results differ between Internet- and Lab-based settings. In this paper, we examine convergent validity between in-lab and out-of-lab data collection with professional accountant participants. Using data from an Internet-based accounting experiment, we test convergent validity on those variables observed to interact with experimental delivery method (in-lab or out-of-lab) in prior research. Our results are consistent with that of the psychology literature. Specifically, no statistically significant interaction is reported between experimental delivery method and the independent variables or the demographic measures. We discuss factors contributing to this result and the implications for future behavioral accounting research.

Keywords: Internet-based experiments, convergent validity

JEL Classification: M4