Experimental Examination of Audit Reporting Standards Data

Description

We conducted an experimental investigation on the impact of a Disclosure requirement (publicly disclosing the auditors actual name) and Sign-Off requirement (requiring auditors to sign-off on the audit report) on auditor misreporting decisions. We used an experimental audit market based on the economic model in Magee and Tseng (1990). Markets were conducted via networked computers utilizing z-tree software (Fischbacher, 2007) and student participants. Two experimental sessions were conducted for each of the four experimental conditions and each session was conducted with two markets operating simultaneously. Consistent with prior research, participants were randomly reassigned to a different market group at the end of each market year (two trading periods) in order to help control reputational concerns among participants, and to remove market group confounds (Calegari, Schatzberg and Sevcik (1998); Schatzberg, Sevcik, Shapiro, Thorne and Wallace (2005); Blay, Gooden, Mellon and Stevens (2019a); Blay, Gooden, Mellon and Stevens (2019b). All manipulations were administered on a between-subjects basis and all participants maintained their original randomly assigned role throughout the experiment. For a detailed description and discussion of the audit market setting, including equilibrium predictions see Blay et al. (2019b). We used the same experimental parameters as the audit markets described in Blay et al. (2019b), with three notable differences. First, we increased the penalty for misreporting from $0.05 to $0.25 in order for the economic penalty for misreporting to be consistent with Blay, Gooden, Mellon and Stevens (2019a). This enables us to better directly compare our results to prior research (Blay et al (2019a) and Schatzberg, Sevcik, Shapiro, Thorne and Wallace (2005)), while also allowing us to understand the interplay between economic penalties and the underlying theoretical mechanisms behind the observed reporting behavior. Second, we eliminated the pre-experimental quiz included in the instructions used in Blay et al. (2019b). Instead, we report the results of the Ethics Position Questionnaire (EPQ), Forsyth (1980), which can be used to measure individual differences in moral thought using its idealism and relativism scales. Lastly, Blay et al. (2019b) presented a 3x1 between-participants experimental design in which disclosure is manipulated at three levels. In particular, Blay et al. (2019b) examined the impact no disclosure, disclosure only, and disclosure with sign-off on auditor misreporting. Instead, we utilized a 2 x 2 between-participants experimental design in which we examined following experimental conditions: No Sign-Off or Disclosure, Sign-Off Only, Disclosure Only, and Sign-Off with Disclosure. We believe that this design choice allows for a more direct comparison between the results of the two studies. In addition to the raw data from the study, the reported Excel file includes a tab with data descriptions.

Steps to reproduce

Utilize the attached z-tree code and experimental materials.

Institutions

Categories

Funding

Licence