The Acceptability of Regression Analysis as Evidence in a Courtroom--Implications for the Auditor.
回顾了法庭中应用回归分析的案例，分析了影响其证据力的因素，并探讨了在审计中使用的可接受性，为审计师提供了模型构建建议。
Abstract Despite numerous academic articles concerning the application of regression analysis as an audit tool, the acceptability of the tool in the legal setting has not been explored. This article reviews court cases involving the application of regression analysis, analyzing factors which influenced the evidentiary force given to testimony that was reliant on regression analysis. Some general issues involved in the use of such statistical evidence, as well as the effect of the "hearsay" rule on individual testimony in jury settings, are discussed. Since no cases have involved regression applications to the audit, no direct proof exists of whether or not the use of regression analysis in analytical review would be considered acceptable in a court of law. However, actual cases where regression analysis has been used in court suggest that if the technique is clearly explained in submitting testimony, if the rationale for the selection of variables is included in the working papers, and if violations of the regression assumptions are investigated, legal support exists for accepting the application of regression as evidence. Further, the ability of the evidence to withstand a technical rebuttal (typical of academicians' critique of empirical studies) is demonstrated. Recommendations are provided as to how to approach model building based on the analysis of relevant court cases.