Haphazard Sampling: Selection Biases Induced by Control Listing Properties and the Estimation Consequences of these Biases

Abstract

ABSTRACT Haphazard sampling is a nonstatistical technique used by auditors to simulate a variety of random sampling techniques when testing the error status of accounting populations. In this study, we compare the properties of haphazard samples selected from control listings with the properties of simple random samples. We hypothesize that control listing entries exhibit salience values that result from the effort required to locate entries and the visual properties of entries. We further hypothesize these salience values influence sample selections and result in sample properties that are different from those of simple random samples. To test these hypotheses, we examine the properties of haphazard samples selected by three participant groups. In each group, sample properties differ from those of simple random sampling and include a lack of independence across sample selections and biased sample inclusion probabilities. We also develop models showing how biased sample inclusion probabilities influence error projections and discuss the estimation consequences of these biases. Data Availability: For information about data availability, please contact the first author.