Main Article Content


To obtain the adequate evidence, auditor does not have to test all existing transaction. Along of cost benefit consideration, it is impossible for auditor to test all transaction evidence. Based on this consideration; then in profession recognized widely that most evidence obtained using sampling. The limited audit sampling research motivated the writer to conduct this research. This research is a development from previous researches by Hall of et al. (2002) and Zarkasyi (1992). Researcher take auditor (BI) as research subjects because sampling problems in audit differ from the practice make an audit of the private sector (Arkin, 1982). The purpose of this research is portraying how sampling audit practice in BI and explore factors affecting the use of sampling method by auditors.

The responses from 122 respondents show 70.5% respondents did not use the statistical sampling method. In nonstatistics sample selection method, two techniques which less get the support empirically namely haphazard and block sampling, in the second (32,6%) and third rank (11,6%). There is indication of selection bias mostly in the size measure and location. Most respondents (76,25%) answered that they did not get formal training in avoiding selection bias. But that way only 36,9% respondents replied they did not use the procedures to mitigate the selection bias.

There are four factors which hypothesized affecting the method used in audit sampling. These factors are auditor perception to statistical sampling method, auditor perception to perceived audit risk, time pressure and experience. Result from logit regression test indicated that among four factors hypothesized, only perception factor to statistical sampling method is significantly influent to the method used in audit sampling.


sampling audit governmental auditor selection bias

Article Details