The chi-square test of independence21 was used to determine whether apparent differences of opinion evident in the samples of role sender group responses relating to the standard of performance of auditors' duties, and the duties which auditors should perform, were statistically significant.
In applying the chi-square test to responses relating to the standard of performance of auditors' duties, the focus of attention was the proportion of the role sender groups who signified that auditors perform their duties poorly. To accommodate this, responses to Part 2, section 1b of the questionnaire were classified on the basis of whether they indicated that auditors perform their duties 'poorly' or 'not poorly'. For each duty, the responses from two role sender groups constituted a 2 x 2 table. The chi-square test was applied to the table to determine whether any difference in the proportion of the respective groups who signified that auditors perform the duty in question poorly was statistically significant.
Similarly, in applying the chi-square test to role sender group responses relating to duties which auditors should perform, the focus of attention was the proportion of each group who signified that a particular duty should be performed by auditors. Responses to Part 2, section 2 of the questionnaire were classified on the basis of whether they affirmed or did not affirm that a particular duty should be performed by auditors. The responses from two role sender groups, with respect to each duty listed in the questionnaire, constituted a 2 x 2 table. The chi-square test was applied to the table to
determine whether any difference in the proportion of the respective groups who signified that auditors should perform the duty in question was statistically significant.
The chi-square test was used to establish the significance or otherwise of apparent differences of opinion between and within auditors' major role sender groups, about the standard of performance of auditors' duties, and the duties which auditors should
perform. A significance level of 0.05 was adopted. If the significance of a calculated
chi-square statistic was less than 0.05, it was concluded that differences evident in the
samples of responses from the two role sender groups in question reflected differences of opinion in the underlying populations. If the significance of the calculated chi-square
statistic was 0.05 or greater, it was accepted that no valid conclusion could be drawn
about differences of opinion in the related populations.
The results of the chi-square test support the detailed deSCriptive and comparative
analysis of the survey findings relating to the performance of auditors' duties, and the
duties which auditors should perform, presented in Chapter 7 (sections 7.3 and 7.4). The
results are presented in tabular form in Appendices 4 and 5 respectively.
6.9 SUMMARY
In this chapter, the objectives and methodology of the empirical research conducted to
test normative propositions relating to the audit expectation-performance gap and the
role of external auditors in society (developed in Chapters 2 to 5), have been outlined. It
has been noted that a mail survey was employed to ascertain the opinions of auditors' role senders about auditors' duties, and their professional standing and specialist function in society. The methods used for identifying auditors' role sender groups and for selecting samples to include in the mail survey have been discussed. Details have also been given about the questionnaire design, the pilot study conducted to test the
questionnaire, the despatch of the questionnaires and the follow-up of non-respondents. Further, an overview of the response rates of the various role sender groups has been provided. Additionally, the steps taken to prepare the survey data for analysis, and the statistical tests performed to ascertain whether generalised conclusions could be drawn from the research findings, have been explained.
In the next two chapters, a detailed analysis of the role sender group responses is provided, and the survey findings are related to the postulated structure of the audit expectation-performance gap and social position of external auditors in society.
CHAPTER 7:
ANALYSIS OF EMPIRICAL RESEARCH RESULTS
PART I: THE DUTIES OF AUDITORS
7.1
INTRODUCTIONThis is the first of two chapters in which the survey results are described and analysed. In these chapters, the expectations of auditors' role senders regarding auditors' duties and social position are established, and differences in expectations between and within auditors' major role sender groups are identified. The survey findings are also examined
in the context of the normative propositions developed in Chapters 2 to 5 relating to the
audit expectation-performance gap and auditors' role in society.
In this chapter, attention is focused on the opinions of auditors' role sender groups regarding auditors' existing duties, the standard of performance of these duties, and the duties which auditors should perform. These issues coincide with objectives 1 to 3 of the empirical research which are set out in Chapter 6 (section 6.2). In Chapter 8, the expectations of auditors' role senders with respect to auditors' professional standing and specialist function in society are analysed, and the duties which may reasonably be expected of auditors are determined. These issues coincide with objectives 4 and 5 of the empirical research (see section 6.2).
The opinions of auditors' role senders with respect to auditors' duties are described and evaluated on the basis of:
( i ) the survey group as a whole; and
( i i ) differences which are evident between and within auditors' major role sender groups (auditors, auditees, financial community and general public audit beneficiaries).
For each of the issues addressed in this chapter (auditors' existing duties, the standard
of performance of these duties, and the duties which auditors should perform), details
of the survey findings, and the results of relevant tests performed to verify the
statistical significance of the data, are presented in Appendices 3 to 5. To facilitate
description and analysis of the survey results, relevant data are extracted from the Appendices and reproduced in the text.
7.2 EXISTING DUTIES OF AUDITORS
7.2.1 Methodology of Analysis
In this section the duties which auditors' role senders consider to be (and not to -be) existing duties of auditors are established, and differences of opinion between and within the major role sender groups are identified. This is accomplished by analysing responses to Part 2, section la, of the questionnaire. The duties which auditors' role sender groups consider to be existing duties of auditors are also compared with auditors' duties as defined by the law and professional promulgations; discrepancies indicate lack of knowledge about auditors' duties by their role senders. Such lack of knowledge has significance for the audit expectation-performance gap as it is perceived to be the major cause of the 'reasonableness' component of the gap.
The duties which auditors' role sender groups consider to be (and not to be) existing
duties of auditors are identified on the basis of the average or representative opinion of
each group. This is indicated by the mean (expressed as a percentage) of the 'Yes', 'No'
and 'Not sure' responses from the group in question, to each of the suggested duties listed
in the questionnaire. A positive mean indicates that the group considers that the
considers that it is not such a duty. The absolute value of the mean indicates how widely the opinion is shared amongst members of the group.
In order to verify that generalised conclusions can be drawn from the samples of responses obtained by the survey, the Wilcoxon signed-ranks test has been applied to the role sender group responses. Similarly, the Mann-Whitney test has been used to
establish whether differences of opinion evident in the samples of responses from the
role sender groups are statistically significant.
In
Appendices 3a to 3g, details of the survey findings with respect to auditors' existing duties, and the results of statistical tests performed on the data, are presented in tabular form. Appendices 3ato
3c provide the following information:3a: the means of responses from the survey group as a whole, and from the four major
role sender groups (auditors, auditees, financial community and general public
audit beneficiaries), to Part 2, section la, of the questionnaire;
3b: the results of the Wilcoxon signed-ranks test, applied to the responses from each of the above groups;
3c: the results of the Mann-Whitney test showing the Significance of differences of opinion reflected in the responses from the four major role sender groups.
Appendices 3d to 3g present similar sets of data for each of the subgroups constituting auditors' major role sender groups.