165 CHAPTER 4 DATA ASSESSMENT AND DESCRIPTIVE STATISTICS 4.1 Introduction This chapter begins by describing the data collecting and survey response rate, as well as the demographic profile of the respondents, which covers both the respondents' and the audit process' characteristics. The chapter then moves on to the results of the first and second-order latent constructs and their relative measurement items, followed by the data screening results in terms of missing values, outliers, and the assessment of the data normality. Moreover, this chapter includes the results of confirmatory factor analysis (CFA) for the measurement models (model 1 and model 2) in terms of uni-dimensionality, reliability, and validity. In addition to the acquired the findings for the assessment of the structural model in the testing of hypotheses on direct and moderation effects are given, the chapter presents the descriptive analysis for all items of the study variables. Finally, the chapter provides a summary of SAIs’ audit reports that related to municipalities either issue by the FACB or MOLG- GDCG. 4.2 Analysis of Survey Response The following subsections discuss the data collection including the respond rate and the demographic profile for respondents including characteristics of the respondents and the characteristics of audit process. 166 4.2.1 Data Collection The MOLG website registered 155 municipalities in Palestine, and the questionnaires were sent to these municipalities via their official email addresses, mainly, directed the URL of the Google Form to key persons in accounting and internal audit departments (Refer to Table 3.8). As a result of personal communication with accountants and internal auditors by telephone, mobiles, emails, WhatsApp groups, and other social media, the total of 186 questionnaires were collected, yielding a general response rate of 60.2%, but the respond rate in class C was 89%, 68% in class A, and 39% in class B as appeared in the Table 4.1. The response rate by official emails without following with telephone or other media around 17% was used by many scholars (Carini et al., 2018), but response rate increases after reminders sending (Saleh & Bista, 2017), if the first email followed by other email, the response rate will increase by 11.8% (Converse et al., 2008). Prior researchers accepted response rates of 31% and 26% when using email surveys distributed to financial statement preparers (accountants), and 21% when using email surveys distributed to financial statement users (Al-Dhubaibi, 2020). In light of this, the study's response rate of 60.2% seems appropriate. Table 4.1: Responses Rate Class A B C Total Estimated number of employees (Population) 75 141 93 309 The number of respondents 51* 52 83 186 The respond rate 68% 39% 89% 60.2% The number of municipalities 15 47 93 155 Average number of respondents in one municipality 3.4 1.1 0.89 1.2 *22 respondents in class A+(center of area) and 29 in class A (center of governance) Source: Author 167 The relative decrease in response rate in municipalities class B and class A can be attributed to the municipality's management notion that few respondents from the municipality is sufficient, particularly from the main accountants who represent the municipality. And the overall average number of respondents in one municipality was 1.2 employees, indicating that the majority of municipalities were participating in the questionnaire response. As a result, the response rate is valid and representative of the study population. All of the questionnaires that were collected were used for the analysis in the study because each questionnaire had been verbally scanned to remove any missing responses, but there were no missing values in the study's variables because each question's response was eligible. As a result, all questionnaires were immediately verified using the Google occlusion tool. The general rule of thumb for determining sample size, according to Sekaran and Roger (2003), is to multiply the number of constructs by 10. Given that there are 10 constructs (variables) in this study, the required sample size should be at least 100 observations (10*10). However, the 186 usable measured values in the current study met the aforementioned criteria, allowing the researcher to move forward with additional analyses. 4.2.2 Demographic Profile All accountants and internal auditors of municipalities in Palestine make up the study’s population. In order to see the description of the demographic profile of respondents and the audit process in the municipalities, frequency analysis was performed using SPSS version 27. Table 4.2 displays the demographic profile of this study, which is divided into two categories: (1) respondents' characteristics, which relate to the description of the respondents' personal qualifications, and (2) audit 168 process characteristics, which relate to the respondents' experience with audit processes performed by external auditors, internal auditors, and SAI auditing. Table 4.2: Sample Profile (N = 186) Group Frequency Percentage Cumulative Percentage The first Group: Respondents Characteristics 1- Occupation Accountant 44 23.7 23.7 Senior Accountant 57 30.6 54.3 Accounting Department Head 72 38.7 93 Internal Auditor 13 7.0 100 2- Gender Male 132 71.0 71 Female 54 29.0 100 3- Age Less than 30 years old 26 14.0 14 30-40 years old 62 33.3 47.3 41-50 years old 65 34.9 82.3 More than 50 years old 33 17.7 100 4- Qualification Less than Bachelor Degree 1 .5 0.5 Bachelor’s Degree or equivalent 144 77.4 77.9 Master Degree 36 19.4 97.4 PhD Degree 3 1.6 99 Bachelor’s Degree in other field 2 1.1 100 5- Experience Less than 5years 20 10.8 10.8 5-10 years 40 21.5 32.3 11-15 years 43 23.1 55.4 More than 15 years 83 44.6 100 Second Group: Audit Process Characteristics 1- Municipality Class Class A+ 22 11.8 11.8 Class A 29 15.6 27.4 Class B 52 28.0 55.4 Class C 81 43.5 98.9 Class D 2 1.1 100 2- Audit Fees in USD Less than 2000 102 54.8 54.8 From 2001 to 4000 47 25.3 80.1 From 4001 to 6000 13 7.0 87.1 More than 6000 18 9.7 96.8 I do not know 6 3.2 100 3- Accounting Basis Cash Basis 79 42.5 42.5 Accrual Basis 63 33.9 76.4 Modified Accrual Basis 31 16.7 93.1 Mix as the type of budget 13 7.0 100 4- Number External Auditor in the team Two auditors 104 55.9 55.9 Three auditors 48 25.8 81.7 Four auditors 21 11.3 93 Five auditors or more 13 7.0 100 169 Table 4.2, continued Group Frequency Percentage Cumulative Percentage 5- Internal Auditors Number None 97 52.2 52.2 One employee 48 25.8 78 Two employees 19 10.2 88.2 Three employees or more 22 11.8 100 6- Last Year Audit Report 2018 6 3.2 3.2 2019 2 1.1 4.3 2020 32 17.2 21.5 2021 138 74.2 96 Never Audited 8 4.3 100 7- Last Year Auditor’s Report type Standard Unmodified 131 70.4 70.4 Unmodified with Emphasis Matter 12 6.5 76.9 Qualified Opinion 15 8.1 85 Adverse Opinion 2 1.1 86 Disclaimer 7 3.8 90 No audit in the municipality 19 10.2 100 8- Last Year SAIs Audit 2019 23 12.4 12.4 2020 33 17.7 30.1 2021 72 38.7 68.8 2022 43 23.1 92 Never Audited 14 7.5 100 Source: SPSS 27 Software 4.2.2.1 The First Group: Demographic Characteristics The first demographic question was the subject of the respondents’ employment position (occupation). According to Table 4.2, the Accounting Department Head (38.7%) received the most responses, followed by Senior Accountants (30.1%), Accountants (23.7%), and Internal Auditors (7%). Given that the head of the accounting department and senior accountant are constantly deeply involved in the preparation of the financial statements as well as communication with external auditors in addition to internal auditors, this suggests that the respondents were competent in responding to the questionnaires that were distributed. The second demographic question was the subject of the gender of the respondents. The results of Table 4.2 show that 71% of the respondents were men and 170 29% were women. As a result, men make up the bulk of the respondents in this study. This might be a result of the Palestinian culture, which discourages women from working outside the home as employees, especially in municipalities. However, in recent years, this culture has changed as a result of women attending universities and earning degrees that qualify them for high-level positions in the workforce. The third question asked respondents to enter their age. According to frequency statistics, the majority of respondents were between the ages of 41 and 50 (34.9%), followed by those between the ages of 31 and 40 (33.3%), those over 50 (17.7%), and those under 30 (14%) respectively. This result as Table 4.2 shows means that younger people now play a less significant role than older people in the accounting departments of municipalities. This goes back to the era of municipality establishment, which began following the establishment of the PNA in 1993. The elderly is never in favor of implementing new accounting methods like accrual accounting, new accounting software, and the adoption of (IPSASs). However, older accountants may have more practical experience and be more qualified to respond to this survey with reliability. Regarding the fourth question, which related to the respondents' work experience. Table 4.2 shows that the majority of respondents (44.6%) had more than 15 years of experience, followed by those with 11 to 15 years of experience (23.1%), those with 5 to 10 years of experience (21.5%), and those with less than 5 years of experience (10.8%). This result suggests that the respondents have relevant experience working in municipal accounting. As a result, this shows that the respondents have sufficient knowledge of audit quality and its factors, and raises the credibility of the responses provided on the distributed questionnaires. 171 The fifth demographic question asked about the respondents' current educational status in relation to their level of education. The most common level of education among respondents was a bachelor's degree (77.4%), a master's degree (19.4%), a doctorate (1.6%), a bachelor's degree in another field (1.1%), and respondents with less than a bachelor's degree (0.5%). Although the law of local governmental units and the regulations permit the employment of accountants from diploma degree if the accountant was employed before year of 2009 (Office, 2020) , as shown in Table 4.2, the municipalities in Palestine were concerned about the educational level of the accountants and the internal auditors. This suggests that the respondents were competent in responding to the questionnaires that were distributed. 4.2.2.2 The Second Group: Audit Process Characteristics In the first question of audit process characteristics, participants were asked to enter the classification of the municipality where the participant works. Frequency statistics demonstrates that the majority of municipalities was classified as class (C) for (43.5%), followed by class (B) for (28%), class (A) for (15.6%), class (A+) for (11.8%) and lowest class (D) was (1.1%) which is transmitted to class (C) according the minister of the Local government ministry in Palestine. The structure of these percentages alignment with the actual structure classes of municipalities of Palestine. This structure of participants gives more credible for the answers of the questionnaire, in class (C), always there is one accountant who responsible on the accounting system and the communication with the external auditors, therefore he will have qualified perfectly to answer the questions. The second question of the audit process characteristics was the subject of the audit fees. According to Table 4.2, most municipalities audit fees were in lowest 172 category less than 2000 USD (54.8%), followed by category from 2001USD to 4000 USD (25.3%), category from 4001 USD to 6000 USD (7.0%), category more than 6000 USD (9.7%) and (3.2%) the respondents did not know the audit fees. These rates reflect the municipalities size and their classification, and in general, the amount of audit fees in the municipalities is low when we compare it to the level of audit fees in the business organization in Palestine. The use of accounting bases was the subject of the third audit process query. According to Table 4.2, the majority of municipalities in Palestine (42.5%) still use the cash basis for accounting, which is followed by accrual basis (33.9%), modified accrual basis (16.7%), and mixed basis (7.0%), respectively. Participants are better able to respond to questions about the accounting basis as a factor of internal control effectiveness and audit quality in municipalities as a result of their growing familiarity with various accounting bases and their impact on the accuracy of financial statements and the quality of audits. Regarding the fourth question, which inquired about the number of audit team individuals in the audit engagement in the municipality, table 4.2 shows that the majority of municipalities (55.9%) were audited by two auditors, followed by three auditors (25.8%), four auditors (11.3%) and (7.0%) five auditor or more. This result means that most municipalities are audited by small audit firms which they have limited number of auditors, and may reflects the simplicity and small size of most municipalities in Palestine. The fifth question of the audit process was related to number of internal auditors in the municipality, table 4.2 shows that (52.2%) of the municipalities have not internal audit as a separate function, because the function of internal audit is not required from the municipalities by law and regulations, However, MOLG issued 173 organizational structure models for municipalities based on their size and class to serve as guidelines for preparing a proper organizational structure for each municipality. According to these models, MOLG required all classes of municipalities to form a committee of municipal council members to perform at least the function of internal auditing and controlling, and required class (A) and recommended class (B) to establish an internal audit department, either supervised by the council or the financial manager (Office, 2020). Also, 25.8% of the participants have one internal auditor in their municipalities, 10.2% have two internal auditors, and 11.8% of the participants have three internal auditors. The sixth question of the audit process characteristics was the subject of the last year audit report is issued by the external auditor for the municipality. According to Table 4.2, most municipalities audited their financial statements in year 2021, this means that most municipalities have recent experience in the external audit process and make audit regularly, therefore the percentage of participants who finished the external audit for 2021 year in the last quarter of 2022 year is (74.2%), and 17.2% of participants have audit report for year 2020, but (4.3%) have not external auditing, and (3.2%) did not audit since 2018, and (1.1%) since 2019. The seventh question of the audit process characteristics was the subject of the last year audit report type is issued by the external auditor for the municipality. According to Table 4.2, most municipalities get a standard unmodified audit report which reached (70.4%) of participants who get unmodified audit report, followed by qualified opinion was (8.1%), unmodified with emphasis matter was (6.5%), disclaimer was (3.8%), and (1.1%) for the adverse opinion. The last question of audit process was related with to last year the municipality are audited by SAIs auditors in order to know the extent of the experience of the 174 participants with audit of FACB as SAI in Palestine, most of participants have recent experience with the audit of FACB, this means that the participants able to evaluate the impact of auditing of FACB on the external audit quality. Table 4.2 shows that (23.1%) of the participants exposed for the audit of FACB in 2022, but 38.7% in year 2021, 17.7% in year 2020, 12.4% in 2019, and 7.5% of the participants have not exposed to this type of audit. 4.3 Construct Measures The primary construct measures were built upon already-in-use tools. The measurement components for the research variables, as well as the first and second order constructs, are summarized in Table 4.3. Table 4.3: List of Constructs and Measurement Items 2nd Order Construct 1st Order Construct Items Number (50) Measurement Scale Audit Quality (AQ) 8 5-Point Likert Supreme Audit Institutions (SAI) 11 // Auditor Characteristics (ACH) Ethics (ET) 6 // Independence (IN) 6 // Competency (CM) 7 // Audit Firm Attributes (AFA) Audit Fees (AF) 2 // Audit Firm Size (AFS) 2 // Effectiveness of the Municipal Internal Control (EMIC) Internal Auditing (IA) 2 // Accounting Basis (AB) 3 // Laws and Regulation (LR) 3 // Source: Author 4.4 Data Screening In order to ensure that data are correctly entered and free of missing values, data screening is required. This section also looked at normality, univariate outliers, and multivariate outliers. 175 4.4.1 Missing Values For administering or distributing the survey to the respondents in the current study, a self-administered method was used through using the information technology and current communication tools such as Emails, WhatsApp, Telephone Calls, and other social media. But if any of the survey participants appeared to be having trouble understanding a particular question or statement, they were given personal assistance to clarify it. And all questionnaires were immediately verified using the Google occlusion tool, therefore no missing values in the study's items which related to variables of the study, because each question's response was eligible. Following the collection of data via the survey, the data was coded and labeled according to the various sections and item numbers of the questionnaire. The researcher then checked the data file for any missing information by entering the frequency of occurrence of each indicator into SPSS. The results of the descriptive analysis showed that there are no invalid or missing entries, thereby attesting to the respondents' full cooperation and the high level of accuracy of their answers. The appropriateness of the items, suitability of the questions, and choice of respondents all had an impact on these results. 4.4.2 Outliers The treatment of outliers is an essential step in the data screening process. Outliers are observations that have a distinct set of characteristics that distinguish them from the rest of the observations (Hair et al., 1998). According to Hawkins (1980), an outlier is an observation that differs so significantly from other observations that it raises questions about whether it was produced by a different mechanism. An extreme response from a participant to any or all questions is 176 considered an outlier (Hair et al., 2019). It might also be a distinct subcategory of the sample (Hair et al., 2019). Outliers were identified using univariate (histograms, box- plots and standardized z score) and multivariate detections (Mahalanobis D2 distance). 4.4.2.1 Univariate Outliers The term "univariate outliers" describes observations with a single variable's unusual value (Tabachnick & Fidell, 2007). In addition to looking at histograms and box plots, each variable's standardized (z) score was looked at for univariate detection (Tabachnick & Fidell 2007). A case is considered an outlier in accordance with Hair et al., (2006) if its standard score is ±3.0 or beyond. As a result, any Z-score that is either greater than 3 or lower than -3 is regarded as an outlier. Table 4.4 provides a summary of the standardized (z) scores for each item in each construct. Table 4.4: Result of Univariate Outlier Based on Standardized Values 1st Order Construct Item Standardized value (Z-Score) Lower Bound Upper Bound Ethics (ET) ET1 -2.607 1.434 ET2 -2.391 1.188 ET3 -2.297 1.278 ET4 -2.471 1.289 ET5 -2.432 1.361 ET6 -2.262 1.266 Independence (IN) IN1 -2.480 1.372 IN2 -2.386 1.252 IN3 -2.483 1.295 IN4 -2.439 1.310 IN5 -2.521 1.299 IN6 -2.316 1.319 Competency (CM) CM1 -2.494 1.188 CM2 -2.432 1.330 CM3 -2.520 1.530 CM4 -2.291 1.557 CM5 -2.296 1.604 CM6 -2.245 1.483 CM7 -2.340 1.511 Audit Fees (AF) AF1 -2.692 1.145 AF2 -2.638 1.249 Audit Firm Size (AFS) AFS1 -2.627 1.353 AFS2 -2.610 1.289 Internal Auditing (IA) IA1 -2.657 1.436 IA2 -2.639 1.288 177 Table 4.4, continued 1st Order Construct Item Standardized value (Z-Score) 1st Order Construct Lower Bound Upper Bound Accounting Basis (AB) AB1 -2.476 1.299 AB2 -2.708 1.421 AB3 -2.600 1.324 Laws and Regulation (LR) LR1 -2.605 1.383 LR2 -2.701 1.425 LR3 -2.436 1.233 Supreme Audit Institutions (SAI) SAI1 -1.673 1.188 SAI2 -1.768 1.364 SAI3 -1.798 1.372 SAI4 -1.759 1.202 SAI5 -1.672 1.142 SAI6 -1.665 1.189 SAI7 -1.721 1.472 SAI8 -1.744 1.212 SAI9 -1.661 1.247 SAI10 -1.721 1.349 SAI11 -2.324 1.040 Audit Quality (AQ) AQ1 -2.309 1.162 AQ2 -2.103 1.123 AQ3 -2.211 1.342 AQ4 -2.224 1.188 AQ5 -2.300 1.364 AQ6 -2.450 1.332 AQ7 -2.343 1.244 AQ8 -2.181 1.271 N = 186 Source: Smart PLS3 As can be seen in Table 4.4, the findings showed that the cases' standardized (z) scores for the research variables ranged from -2.708 to 1.604, meaning that none of the items' values exceeded the threshold of ±3.0. So none of the 186 cases contain a single univariate outlier. 4.4.2.2 Multivariate Outliers Since the variables in the current study were measured using a 5-point Likert scale, outliers were expected because some participants might have had an extreme or different opinion about a given question by selecting a response of 1 or 5. Thus, the Mahalanobis distance measure was employed in the current study to identify outliers. 178 The multivariate outliers have been successfully identified using Mahalanobis distance. To choose the best empirical values for the current study, the table of chi- square statistics was first applied. Two techniques exist to recognize outliers: (1) Based on the number of measurements in the questionnaire; (2) Based on the number of study variables. The results indicated that the most significant Mahalanobis value was 27.386 (belonged to case#28) significant at 0.01 level. No any cases having Mahalanobis value less than 27.368 was found in this study, indicating the absence of any multivariate outliers, according to (Kline, 2010). 4.4.3 Assessment of the Data Normality To ascertain whether the data for a variable are distributed according to a normal curve, the normality test was performed, either univariate or multivariate normality. 4.4.3.1 Univariate Normality Due to the existence of kurtosis variables, data with non-normal distribution would appear to either skew to the left or to the right (Brown, 2012), leading to misleading results regarding the relationships between the variables under study and the significance of these relationships. Skewness and kurtosis values are used to evaluate the univariate normality. The values of skewness and kurtosis should both fall within the range of ±2 and ±7, respectively (HO, 2006; Olsson et al., 2000; Oppenheim, 1966). The data seem to support this hypothesis with sufficient normality. The values for skewness and kurtosis for each item are summarized in Table 4.5. 179 Table 4.5: Assessment of Normality of All Items 1st Order Construct Item Skewness Std. Error of Skewness Kurtosis Std. Error of Kurtosis Ethics (ET) ET1 -0.666 0.178 0.013 0.355 ET2 -0.733 0.178 -0.219 0.355 ET3 -0.61 0.178 -0.279 0.355 ET4 -0.683 0.178 -0.002 0.355 ET5 -0.647 0.178 -0.077 0.355 ET6 -0.724 0.178 -0.265 0.355 Independence (IN) IN1 -0.657 0.178 0.056 0.355 IN2 -0.566 0.178 -0.279 0.355 IN3 -0.703 0.178 0.041 0.355 IN4 -0.553 0.178 -0.223 0.355 IN5 -0.719 0.178 0.035 0.355 IN6 -0.481 0.178 -0.415 0.355 Competency (CM) CM1 -0.73 0.178 -0.085 0.355 CM2 -0.582 0.178 -0.111 0.355 CM3 -0.599 0.178 0.083 0.355 CM4 -0.614 0.178 -0.003 0.355 CM5 -0.514 0.178 -0.041 0.355 CM6 -0.583 0.178 -0.062 0.355 CM7 -0.558 0.178 -0.177 0.355 Audit Fees (AF) AF1 -0.733 0.178 -0.056 0.355 AF2 -0.909 0.178 0.555 0.355 Audit Firm Size (AFS) AFS1 -0.583 0.178 -0.147 0.355 AFS2 -0.625 0.178 -0.067 0.355 Internal Auditing (IA) IA1 -0.521 0.178 -0.37 0.355 IA2 -0.77 0.178 0.098 0.355 Accounting Basis (AB) AB1 -0.412 0.178 -0.537 0.355 AB2 -0.841 0.178 0.486 0.355 AB3 -0.925 0.178 0.364 0.355 Laws and Regulations (LR) LR1 -0.915 0.178 0.507 0.355 LR2 -0.681 0.178 0.235 0.355 LR3 -0.747 0.178 -0.083 0.355 Supreme Audit Institutions (SAI) SAI1 -0.383 0.178 -1.215 0.355 SAI2 -0.48 0.178 -0.862 0.355 SAI3 -0.553 0.178 -0.905 0.355 SAI4 -0.393 0.178 -1.031 0.355 SAI5 -0.415 0.178 -1.212 0.355 SAI6 -0.35 0.178 -1.187 0.355 SAI7 -0.399 0.178 -0.97 0.355 SAI8 -0.401 0.178 -1.115 0.355 SAI9 -0.4 0.178 -1.135 0.355 SAI10 -0.34 0.178 -1.029 0.355 SAI11 -0.877 0.178 0.001 0.355 Audit Quality (AQ) AQ1 -0.829 0.178 -0.133 0.355 AQ2 -0.748 0.178 -0.478 0.355 AQ3 -0.685 0.178 -0.402 0.355 AQ4 -0.699 0.178 -0.404 0.355 AQ5 -0.544 0.178 -0.423 0.355 AQ6 -0.728 0.178 -0.123 0.355 AQ7 -0.772 0.178 -0.121 0.355 AQ8 -0.603 0.178 -0.428 0.355 N = 186 Source: Smart PLS3 180 The result demonstrates that all 50 items' skew and kurtosis fell between ±2 and ±7, respectively. Therefore, it can be said that a normal distribution accurately described the entire data set of the items. The skew ranged from -0.925 to -0.340, and the kurtosis ranged from -1.215 to 0.555, as shown in Table 4.5. 4.4.3.2 Multivariate Normality Mardia's procedures, which are regarded as a common test for multivariate normality in regard to skewness or kurtosis as suggested by Hair et al., (2017) and Cain et al., (2018) are used to analyze multivariate skewness and kurtosis, according to Mardia (1970) and Mardia (1974). In these procedures, it can be concluded that the data is not multivariate normal if the p-value of either multivariate skewness or kurtosis is lower than the significance level of 0.05, and thus suitable to use SmartPLS 3 using a 1,000-sample re-sample bootstrapping procedure (Hair et al., 2019; Ramayah et al., 2018). The following link provides suitable software to assess the multivariate skewness and kurtosis as suggested by Hair et al. (2017) and Ngah et al. (2020) (https://webpower.psychstat.org/models/kurtosis/results.php?url=c6c8ce84a2efb7ec83 569e241bed548a). The result of applying Mardia's multivariate normality on the collected data according to the table 4.6 and table 4.7 was β = 19.197 and p < 0.00001 for the multivariate skewness, and β = 188.074 and p = 0.01677 for the multivariate kurtosis, this is yielded the conclusion that the multivariate skewness was not normal due to the p-value being less than 0.05. Additionally, because the multivariate kurtosis had a p- value of less than 0.05. Accordingly, the collected data was not multivariate normal. 181 Table 4.6: Mardia's Multivariate Normality 1st Order Construct Skewness SE_skew Z_skew Kurtosis SE_kurt Z_kurt AB -0.943 0.178 -5.293 0.158 0.355 0.446 ACH -1.001 0.178 -5.621 0.140 0.355 0.395 AF -1.013 0.178 -5.684 0.563 0.355 1.589 AFA -1.103 0.178 -6.191 0.528 0.355 1.490 AFS -0.719 0.178 -4.035 -0.138 0.355 -0.391 AQ -1.070 0.178 -6.006 0.371 0.355 1.048 CM -0.960 0.178 -5.390 0.256 0.355 0.722 EMIC -1.156 0.178 -6.489 0.399 0.355 1.125 ET -0.820 0.178 -4.604 -0.123 0.355 -0.346 IA -0.754 0.178 -4.232 -0.074 0.355 -0.209 IN -0.846 0.178 -4.750 0.039 0.355 0.109 LR -0.971 0.178 -5.449 0.330 0.355 0.932 SAI -0.418 0.178 -2.349 -1.345 0.355 -3.794 Table 4.7: Mardia’s Multivariate Skewness and Kurtosis Skewness and Kurtosis b z p-value Skewness 19.19667 595.096915 0.00001 Kurtosis 188.07378 -2.391609 0.01677 Source: https://webpower.psychstat.org As a result, Smart PLS, a second generation non-parametric analysis software that can be used in this study to examine complicated models with latent variables and does not require normally distributed data (Hair et al., 2019). According the suggestions of (Hair et al., 2019), the structural model's path coefficients, standard deviation, t-values, and p-values were reported using a 1,000-sample re-sample bootstrapping procedure (Hair et al., 2019; Ramayah et al., 2018). 4.5 Common Method Bias (Harman’s single-factor test) Common method bias, which is described as variance due to the measurement technique rather than the constructs the measure represents, may pose a problem in behavioral studies (Podsakoff et al., 2012). The phenomdescribes a bias in the dataset brought on by a factor independent of the measurements. It's possible that something 182 unrelated to the question had an effect on the answer. This study's data collection method, an online questionnaire survey using Google Form, may have introduced systematic response bias, which could have impacted or inflated responses. As this study used a one-wave self-reported design, in which all the data for all the variables were collected at the same time, Harman's single-factor test (Hoyle, 1995) was used to determine whether common method variance was a significant issue. The results of Harman's single factor test suggested that common method variance was not a major problem because one factor model explained 46.48% of the total variance, which was below 50% (Hoyle, 1995). Harman's single-factor test's results are shown in Appendix 5. 4.6 Measurement Model (Confirmatory Factor Analysis) – Stage 1 of SEM To determine the relationships between manifest or observed and latent or unobserved variables, the measurement model or confirmatory factor analysis (CFA) is used. Therefore, it could be said that the measurement model specifies how latent or unobserved variables are evaluated in relation to the manifest variables (HO, 2006). The process of ensuring accuracy includes the operationalization of constructs, which is a crucial step (Hair et al., 2006). In an effort to ensure theoretical accuracy, researchers can choose from a number of recognized scales. Although there are many different scales available, researchers are frequently constrained by the problem of a lack of well-established scales, which forces them to either create new measurement scales from scratch or significantly modify existing scales to fit a new context. Given all of these factors, the selection of items to measure the constructs serves as the foundation for the SEM analysis (Hair et al., 2006). 183 Each of the constructs in the CFA models had its reliability and validity evaluated. Cronbach's alpha, construct reliability (CR), and average variance extracted (AVE) are used to measure reliability, while constructs, including convergent and discriminant functions, are used to measure validity. This research included two overall measurement models, as well as the two research structural models depicted in Section 3.4. The following subsections go over the evolution of each measurement model. The results of testing the unidimensionality of each construct using SmartPLS 3 are presented. 4.6.1 Measurement Model 1 Confirmatory factor analysis was used to assess the overall measurement model 1. The overall measurement model 1 including all latent constructs with their indicators was portrayed by Smart PLS3 as the following figure 4.1 of the Initial Measurement Model 1 before omitting the item SAI11. 184 Source: Smart PLS3 Figure 4.1: The Initial Measurement Model 1 4.6.1.1 Convergent Validity and Reliability Table 4.8 represents the result of convergent validity and Cronbach alpha for the measurement model 1after omitting item SAI11 as was portrayed by Smart PLS3 in appendix 7. 185 Table 4.8: Convergent Validity and Cronbach Alpha for Measurement Model 1 Construct Item / 1st Order Construct Factor Loading Average Variance Extracted (AVE)a Composite Reliability (CR)b Internal Reliability Cronbach Alpha 1st Order Constructs Ethics (ET) ET1 0.882 0.790 0.957 0.947 ET2 0.894 ET3 0.884 ET4 0.902 ET5 0.882 ET6 0.888 Independence (IN) IN1 0.867 0.781 0.955 0.944 IN2 0.881 IN3 0.887 IN4 0.908 IN5 0.878 IN6 0.884 Competency (CM) CM1 0.874 0.727 0.949 0.937 CM2 0.861 CM3 0.842 CM4 0.861 CM5 0.843 CM6 0.828 CM7 0.857 Audit Fees (AF) AF1 0.930 0.860 0.925 0.837 AF2 0.924 Audit Firm Size (AFS) AFS1 0.932 0.873 0.932 0.855 AFS2 0.937 Internal Auditing (IA) IA1 0.951 0.903 0.949 0.893 IA2 0.950 Accounting Basis (AB) AB1 0.883 0.814 0.929 0.885 AB2 0.913 AB3 0.910 Laws and Regulation (LR) LR1 0.918 0.834 0.938 0.900 LR2 0.930 LR3 0.892 Supreme Audit Institutions (SAI) SAI1 0.780 0.739 0.966 0.969 SAI2 0.804 SAI3 0.907 SAI4 0.867 SAI5 0.926 SAI6 0.902 SAI7 0.880 SAI8 0.825 SAI9 0.833 SAI10 0.864 SAI11 0.347c Audit Quality (AQ) AQ1 0.836 0.717 0.953 0.944 AQ2 0.837 AQ3 0.817 AQ4 0.885 AQ5 0.870 AQ6 0.820 AQ7 0.847 AQ8 0.859 186 Table 4.8, continued Construct Item / 1st Order Construct Factor Loading Average Variance Extracted (AVE)a Composite Reliability (CR)b Internal Reliability Cronbach Alpha 2nd Order Constructs Auditor Characteristics (ACH) Ethics (ET) 0.945 0.875 0.954 0.928 Independence (IN) 0.939 Competency (CM) 0.921 Audit Firm Attributes (AFA) Audit Fees (AF) 0.926 0.858 0.923 0.834 Audit Firm Size (AFS) 0.926 Effectiveness of the Municipal Internal Control (EMIC) Internal Auditing (IA) 0.906 0.843 0.942 0.907 Accounting Basis (AB) 0.933 Laws and Regulation (LR) 0.915 a: Average Variance Extracted = (summation of the square of the factor loadings)/{(summation of the square of the factor loadings) + (summation of the error variances)}. b: Composite reliability = (square of the summation of the factor loadings)/{(square of the summation of the factor loadings) + (square of the summation of the error variances)}. c: denotes an item that was discarded because it didn't have enough factor loading to meet the cutoff of 0.6. Source: Smart PLS3 The initial standardized factor loading of the SAI11 was 0.347, below the cut- off 0.6, as shown in Table 4.8 analysis of the standardized factor loadings of the model's items. Therefore, as advised by Hair et al. (2006) this item was taken off the model. Compared to the overall number of items in the constructs, the number of deleted items was not significant. Furthermore, the removal had little effect on the conceptualization of the constructs' content. The remaining 49 items and 8 first order constructs all had standardized factor loadings above 0.6, ranging from 0.780 (for SAI1) to 0.951 (for IA1). Each of the constructs was evaluated for reliability after the unidimensionality of the constructs was achieved. Average variance extracted (AVE), construct reliability (CR), and Cronbach's alpha are used to evaluate reliability. According to Hair et al., (2006), the cut-off value for first and second order constructs is 0.5. Table 4-8 demonstrate that the AVE values, which reflect the overall amount of variance in 187 the indicators accounted for by the latent construct, were above this cutoff and ranged between 0.717 (for Audit Quality (AQ)) and 0.903 (for Internal Auditing (IA)). The composite reliability values, which show how well the construct indicators predict the latent construct, were higher than Bagozzi and Yi (1988) recommended value of 0.6 for all first and second order constructs, ranging from 0.923 for the Audit Firm Attributes (AFA) to 0.966 for the Supreme Audit Institutions (SAI). According to Nunnally and Bernstein, (1994), the Cronbach's Alpha values, which indicate how error-free a measure is, were higher than the cut-off point of 0.7 for all first and second order constructs. These values ranged from 0.834 for the Audit Firm Attributes (AFA) to 0.969 for the Supreme Audit Institutions (SAI). 4.6.1.2 Discriminant Validity A construct's discriminant validity describes how it differs from other constructs based on the correlation and square root of AVE values that were determined. It indicates sufficient discriminant validity when the square root of AVE for both constructs is greater than the correlation between the two constructs (Fornell and Larcker 1981; Hair et al., 2006) As shown in Appendix 6, who represents the results of cross loadings of the indicators to assess the discriminant validity of all Items and 1st order constructs. The cross loadings of the indicators specified that an indicator's outer loading on the associated construct was greater than all of its loadings on other constructs on each item row. These results demonstrated no any discriminant validity problem (Hair et al., 2011). 188 4.6.1.2.1 Fornell-Larcker Criterion The results of the Fornell-Larcker criterion to assess the discriminant validity of the measurement model are shown in Table 4.9. Table 4.9: Results of Fornell-Larcker Criterion in Measurement Model 1 ACH AFA EMIC SAI AQ Auditor Characteristics (ACH) 0.935 Audit Firm Attributes (AFA) 0.704 0.926 Effectiveness of the Municipal Internal Control (EMIC) 0.755 0.762 0.918 Supreme Audit Institutions (SAI) 0.056 -0.012 0.062 0.860 Audit Quality (AQ) 0.842 0.774 0.831 0.096 0.847 Note: The diagonals represent the square root of the average variance extracted, while the other entries represent correlations. Source: SmartPLS 3 The inter-correlations between the five hypothesized latent constructs in measurement model 1 ranged from -0.012 to 0.842, as shown in Table 4.9, falling short of the cut-off of 0.85 (Kline, 2005). The analysis also revealed, as shown in Table 4.9, that the value of the off-diagonal elements was lower than the value of the AVE square root. Thus, it demonstrates that each latent construct measurement was completely discriminatory with respect to one another based on the Fornell-Larcker approach (Fornell and Larcker 1981; Hair et al., 2014). 4.6.1.2.2 HTMT Discriminant Criteria The findings of the HTMT discriminant criteria used to evaluate the measurement model 1's discriminant validity are shown in Table 4.10. 189 Table 4.10: Results of HTMT Discriminant Criteria in Measurement Model 1 ACH AFA EMIC SAI AQ Auditor Characteristics (ACH) Audit Firm Attributes (AFA) 0.800 Effectiveness of the Municipal Internal Control (EMIC) 0.823 0.876 Supreme Audit Institutions (SAI) 0.073 0.068 0.042 Audit Quality (AQ) 0.899 0.872 0.898 0.067 Source: Smart PLS3 All of the HTMT values between the five hypothesized latent constructs in measurement model 1 were below 0.90, ranging from 0.042 to 0.899, as shown in Table 4.10. Thus, it demonstrates that each latent construct measurement was completely discriminatory with respect to one another (Henseler et al., 2015). After looking at the measurement model 1's convergent validity and discriminant validity, it can be said that the modified measurement model 1 is valid and reliable for evaluating the constructs, their related items, and sub-constructs. The modified measurement model 1 is shown in Appendix 7 with uniform factor loadings for all latent constructs and related items. 4.6.2 Measurement Model 2 The overall measurement model 2 was evaluated using confirmatory factor analysis. 4.6.2.1 Reliability and Convergent Validity All of the constructs in measurement model 2 have already been examined in measurement model 1 for standardized factor loading, Cronbach alpha, and convergent validity as shown in Table 4.8. 190 4.6.2.2 Discriminant Validity Fornell-Larcker Criterion and HTMT Discriminant Criteria are used to evaluate the validity of the measurement model 2. 4.6.2.2.1 Fornell-Larcker Criterion The findings of the Fornell-Larcker criterion used to evaluate the measurement model 2's discriminant validity are shown in Table 4.11. Table 4.11: Fornell-Larcker Criterion in Measurement Model 2 AB AF AFS AQ CM ET IA IN LR AB 0.902 AF 0.698 0.927 AFS 0.691 0.715 0.935 AQ 0.798 0.721 0.711 0.847 CM 0.715 0.647 0.633 0.797 0.852 ET 0.669 0.618 0.594 0.780 0.803 0.889 IA 0.770 0.597 0.648 0.735 0.623 0.577 0.950 IN 0.681 0.597 0.572 0.787 0.786 0.849 0.629 0.884 LR 0.793 0.635 0.610 0.757 0.675 0.620 0.728 0.648 0.913 Note: Diagonals represent the square root of the average variance extracted while the other entries represent the correlations Source: Smart PLS3 The inter-correlations between the nine hypothesized latent constructs in measurement model 2 ranged from 0.572 to 0.849, as shown in Table 4.11, falling below the cut-off of 0.85 (Kline, 2005). The analysis also revealed, as shown in Table 4.11, that the value of the off-diagonal elements was lower than the value of the AVE square root. This demonstrates that each latent construct measurement was completely discriminatory to each order based on the Fornell-Larcker approach (Fornell and Larcker, 1981; Hair et al., 2014). 191 4.6.2.2.2 HTMT Discriminant Criteria The findings of the HTMT discriminant criteria used to evaluate the measurement model 2's discriminant validity are shown in Table 4.12. Table 4.12: HTMT Discriminant Criteria in Measurement Model 2 AB AF AFS AQ CM ET IA IN LR AB AF 0.811 AFS 0.797 0.845 AQ 0.871 0.812 0.792 CM 0.783 0.729 0.706 0.845 ET 0.730 0.694 0.659 0.825 0.851 IA 0.867 0.690 0.741 0.800 0.681 0.627 IN 0.745 0.672 0.637 0.833 0.835 0.898 0.685 LR 0.888 0.732 0.695 0.821 0.733 0.671 0.812 0.702 Source: Smart PLS3 All of the HTMT values between the nine hypothesized latent constructs in measurement model 2 were below 0.90, ranging from 0.627 to 0.898, as shown in Table 4.12. Thus, it demonstrates that each latent construct measurement was completely discriminatory with respect to one another (Henseler et al., 2015). After analysing the convergent validity and discriminant validity of the measurement model 2, it can be said that the modified measurement 2 is a valid and reliable method for evaluating the constructs, their related items, and sub-constructs. The modified measurement model 2 is shown in Figure 4.2 with uniform factor loadings for all latent constructs and associated items. 192 Source: Smart PLS3 Figure 4.2: Measurement and Structural Model 2 4.7 Descriptive Analysis To account for all of the variables in this analysis, the descriptive function was computed using the covariance matrix method. The variables' composite scores were calculated by parcelling the original measurement item scores. Parcels are summation or averages of several individual indicators or items based on their factor loadings on 193 the construct (Coffman & Maccallum 2005; Hair et al., 2006). Table 4.13 displays the mean and standard deviation of the constructs, assessed on a 5-point Likert scale: Table 4.13: Results of Descriptive Statistic for Variables Constructs Mean Standard Deviation Minimum Maximum Auditor Characteristics (ACH) 3.560 0.872 1.365 4.746  Ethics (ET) 3.597 0.959 1.167 5  Independence (IN) 3.603 0.945 1 5  Competency (CM) 3.480 0.891 1.143 5 Audit Firm Attributes (AFA) 3.710 0.884 1.25 4.75  Audit Fees (AF) 3.761 0.960 1 5  Audit Firm Size (AFS) 3.659 0.949 1 5 Effectiveness of the Municipal Internal Control (EMIC) 3.635 0.856 1.222 4.889  Internal Auditing (IA) 3.642 0.949 1 5  Accounting Basis (AB) 3.633 0.916 1 5  Laws and Regulations (LR) 3.629 0.932 1 5 Supreme Audit Institutions (SAI) 3.299 1.184 1 4.9 Audit Quality (AQ) 3.576 0.965 1.125 4.75 N = 186 Source: Smart PLS3 As a measure of central tendency, the mean was used, and it showed that all constructs' mean values were higher than the midpoint of 3 on a 5-point Likert scale. The phenomenon showed that the consensus respondents had a more favorable perception toward these variables were above the average. Audit Fees (AF), which had the highest mean score of (3.761), was followed by Audit Firm Attributes (AFA) (3.71), and Audit Firm Size (AFS) (3.659). Supreme Audit Institutions (SAI) had the lowest mean rating, with a mean score of (3.299). The standard deviation was used as a dispersion index to show how much deviations within each variable are from the mean of the variable. The Supreme Audit Institution's (SAI) individual value deviated from the mean the most of any of the variables under study (SD = 1.184). The standard deviation indicated that respondents' perceptions of the Supreme Audit Institutions (SAI) varied somewhat. In other words, 194 the survey respondents' responses to this variable varied the most from one another. On the other hand, Effectiveness of Municipal Internal Control (EMIC), with a standard deviation of 0.856, had the lowest deviation from the mean. The mean of all constructs and their standard deviations are well represented in Figure 4.3 along with their respective ranges. Source: Excel Figure 4.3: Means and Standard Variations of All Constructs 4.7.1 Descriptive Analysis of Auditor Characteristics (ACH) Items Table 4.14 shows the mean, standard deviation, minimum and maximum of all items on ACH. The obtained mean values exceeded the three-point mark (above average), ranging from 3.35 (CM5) to 3.71 (CM1). Furthermore, ET6 was found to have the highest deviation from its mean value (SD = 1.134), indicating that the responses obtained from respondents for ET6 varied the most from one another, whereas CM3 recorded the lowest deviation from its mean value (SD = 0.988). 195 Table 4.14: Results of Descriptive Statistic for the Items of ACH Constructs Constructs Code Mean Standard Deviation Minimum Maximum Auditor Characteristics ACH 3.560 0.872 1.365 4.746 Ethics ET 3.597 0.959 1.167 5 The overall reputation of the audit firm is positive. ET1 3.58 0.99 1 5 The audit team members as a group always exercise due care throughout the engagement. ET2 3.67 1.117 1 5 The audit firm has strict guidelines on the procedures that must be completed before signing the audit report. ET3 3.57 1.119 1 5 The audit firm actively encourages staff members to take courses and attend seminars in fields where the firm has major clients. ET4 3.63 1.064 1 5 The senior auditors supervise junior audit staff. ET5 3.56 1.055 1 5 The engagement auditors maintain high ethical standards. ET6 3.56 1.134 1 5 Independence (IN) IN 3.603 0.945 1 5 The audit firm has a skeptic's mindset, not a client advocate's mindset. IN1 3.58 1.038 1 5 The audit fee is less than 10% of the total revenue of the audit firm. IN2 3.62 1.1 1 5 The audit firm and individual audit team members never participate in any conduct that might undermine its/their independence, either in fact or in appearance, in any of your contact with them. IN3 3.63 1.059 1 5 The audit firm performing the audit does not provide consultancy services to the municipality. IN4 3.6 1.067 1 5 The audit firm has a high audit staff turnover rate. IN5 3.64 1.047 1 5 Members of the audit team are cycled off the audit on a regular basis. IN6 3.55 1.101 1 5 Competency CM 3.480 0.891 1.143 5 The audit team assigned to the audit engagement (partner, manager, and supervisor) is well educated on local government units. CM1 3.71 1.086 1 5 Other municipalities are audit clients of the auditor that is conducting the audit. CM2 3.59 1.063 1 5 The auditors assigned to the engagement have extensive understanding of accounting and auditing standards, as well as professional certifications such as the CPA. CM3 3.49 0.988 1 5 196 Table 4.14, continued Constructs Code Mean Standard Deviation Minimum Maximum The audit team members as a whole have a good understanding of the municipality's operations. CM4 3.38 1.039 1 5 In completing the audit, the audit company makes considerable use of computers and statistical methodologies. CM5 3.35 1.026 1 5 Each audit area has a strict time budget that the audit firm wants its auditors to stick to. CM6 3.41 1.073 1 5 The total number of hours spent on the audit by the audit team (from the beginning of field work to the audit report date). CM7 3.43 1.039 1 5 N = 186 Source: Smart PLS3 According to the results of Table 4.14, the majority of respondents believe that auditor characteristics (ethics, independence, and competence) influence audit quality and that these characteristics can determine audit quality in the municipalities. 4.7.2 Descriptive Analysis for of Audit Firm Attributes (AFA) Items Table 4.15 shows the mean, standard deviation, minimum and maximum of AFA. Table 4.15: Descriptive Statistic for the Items of AFA Constructs Constructs Code Mean Standard Deviation Minimum Maximum Audit Firm Attributes AFA 3.710 0.884 1.25 4.75 Audit Fees AF 3.761 0.960 1 5 The amount of audit fees that is paid AF1 3.81 1.042 1 5 The amount of audit fees is related to the efforts of the auditors in the audit engagement. AF2 3.72 1.029 1 5 Audit Firm Size AFS 3.659 0.949 1 5 The suitable number of professionals in the audit team to achieve audit quality AFS1 3.64 1.005 1 5 The legal form of the audit firm and its size affect audit quality AFS2 3.68 1.026 1 5 N = 186 Source: Smart PLS3 197 Table 4.15 presents the mean and standard deviation of all items on AFA. The obtained mean values exceeded the three-point mark (above average), ranging from 3.64 (AFS1) to 3.81 (AF1). Furthermore, AF1 was found to have the highest deviation (SD = 1.042) from its mean value, indicating that the responses obtained from respondents for AF1 varied the most from one another, whereas AFS1 recorded the lowest deviation (SD=1.005) from its mean value. According to the results of Table 4.15, the majority of respondents believe that audit firm attributes (audit fees, audit firm size) influence audit quality and that these attributes can determine audit quality in the municipalities. 4.7.3 Descriptive Analysis of Effectiveness of Municipal Internal Control (EMIC) Items Table 4.16 shows the mean, standard deviation, minimum and maximum of EMIC. It presents the mean and standard deviation of all items on EMIC. The obtained mean values exceeded the three-point mark (above average), ranging from 3.60 (IA1) to 3.69 (IA2). Furthermore, LR3 was found to have the highest deviation (SD = 1.09) from its mean value, indicating that the responses obtained from respondents for LR3 varied the most from one another, whereas AB recorded the lowest deviation (SD=1.005) from its mean value. According to the results of Table 4.16, the majority of respondents believe that effectiveness of municipal internal control (internal auditing, accounting basis, laws and regulations) influence audit quality and that these attributes can determine audit quality in the municipalities. 198 Table 4.16: Results of Descriptive Statistic for the Items of EMIC Constructs Constructs Code Mean Standard Deviation Minimum Maximum Effectiveness of the Municipal Internal Control EMIC 3.635 0.856 1.222 4.889 Internal Auditing IA 3.642 0.949 1 5 The nature and type of the internal audit function in the municipality. IA1 3.6 0.977 1 5 External auditors work closely with internal auditors. IA2 3.69 1.019 1 5 Accounting Basis AB 3.633 0.916 1 5 The accounting basis used in the municipality’s accounting system. AB1 3.62 1.059 1 5 The transition from cash basis to accrual basis improves the relevance and reliability of the financial statements. AB2 3.62 0.969 1 5 Accrual basis requires the auditor to increase his efforts in the auditing process. AB3 3.65 1.019 1 5 Laws and Regulations LR 3.629 0.932 1 5 The existence of appropriate laws and regulations increases the audit quality. LR1 3.61 1.003 1 5 The commitment of the client to the laws and regulations enhances audit quality. LR2 3.62 0.97 1 5 The commitment of the auditors with the investigation of client’s adherence with applicable laws and regulation increases audit quality. LR3 3.66 1.09 1 5 N = 186 Source: Smart PLS3 4.7.4 Descriptive Analysis of Supreme Audit Institutions (SAI) Items Table 4.17 shows the mean, standard deviation, minimum and maximum of SAI. It presents the mean and standard deviation of all items on SAI. The obtained mean values exceeded the three-point mark (above average), ranging from 3.16 (SAI7) to 3.76 (SAI11). Furthermore, SAI5 was found to have the highest deviation (SD = 1.421) from its mean value, indicating that the responses obtained from respondents for SAI5 varied the most from one another, whereas SAI11 recorded the lowest deviation (SD=1.189) from its mean value. According to the results of Table 4.17, the majority of respondents believe that Supreme Audit Institutions audit influence the relationship between audit quality and its determinants of auditor characteristics, audit firm attributes, and effectiveness of municipal internal control. 199 Table 4.17: Results of Descriptive Statistic for the Items of SAI Constructs Constructs Code Mean Standard Deviation Minimum Maximum Supreme Audit Institutions SAI 3.299 1.184 1 4.9 The SAIs and choosing of a good reputation auditor with a high professional ethics SAI1 3.34 1.398 1 5 The SAIs and choosing of an independent auditor either in his mind and appearance SAI2 3.26 1.277 1 5 The SAIs and choosing of a high professional competence auditor SAI3 3.27 1.262 1 5 The SAIs and choosing of a highly qualified and professional audit team. SAI4 3.38 1.351 1 5 The SAIs and choosing of an audit firm whose audit fees are reasonable and fair. SAI5 3.38 1.421 1 5 The SAIs and choosing of a large-size audit firm such as the Big 4 SAI6 3.33 1.401 1 5 The SAIs and establishing an internal audit unit in the municipality, and works to increase its efficiency and effectiveness SAI7 3.16 1.253 1 5 The SAIs audit affects the municipal administration in order to adopt the accrual basis of accounting. SAI8 3.36 1.353 1 5 The SAIs and complying with the applicable laws and regulations. SAI9 3.28 1.375 1 5 The audit team always relies on the reports and findings of the SAIs audit in the audit engagement process. SAI10 3.24 1.303 1 5 The SAIs audit supports and increases the quality of the external audit in general. SAI11 3.76 1.189 1 5 N = 186 Source: Smart PLS3 4.7.5 Descriptive Analysis of Audit Quality (AQ) Items Table 4.18 shows the mean, standard deviation, minimum and maximum of AQ. Table 4.18 presents the mean and standard deviation of all items on AQ. The obtained mean values exceeded the three-point mark (above average), ranging from 3.49 (AQ3) to 3.66 (AQ1). Furthermore, AQ2 was found to have the highest deviation (SD = 1.24) from its mean value, indicating that the responses obtained from respondents for AQ2 varied the most from one another, whereas AQ6 recorded the lowest deviation (SD=1.085) from its mean value. According to the results of Table 4.18, the majority of respondents believe that audit quality will be achieved if the auditors detect and report the deficiencies, advise 200 the municipal management with new accounting standards, and satisfy the audit committee through effective communication. Table 4.18: Results of Descriptive Statistic for the Items of AQ Constructs Constructs Code Mean Standard Deviation Minimum Maximum Audit Quality AQ 3.576 0.965 1.125 4.75 Audit quality detects and reports the material errors and fraud in the client’s financial statements. AQ1 3.66 1.152 1 5 Audit quality detects and reports the material weakness of the internal control system. AQ2 3.61 1.24 1 5 The audit firm agrees to complete the audit by a deadline stipulated by the client. AQ3 3.49 1.126 1 5 The audit team and the audit committee of the council communicate often. AQ4 3.61 1.173 1 5 There is a communication between the audit team and the council's management. AQ5 3.51 1.092 1 5 Throughout the year, the audit firm keeps the council management informed about accounting and financial reporting developments that have an impact on the council. AQ6 3.59 1.058 1 5 During the audit, the audit engagement partner and manager conduct numerous visits to the council. AQ7 3.61 1.115 1 5 The auditor adds benefits to the municipality by generating useful improvement ideas. AQ8 3.53 1.159 1 5 N = 186 Source: Smart PLS3 4.8 Reports of SAIs in Palestine Financial and Administrative Control Bureau (FACB) and MOLG - GDGC issue annual, interim, and specialized reports. The reports that SAIs produce, the effects they have on society, and their capacity to fight corruption, protect public finances, and less an abuse of public office all have an impact on how strong and effective they are. The degree to which the recommendations in these reports are carried out as soon as possible will determine how well SAIs work. It is also evaluated based on the Legislature's capacity to act on recommendations and comments made in reports. The study summarized the SAIs reports to demonstrate the influence of these 201 reports on the study variables, as well as to support the study's data analysis in the effect of the SAIs as a moderator variable between audit quality and its factors. 4.8.1 FACB Reports Annual reports of FACB issued semi-regularly in the period from 2006 to 2021, and the interim and the specialized reports which were issued sometimes in this period, these reports are available at FACB’s (old name SAACB) web site https://www.saacb.ps/BruRptsTestSAACB/IndexRPTArabic). The FACB wants to make audit findings available to decision-makers and stakeholders because doing so will encourage an audit culture at audited institutions and result in more recommendations being followed through on. This will facilitate the use of preventative measures (FACB, 2014). The FACB has posted its fifteenth report online as evidence of compliance with the requirements of FACB Law no. 15 for 2004 since the publication of the FACB's reports which began in 2006 (FACB, 2020). Prior to 2011, these reports were not distributed on a regular basis or in a consistent format. Following that, the reports became more regular in format and subject matter, and they were issued on an annual basis except report of 2013. All FACB reports either annual report or interim reports on LGUs were examined and summarized in the Table 4.19 and Table 4.20. These tables show how the FACB influenced audit quality in the municipalities and the selected determinants of audit quality, auditor characteristics, audit firm attributes, and effectiveness of internal control, through its notes and recommendations, as well as the procedures implemented to address violations of laws and regulations and strengthen internal controls in LGUs. 202 Table 4.19: General Data of FACB Audit Reports The years/ Items 2011 2012 2014 2015 2016 2017 2018 2019 2020 Average Total FACB reports 156 123 104 118 119 123 139 125 115 125 Audit report related to LGUs 55 50 25 37 43 70 71 63 60 53 LGUs reports % 35% 41% 24% 31% 36% 57% 51% 50% 52% 42% The responds rate to audit reports 70% 66% 72% 71% 79% 63% 79% 81% 72% 73% Municipality Audited 7 35 12 16 17 20 25 21 17 19 Percentage of municipalities to audit reports for LGU 13% 70% 48% 43% 40% 29% 35% 33% 28% 38% Complaints received 306 267 352 360 485 412 360 319 174 337 Complaints of LGUs * 28 64 111 106 143 152 95 76 97 Percentage of Complaints of LGUs * 10% 18% 31% 22% 35% 42% 30% 44% 29% Complaints of Municipalities * * * * 51 20 19 16 16 24 Percentage of municipalities complaints to LGUs * * * * 48% 14% 13% 17% 21% 22% Orders of Anti- Corruption Commission-ACC transferred to FACB for auditing * * * * 120 65 125 52 36 80 Cases are transferred to ACC 33 29 37 27 24 * * * * 30 LGU cases transferred to ACC 13 17 * 11 14 23 11 26 19 Financial impact LGU in thousand USD 12,379 3,922 689 895 * * 19,432 * * 7,464 Total Financial impact in thousand USD 22,624 7,828 4,441 20,316 * * * * * 13,803 Attendance of tenders 414 550 676 * 368 * * * * 402 Source: Author According to Table 4.19, the average percentage of audit reports from local government units was 42%, while the average percentage of complaints was 29% of total reports. This demonstrates the importance of local government units in FACB auditing and the importance of this type of audit through the high percentage of responses to FACB reports, which average was 73%. Furthermore, the cooperation between ACC and FACB in dealing with corruption cases increased the importance of FACB auditing, because municipal councils recognized that compliance with laws, 203 regulations, and regulator recommendations is critical in order to avoid punishments and fines. Furthermore, the employees of FACB attend the bidding meetings in order to control the tendering policies and procedures for public sector organizations, particularly municipalities including the external audit bid for hiring the external auditors. The average number of bidding meetings was 402. The financial effect of the cases under audit is sometimes shown in FACB audit reports; for example, the financial effect in year 2018 was 19,432,978 USD, but the average was 7,463,627 USD. Most of FACB auditing reports related to compliance auditing, and few of them related to financial statements auditing. Table 4.20 summarizes general auditing notes and recommendations which they appeared in the annual reports of FACB since 2006 and related with the audit quality and its selected determinants. Table 4.20: General Notes and Recommendations of FACB on AQ # The Notes and the Recommendations The Audit Quality Attributes 1 Employees in certain local governments abused their authority and misappropriated funds. Weakness of internal auditing 2 Occasionally, spending can be done without all the required paperwork and necessary documents. Weakness of internal auditing 3 Violation of the provisions of the Building and Organization Code 1996 for Local Authorities in terms of licensing fees, violation fees, and granting discounts. Failure to comply with the laws and regulations 4 Some local governments did not put the code of conduct for local government employees into effect. Failure to comply with the laws and regulations 5 When hiring new employees, some municipal governments do not always adhere to conceptual knowledge and legal procedures. Failure to comply with the laws and regulations 6 Violation of code provisions for supplies and project implementation at local governments in terms of supplying, executing, or servicing. Failure to comply with the laws and regulations 7 Some local government entities have a lack of internal control and a robust internal control system that protects assets. Weakness of internal control 8 Violation of laws, regulations, and ordinances governing budgeting and revenue/expense measurement. Failure to comply with the laws and regulations 204 Table 4.20, continued # The Notes and the Recommendations The Audit Quality Attributes 9 Some municipalities do not have external auditors. Laws and Regulations. Public Interest Theory for auditing. 10 Local government accounts are untrustworthy and raise concerns about accuracy, authenticity, and occurrence due to a lack of corroborating documents and a governing documentation cycle. Accounting Basis 11 Some local government entities failed to collect fees mandated by applicable laws and regulations, particularly fees for billboards, crafts, and industries. Failure to comply with the laws and regulations 12 Accounting software might not meet all the requirements of local governments because it does not assign user rights or incorporate actions, making financial statements susceptible to loss, damage, and deletion as well as casting doubt on their objectivity and fairness. Weakness in accounting Information System and Accounting basis 13 The Municipality did not follow laws and decisions regarding its participation in licensed electricity distribution companies. Failure to comply with the laws and regulations 14 The Municipality failed to comply with the Council of Ministers' 2017 electricity tariff. Failure to comply with the laws and regulation 15 Internal supervision and audit system weakness. Audit quality 16 A flaw in the financial system's application. Accounting basis 17 Insufficient promises made by the financial system to local governments in terms of spending, budget planning, and document reinforcement Weakness in Internal Auditing 18 Failure to prepare financial statements in accordance with regulations and legislation. Accounting Basis and violation of laws and regulations 19 Weakness in audit regulations that govern spending, resulting in a lack of a tight internal control system. Internal control and internal auditing 20 Some local governments may fail to perform proper bank reconciliations in order to keep track of their bank accounts. Weakness in Internal Auditing 21 Many local governments fail to manage public finances due to a lack of control systems and a division of powers, resulting in cases of misappropriation, credit misuse, and public funds theft. Weakness in the internal control system 22 The municipality violated international accounting rules by failing to disclose the accounting policies used to record the grant in the financial statements and failing to describe the nature of the grant. Accounting Basis 23 Despite the accounting accrual concept, waste charges from previous years were recorded in the current books. Accounting Basis 24 As required by accrual accounting, the municipality did not record expenses and allowance for doubtful debts for current books. Accounting Basis Source: Author The majority of the preceding notes and recommendations are concerned with the internal control system and its dimensions: internal auditing, accounting basis and compliance with applicable laws and regulations. 205 These notes and recommendations serve as a warning to the management of all local government units to avoid them and improve their internal controls and accounting information systems. This is leading to an improvement in external audit quality by producing high-quality financial statements that are used as input in the external auditing process. However, the FACB audit reports notes and recommendations do not specifically and clearly mention the other inputs of the auditing process, such as the auditor characteristics (auditors' ethics, competency, and independence) and the audit firm attributes (audit fees and audit firm size), but the FACB auditors may take into account the audit firm's attributes and the auditor's characteristics as specified in the Palestinian government's auditing standards and the MOLG approved guidelines of ToR for hiring external auditors in the LGUs. 4.8.2 MOLG-GDCG Reports MOLG prepares periodic (annual or semi-annual) reports on local government units (LGUs) through GDCG auditors, but these reports remain confidential and are not available to the public. The researcher obtained some of them for the study through personal contact with some municipalities. The audit report is a semi- structured document with many questions pertaining to the audit scope, including the reviewing of the financial aspects such as cash balances, debts, inventories, checks, accounting records of revenues and expenses. Also, this type of auditing includes the examination of the budget process including the compliance with stated expenditures and revenues amounts as appeared in the budget, internal controls and procedures in the accounting system, external audit reports, FACB audit reports, and the compliance with the applicable laws and regulations. 206 In comparison to the notes and recommendations of the FACB, the GDCG's are more precise, thorough, and detailed. The majority notes of this type of auditing referring to violations of the related laws, rules, policies, and MOLG directives, as a result, the focus of this audit is on operational and compliance audits, including internal control system audits. Table 4.21 shows the summary of the most notes and the recommendations of the GDCG. Table 4.21: The Notes and the Recommendations of MOLG GDCG # The Notes and the Recommendations The Audit Quality Attributes 1 Payment vouchers may be issued in the absence of all necessary paperwork, official authority approvals, beneficiary signatures, dates, and other data. Weakness of internal auditing 2 Laws, regulations, and ordinances governing budgeting and revenue/expense measurement are being broken. Failure to comply with the laws and regulations 3 Some municipalities use Excel to keep track of paper records instead of appropriate accounting software because they lack the necessary internal controls. Internal control over accounting system 4 Due to an insufficiency of supporting documentation and a cycle for governing documentation, local government entities' accounts are unreliable and raise questions about their accuracy, authenticity, and occurrence. Accounting Basis 5 Violation of some storekeeping procedures, particularly complete records, physical counting, and item evaluation and organization. Failure to comply with the laws and regulations. And weakness of internal auditing 6 Accounting software that does not incorporate actions or assign user authorities may not meet the needs of some local governments, leaving financial statements vulnerable to data loss, destruction, and deletion, raising questions about their legitimacy and fairness. Weakness in accounting Information System and Accounting basis 7 Spending more cash than the limit of 50 JOD without using current checks, and possibly using postponed checks, is a violation of financial regulations. Failure to comply with the laws and regulations. And weakness of internal auditing 8 violation of income tax for council members' and employees' salaries and wages Weakness in Internal Auditing 9 Not producing the financial statement in accordance with the rules and regulations that apply. Accounting Basis and violation of laws and regulations 10 Both real cash counting and cash insurance are not practices on a regular basis. For the purpose of monitoring their bank accounts, some municipal governments might not carry out proper bank reconciliations. Weakness in Internal Auditing Source: Author 4.9 Summary of Chapter Four In this chapter, there are two main stages to the data analysis process. An initial analysis of the data was part of the first stage. In order to use SEM effectively, the 207 data must adequately meet the fundamental assumptions. The entire data set of the items was, in general, normally distributed and devoid of errors, missing values, and univariate outliers. The two SEM stages were applied in the second phase. The first step involved creating measurement models for the research's latent constructs. Following the first stage's confirmation of the constructs' unidimensionality, reliability, and validity, the second stage was created to put the research hypotheses to the test by creating structural models. This chapter analyzes and discusses SAIs’ reports in addition to the descriptive analysis of each variable in the research.