10 meaning the multicollinearity is not concerned (Hair and others, 2006).
Mahasarakham University
70
Correlation analysis. This research uses correlation analysis to test correlation
among all variables and provide a correlation matrix that shows the intercorrelations
among all variables for the initial analysis. As the variables become highly correlated,
the multicollinearity problem may occur. This problem occurs when any single
independent variable is highly correlated with a set of other independent variables. As
multicollinearity increases, it complicates the interpretation of the variables because a
variable can be explained by the other variables in the analysis. Multicollinearity is
indicated when the inercorrelation between explanatory variables exceeds 0.80 (Berrt
abd Feldmann, 1985). Consequently, factor analysis is used to group highly correlated
variables together and the factor score of all variables are prepared to avoid the
multicollinearity problem. Then, they are evaluated by the Ordinary Least Squares
(OLS) regression analysis.
Regression analysis. The Ordinary Least Squares (OLS) regression analysis is
used to test all hypotheses following the conceptual model. Because both dependent and
independent variables in this research are categorical data and interval data, OLS is an
appropriate method for examining the hypotheses. The related model and the
hypotheses in this research are transformed to ten equation models. Moreover, two
control variables, i.e., gender and age are included in all of those equations for
hypothesis testing.
The investigation of the relationship among four dimensions composed in best
audit practices and audit credibility is depicted in Equation 1 as related to H1-H4 as
follows:
Equation 1:
AC = β01 + β1APW + β2APN + β3AIL + β4AMR + β5GEN + β6AGE+ ε1
The examination of the relationship between four dimensions composed in best
audit practices and audit independence is depicted in Equation 2 as related to H5-H8a as
follows:
Mahasarakham University
71
Equation 2:
AI = β02 + β7APW + β8APN + β9AIL + β10AMR + β11GEN +β12AGE + ε2
The investigation of the relationship between four dimensions composed in
best audit practices and audit judgment is depicted in Equation 3 as related to H5-H8b
as follows:
Equation 3:
AJ = β03 + β13APW + β14APN + β15AIL + β16AMR + β17GEN +β18AGE + ε3
The examination of the relationship between four dimensions composed in best
audit practices and audit performance is depicted in Equation 4 as related to H5-H8c as
follows:
Equation 4:
AP = β04 + β19APW + β20APN + β21AIL + β22AMR + β23GEN +β24AGE + ε4
The investigation of the relationship between audit independence, audit
judgment, audit performance, and audit credibility is depicted in Equation 5 as related to
H9-H11 as follows:
Equation 5:
AC = β05 + β25AI + β26AJ + β27AP + β28GEN + β29AGE + ε5
Hence, this equation which used to examine the relationship between audit
independence, audit judgment, and audit performance is depicted in Equation 6 as
related to H12-H13 as follows:
Equation 6:
AP = β06 + β30AI + β31AJ + β32GEN+ β33AGE + ε6
Mahasarakham University
72
The third sub-model that examination determined is the role of moderating
effects namely audit profession climate and proactive individual goal which moderates
the relationship between antecedents and best audit practices of four dimensions which
are audit profession well-roundedness, audit practice norms, audit innovation learning,
and audit morality reasoning is depicted in Equations 7-10 as related to H14a-d to
H22a-d, as follow:
Equation 7:
APW = β07 + β34CIM + β35PC + β36AE + β37APC + β38PIG +
β39(CIM*APC) + β 40(PC*APC) + β41(AE*APC) + β42(CIM*PIG)
+ β43(PC*PIG) + β44(AE*PIG) + β45GEN + β46AGE + ε7
Equation 8:
APN = β08 + β47CIM + β48PC + β49AE + β50APC + β51PIG +
Β52(CIM*APC) + β 53(PC*APC) + β54(AE*APC) + β55(CIM*PIG)
+ β56(PC*PIG) + β57(AE*PIG) + β58GEN + β59AGE + ε8
Equation 9:
AIL = β09 + β60CIM + β61PC + β62AE + β63APC + β64PIG +
Β65(CIM*APC) + β 66(PC*APC) + β67(AE*APC) + β68(CIM*PIG)
+ β69(PC*PIG) + β70(AE*PIG) + β71GEN + β72AGE + ε9
Equation 10:
AMR = β10 + β73CIM + β74PC + β75AE + β76APC + β77PIG +
Β78(CIM*APC) + β 79(PC*APC) + β80(AE*APC) + β81(CIM*PIG)
+ β82(PC*PIG) + β83(AE*PIG) + β84EN + β85AGE + ε10
Mahasarakham University
73
Table 5: Descriptive of Abbreviation Variables
Abbreviations Full words Abbreviations Full words
BAP Best Audit Practices CIM Continuous Improvement
Mindset
APW Audit Profession
Well-Roundedness
PC Professional Commitment
APN Audit Practice Norms AE Audit Experience
AIL Audit Innovation
Learning
APC Audit Profession Climate
AMR Audit Morality
Reasoning
PIG Proactive Individual Goal
AI Audit Independence GEN Gender
AJ Audit Judgment AGE Age
AP Audit Performance ε Error Term
AC Audit Credibility
Summary
This chapter summarizes the research methods used in this research for
gathering data and examining all constructs in the conceptual model to answer the
research question. The details consist of the sample selection and data collection
procedure including population and sample, data collection, and test of non-response
bias. Furthermore, the variable measurements are followed for each of all variables in
the conceptual model. Additionally, the instrumental verifications include the test of
validity and reliability and the Ordinary Least Squares (OLS) regression analysis is used
to test hypotheses are presented. In conclusion, CPAs as the sample is selected from the
Federation of Accounting Professions under the Royal Patronage of His Majesty the