3. Methodology
Regression Models: This study uses a logistic regression model to examine the relationship between
pressure and opportunity factors and the level of fraud occurrence.
Fraud= β0+ β1 SALAR + β2 LEV + β3 AUDCSIZE + β4 BRDSIZE+ ε
Where:
FRAUD = Coded 1 if the firm experienced fraud previously and 0 otherwise.
SALAR= Sales to Accounts receivables
LEV= Total debt to Total assets
AUCSIZE= Number of audit committee members
BRDSIZE= Number of board of directors members
The FRAUD as the dependent variable is a dummy variable representing whether the firm has experienced
fraud. Hypothesis 1 predicts that financial statement fraud is more likely to occur when the pressure is high.
This study assumes that the coefficients on sales to accounts receivables ratio and leverage are positively
associated with the fraud occurrence in organizations. Hypothesis 2 expects that organizations experience
high level of fraud when there is more opportunity for fraud commitment. We expect that the numbers of
audit committee and board of directors’ members are negatively associated with the likelihood of financial
statement fraud occurrence. To examine our hypotheses, the regression model utilized to test the relationship
between defined proxies and the probability of financial statement fraud occurrence.
Sample selection: The sample of this research is divided into two groups of fraud and non-fraud companies.
The list of fraud companies is originated from a study in Malaysia which investigated the governance and
company value (Kwan & Kwan, 2011). Another source for updating fraud companies list is Security
Commission Malaysia (SC) website. Data on leverage, total sale and account receivable is collected through
DATASTREAM database. Another source of collecting these data is companies’ annual reports getting from
Bursa Malaysia website. Overall, the sample includes 40 fraud companies and 100 non-fraud companies for
the period of 2002 to 2012 for this study.
đang được dịch, vui lòng đợi..