Skip to main content

Table 4 Multinomial regression results for 2006–2010

From: Voluntary disclosure of internal control and auditor’s attestation: evidence from China

 

Separate logistic regressions

Multinomial logistic regressions (Model 3)

Model 1

Discloser versus non-disclosers

Model 2

Self-assessment report only versus Self-assessment report with auditor’s attestation

(1)

Self-assessment report only versus non-disclosers (ICVD1)

(2)

Self-assessment report with auditor’s attestation versus non-disclosers (ICVD2)

(3)

Diff between (1) and (2) Chi-squared

SEO

0.75*** (3.40)

0.54 (1.52)

0.28 (0.90)

1.05*** (4.29)

5.33**

FUND

4.63*** (3.84)

5.12*** (2.67)

1.80 (1.12)

5.94*** (4.29)

5.58**

GOV

0.22 (1.45)

0.47* (1.91)

−0.05 (−0.24)

0.42** (2.10)

3.54*

GOV*CENTRL

0.41** (2.25)

0.93*** (3.60)

−0.22 (−0.85)

0.69*** (3.53)

11.69***

GOV*SEO

−0.35 (−1.28)

−0.50 (−1.16)

0.14 (0.36)

−0.66** (−2.19)

3.85**

SIZE

0.55*** (8.57)

−0.03 (−0.36)

0.62*** (7.29)

0.53*** (7.17)

 

GROW

0.01 (0.26)

0.08 (0.88)

−0.16 (−1.64)

0.01 (0.74)

 

ROE

−0.01 (−0.55)

0.14 (0.72)

−0.01 (−0.48)

−0.01 (−0.44)

 

LEV

−1.49*** (−4.27)

0.37 (0.68)

−1.72*** (−3.52)

−1.32*** (−3.33)

 

INDIR

0.71 (0.61)

1.23 (0.78)

0.01 (0.01)

1.44 (1.00)

 

CLIST

1.36** (2.11)

0.37 (0.47)

1.42** (1.97)

1.75** 2.43)

 

BIG4

0.19 (0.82)

−1.95*** (−5.14)

1.01*** (3.41)

−0.75** (−2.38)

 

INV

0.90** (2.53)

−0.06 (−0.09)

0.86* (1.82)

0.90** (2.01)

 

AGE

−0.07*** (−4.21)

−0.11*** (−4.20)

−0.01 (−0.77)

−0.11*** (−5.48)

 

Industry MBR

0.03 (0.37)

0.08 (0.68)

−0.33 (−0.31)

0.04 (0.80)

 

# of Obs.

3994

1198

3994

  

Pseudo R2

23.26%

17.67%

21.99%

  
  1. This table presents the results of logistic regressions for the full sample period. Model 1 is a simple logistic regression examining the likelihood of disclosing an IC self-assessment report. Model 2 is also a simple logistic regression examining the likelihood of attaining an auditor’s attestation on the self-assessment report among disclosers. Model 3 is for multinomial logistic regressions; we use non-disclosers as our reference group, and examine the likelihood simultaneously for firms only disclosing an IC self-assessment report (ICVD1) and for firms both disclosing an IC self-assessment report and attaining an auditor’s attestation (ICVD2) on the report. Estimated coefficients and z statistics (in parentheses) are reported
  2. *, **, and ***denote the significance levels at 10, 5 and 1%, respectively