pdf - Nyenrode Business Universiteit

pdf - Nyenrode Business Universiteit pdf - Nyenrode Business Universiteit

30.08.2013 Views

4.4. ANALYSES 105 RPE use ✻ Positive Information Asymmetry - Goal Ambiguity - Measurability of Outputs - Decentralization Size & Sector 4.4.2 Multivariate Analyses Figure 4.3: Statistical Model Negative (H1) Positive Room for Managerial Opportunism In order to test RPE’s opportunism-mitigating effect, I analyse the model that is presented in figure 4.3. Two analyses are run: first, the model is tested with the RPE-Use measure. This model is referred to as the base model. Second, an alternative measure for RPE use is used as a robustness check. Both models fit the data, but do not support the research hypothesis; in my models, RPE use does not reduce the room for managerial opportunism. The results of these analyses and several robustness checks are discussed in the remainder of this section. 4.4.2.1 Analyses with RPE-Use Measure This subsection presents the base model results of the multivariate analysis of the mitigating effect of RPE on the room for managerial opportunism, using the primary RPE- ✻ ❄ ✻ ✻

106 CHAPTER 4. OPPORTUNISM MITIGATION EFFECTIVENESS measure. The multivariate analyses are conducted through structural equation modelling 11 (SEM). The SEM analysis shows overall support for the model. The model explains 14% of the variance in the room for opportunism. The model is presented in table 4.12 - panel A. The model does not support the research hypothesis. Instead, RPE-Use has a positive effect on the room for managerial opportunism; more RPE use is associated with increased room for opportunism. In the conclusions and discussion section of this chapter, this unexpected finding will be discussed and confronted with prior literature. The RPE-inducing effect of information asymmetry is also unsupported by the model. This finding is consistent with the results of chapter 2, where information asymmetry (interacting with the level of comparability of the business unit) does not increase RPE use. Table 4.12: Results of SEM Analysis (RPE-Use) Panel A - base model results Coefficient Std. Error Critical Ratio Prob. RPE use > RFMO 0.100 0.058 1.883 0.06 ‘a’ Inf. Asymmetry > RFMO 0.080 0.070 1.372 0.09 ‘a’ Inf. Asymmetry > RPE use 0.051 0.062 0.901 0.18 ‘a’ Goal Ambiguity > RFMO 0.198 0.127 3.137 0.00 ‘a’ Meas. of Outputs > RFMO -0.216 0.110 -3.265 0.00 ‘a’ Decentralization > RFMO 0.084 0.056 1.446 0.07 ‘a’ BU size > RFMO -0.021 0.030 -0.339 0.73 Firm size > RFMO -0.017 0.028 -0.266 0.79 Dummy prod. > RFMO 0.076 0.162 1.376 0.17 Dummy fin. serv. > RFMO -0.033 0.164 -0.597 0.55 Dummy N.F.P. > RFMO -0.052 0.170 -0.887 0.38 ‘a’ : variable based on directional expectation; significance calculated as onetailed p-value Panel B - base model fit statistics Chi 2 = 5.656 goodness-of-fit statistics Degrees of freedom = 8 GFI = 0.997 Probability level = .686 AGFI = 0.973 CFI = 1.000 Multivariate normality RMR = 0.021 Kurtosis 13.524 (c.r. 7.017) RMSEA = 0.000 Panel C - bootstrap results Bootstrap samples = 500 Bollen-Stine bootstrap p = 0.701 11 The SEM analyses are run in AMOS 18.

106 CHAPTER 4. OPPORTUNISM MITIGATION EFFECTIVENESS<br />

measure. The multivariate analyses are conducted through structural equation modelling 11<br />

(SEM). The SEM analysis shows overall support for the model. The model explains 14% of<br />

the variance in the room for opportunism. The model is presented in table 4.12 - panel A.<br />

The model does not support the research hypothesis. Instead, RPE-Use has a positive<br />

effect on the room for managerial opportunism; more RPE use is associated with increased<br />

room for opportunism. In the conclusions and discussion section of this chapter, this<br />

unexpected finding will be discussed and confronted with prior literature.<br />

The RPE-inducing effect of information asymmetry is also unsupported by the model.<br />

This finding is consistent with the results of chapter 2, where information asymmetry<br />

(interacting with the level of comparability of the business unit) does not increase RPE<br />

use.<br />

Table 4.12: Results of SEM Analysis (RPE-Use)<br />

Panel A - base model results Coefficient Std. Error Critical Ratio Prob.<br />

RPE use > RFMO 0.100 0.058 1.883 0.06 ‘a’<br />

Inf. Asymmetry > RFMO 0.080 0.070 1.372 0.09 ‘a’<br />

Inf. Asymmetry > RPE use 0.051 0.062 0.901 0.18 ‘a’<br />

Goal Ambiguity > RFMO 0.198 0.127 3.137 0.00 ‘a’<br />

Meas. of Outputs > RFMO -0.216 0.110 -3.265 0.00 ‘a’<br />

Decentralization > RFMO 0.084 0.056 1.446 0.07 ‘a’<br />

BU size > RFMO -0.021 0.030 -0.339 0.73<br />

Firm size > RFMO -0.017 0.028 -0.266 0.79<br />

Dummy prod. > RFMO 0.076 0.162 1.376 0.17<br />

Dummy fin. serv. > RFMO -0.033 0.164 -0.597 0.55<br />

Dummy N.F.P. > RFMO -0.052 0.170 -0.887 0.38<br />

‘a’ : variable based on directional expectation; significance calculated as onetailed<br />

p-value<br />

Panel B - base model fit statistics<br />

Chi 2 = 5.656 goodness-of-fit statistics<br />

Degrees of freedom = 8 GFI = 0.997<br />

Probability level = .686 AGFI = 0.973<br />

CFI = 1.000<br />

Multivariate normality RMR = 0.021<br />

Kurtosis 13.524 (c.r. 7.017) RMSEA = 0.000<br />

Panel C - bootstrap results<br />

Bootstrap samples = 500 Bollen-Stine bootstrap p = 0.701<br />

11 The SEM analyses are run in AMOS 18.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!