pdf - Nyenrode Business Universiteit

More documents

Recommendations

Info

4.4. ANALYSES 107 The model does show significant relations for the antecedents for the room for managerial opportunism. The room for managerial opportunism is influenced by information asymmetry (positively), goal ambiguity (positively), measurability of outputs (negatively), and decentralization of decision rights (positively). The directions of these relations are consistent with my expectations. The size and industry control variables do not have significant effects on the room for managerial opportunism. To evaluate the fit of the model, I calculated various fit indexes. Assessing the fit of a SEM is not as straightforward as, for example, an OLS regression analysis. The evaluation of the statistical properties of various fit indexes is still an active research topic, and no single measure provides a definite answer with regard to the fit between a model and the data, as a wide variety of goodness-of-fit statistics are available for SEM (Kline 2005). Kline suggests that multiple measures must be assessed to evaluate the goodness-of-fit. As the indicators of model fit, I use the chi 2 , the goodness-of-fit index (GFI), the adjusted goodness-of-fit index (AGFI), the comparative fit index (CFI), the root mean square residual (RMR), and the root mean square error of approximation (RMSEA). Kline (2005) suggests that the following factors indicate a good fit: an insignificant chi 2 ; a GFI, AGFI, and CFI greater than 0.90; a RMR close to 0; and an RMSEA less than 0.045. As presented in table 4.12 - panel B, the indicators show good model fit. Despite the good fit, the model shows potential non-normality issues. The multivariate kurtosis measure lies above the threshold value of ten suggested by Kline (2005). Kline (2005) argues that kurtosis greater than ten may suggest a problem with the data. The kurtosis in the current model is 13.524. The high kurtosis is due to the inclusion of control variables for business unit and firm size 12 . Because multivariate normality is a strict assumption of maximum likelihood estimation (MLE) in SEM models, I use bootstrapping with 500 samples, which does not assume multivariate normality, to analyse whether the kurtosis affects the MLE estimates (Kline 2005, Bryne 2001). I find that the results are qualitatively similar, and that normality issues did not drive the results. Additionally, the Bollen-Stine bootstrap p does not reject the model. The Bollen-Stine bootstrap is a bootstrap modification of model chi 2 that is used to test model fit by, for example, adjusting for the lack of multivariate normality. Similar to the chi 2 , the Bollen-Stine bootstrap is insignificant at p = 0.701, which indicates a fit between the data and the model. The bootstrap statistics are presented in table 4.12 - panel C. 4.4.2.2 Analyses with RPE-based-Targets Measure To buttress the findings of the base model analyses, I run a robustness check. Here I follow the strategy of chapters 2 and 3 by rerunning the analysis with a different operationalization of RPE use: RPE-based-Targets. 12 Excluding the size controls from the model reduces the kurtosis to 3.928, but does not yield qualitatively different conclusions.
108 CHAPTER 4. OPPORTUNISM MITIGATION EFFECTIVENESS RPE-based-Targets specifically measures the explicit use of RPE for target-setting purposes (i.e., how peer performance influences the difficulty of the performance standards upon which the manager’s performance is evaluated). I measure RPE-based-Targets as the impact of peer performance on the performance target. I ask to what extent the performance of the peers influences the difficulty of the target and/or budget for a variety of performance measures, as described in the appendix (Q2). To interpret the impact of peers on the evaluation as a whole, I weigh the impact of peers on each of the performance metrics by multiplying the scores with their individual importance for the performance evaluation (Q3). RPE-based-Targets does not solely measure ex-post peer comparison. Peer performance can also influence the agent’s target at the beginning of the period. In this case, the target for the upcoming period is based on the peers’ actual performance during the current period. I expect this second RPE measure to correlate with the primary measure ‘RPE-Use’ because they both include explicit RPE use and apply the peer comparison information ex-post. Because of the differences between the measures presented above, the correlation may not be strong. Correlation analysis confirms this inference; the two measures of RPE are correlated with one another but not strongly (0.498; p < 0.01). The analyses with the RPE-based-Targets measure support the main findings of the base model analyses, albeit with a trimmed and marginally significant model. Similar to the base model, the model has a kurtosis that lies above the threshold value of 10 (Kline 2005). In contrast to the previous models described in this thesis, the kurtosis of 14.179 significantly affects the results of the estimated model. The Bolen-Stine bootstrap is significant (p = 0.025). This finding indicates that there is no fit between the data and the model. To reduce the non-normality of the data, I excluded the kurtosis-increasing size controls from the analyses. The trimmed model has a kurtosis of 2.783, which indicates that the trimmed model does not suffer from potential non-normality issues. Both the original and trimmed versions of the robustness check model marginally fit the data. As argued in the previous section, the evaluation of model fit with SEM is not based on a single fit statistic. Relying on multiple statistics can lead to contradictory results, as shown by the current model. The model’s chi 2 , the adjusted goodness-of-fit index (AGFI), and the root mean square error of approximation (RMSEA) index indicate an inadequate fit between the model and the data. However, the goodness-of-fit index (GFI), the comparative fit index (CFI), and the root mean square residual (RMR) suggest that the model and the data do fit. As a result, no definite conclusion about the goodness of fit can be drawn. Still, because the inconclusive set of fit indexes does not refute the model, the results of this model are presented in this section as another illustration of the relation between RPE use and the room for managerial opportunism. The robustness check yields results that are largely similar to those of the model with the primary RPE measure (RPE-Use). The current model explains 14% of the variance of RFMO. Similar to the base model, the model with the proxy for the RPE-based-Target does not support the research hypothesis, as the model shows a positive relation between
Page 2 and 3:
Relative Performance Evaluation in
Page 4 and 5:
NYENRODE BUSINESS UNIVERSITEIT Rela
Page 6 and 7:
Table of Contents Preface 1 General
Page 8:
Preface Why do people make things h
Page 11 and 12:
2 CHAPTER 1. GENERAL INTRODUCTION I
Page 13 and 14:
4 CHAPTER 1. GENERAL INTRODUCTION o
Page 15 and 16:
6 CHAPTER 1. GENERAL INTRODUCTION T
Page 17 and 18:
8 CHAPTER 1. GENERAL INTRODUCTION F
Page 19 and 20:
10 CHAPTER 1. GENERAL INTRODUCTION
Page 22 and 23:
Chapter 2 The Use of Relative Perfo
Page 24 and 25:
2.1. INTRODUCTION 15 is not a promi
Page 26 and 27:
2.2. THEORETICAL BACKGROUND 17 The
Page 28 and 29:
2.2. THEORETICAL BACKGROUND 19 Sinc
Page 30 and 31:
2.2. THEORETICAL BACKGROUND 21 who
Page 32 and 33:
2.2. THEORETICAL BACKGROUND 23 suff
Page 34 and 35:
2.3. SAMPLE AND MEASUREMENT 25 2.2.
Page 36 and 37:
2.3. SAMPLE AND MEASUREMENT 27 2.3.
Page 38 and 39:
2.3. SAMPLE AND MEASUREMENT 29 can
Page 40 and 41:
2.3. SAMPLE AND MEASUREMENT 31 Comp
Page 42 and 43:
2.3. SAMPLE AND MEASUREMENT 33 this
Page 44 and 45:
2.3. SAMPLE AND MEASUREMENT 35 Emph
Page 46 and 47:
2.4. ANALYSES 37 2.4 Analyses This
Page 48 and 49:
2.4. ANALYSES 39 Table 2.8: RPE use
Page 50 and 51:
2.4. ANALYSES 41 2.4.2 Correlations
Page 52 and 53:
2.4. ANALYSES 43 Concerning the con
Page 54 and 55:
2.4. ANALYSES 45 Table 2.13: Result
Page 56 and 57:
2.4. ANALYSES 47 Table 2.15: Summar
Page 58 and 59:
2.5. CONCLUSIONS AND DISCUSSION 49
Page 60 and 61:
2.5. CONCLUSIONS AND DISCUSSION 51
Page 62 and 63:
2.6. CONTRACTIBILITY MIMIC MODEL (A
Page 64 and 65:
2.6. CONTRACTIBILITY MIMIC MODEL (A
Page 66 and 67: 2.7. ANALYSES FOR FOR-PROFIT BUS (A
Page 68 and 69: Chapter 3 Does Relative Performance
Page 70 and 71: 3.1. INTRODUCTION 61 Holmstrom’s
Page 72 and 73: 3.2. DEVELOPMENT OF THE MODEL 63 3.
Page 74 and 75: 3.2. DEVELOPMENT OF THE MODEL 65 RP
Page 76 and 77: 3.3. SAMPLE AND MEASUREMENT 67 3.3
Page 78 and 79: 3.3. SAMPLE AND MEASUREMENT 69 3.3.
Page 80 and 81: 3.3. SAMPLE AND MEASUREMENT 71 Tabl
Page 82 and 83: 3.3. SAMPLE AND MEASUREMENT 73 Tabl
Page 84 and 85: 3.4. ANALYSES 75 3.4 Analyses This
Page 86 and 87: 3.4. ANALYSES 77 Table 3.9: Pearson
Page 88 and 89: 3.4. ANALYSES 79 Table 3.10: Result
Page 90 and 91: 3.4. ANALYSES 81 Table 3.11: Result
Page 92 and 93: 3.4. ANALYSES 83 Larcker & Rusticus
Page 94 and 95: 3.5. CONCLUSIONS AND DISCUSSION 85
Page 96 and 97: Chapter 4 Does Relative Performance
Page 98 and 99: 4.1. INTRODUCTION 89 Moreover, chap
Page 100 and 101: 4.2. DEVELOPMENT OF THE MODEL 91 is
Page 102 and 103: 4.2. DEVELOPMENT OF THE MODEL 93 Sp
Page 104 and 105: 4.3. SAMPLE AND MEASUREMENT 95 4.3
Page 106 and 107: 4.3. SAMPLE AND MEASUREMENT 97 To i
Page 108 and 109: 4.3. SAMPLE AND MEASUREMENT 99 of R
Page 110 and 111: 4.3. SAMPLE AND MEASUREMENT 101 Tab
Page 112 and 113: 4.4. ANALYSES 103 Table 4.8: Items
Page 114 and 115: 4.4. ANALYSES 105 RPE use ✻ Posit
Page 118 and 119: 4.4. ANALYSES 109 RPE and RFMO. Mor
Page 120 and 121: 4.4. ANALYSES 111 room for manageri
Page 126 and 127: Chapter 5 Conclusion 5.1 Summary, C
Page 128 and 129: 5.1. SUMMARY, CONCLUSIONS AND DISCU
Page 134 and 135: 5.2. IMPLICATIONS FOR PRACTICE 125
Page 136 and 137: 5.3. LIMITATIONS AND SUGGESTIONS FO
Page 138 and 139: Appendix Overview of Questionnaire
Page 140 and 141: Q5 How often do the following thing
Page 142 and 143: Q14 When serving your clients, how
Page 144 and 145: Q23 Please indicate whether you agr
Page 146 and 147: Noise Q34 NOISE (primary measure) P
Page 148 and 149: References 1. Abernethy, M.A., Bouw
Page 150 and 151: 26. Diamantopoulos, A. & Siguaw, J.
Page 152 and 153: 52. Himmelberg, C.P. & Hubbard R.G.
Page 154: 83. Towry, K. 2003. ‘Control in a
Page 157 and 158: 148 beïnvloeden, veroorzaken ruis
Page 160 and 161: Biography Hilco van Elten (1979) ho
show all

pdf - Nyenrode Business Universiteit

Create successful ePaper yourself

Delete template?

Save as template?