CHAPTER 5: SAMPLE CHARACTERISTICS AND MEASUREMENT SCALES
5.6. Descriptive statistics and collinearity assessment
The descriptive statistics provide simple summaries about the sample data and measures. Descriptive statistics summarize the data and formed the basis of quantitative analysis of data. In addition, to analyse the association between the dependent and independent variables, a correlation analysis (Pearson) is also undertaken. Moving forward, the issue of potential collinearity problem is also tested to consider removing one of the corresponding indicators before conducting the further regressions.
Appendix 22 contains descriptive statistics for all the variables used in this study. As can be observed from Table 5.2, the valid number of observations for each variable is 174 samples. Mean, median, maximum, minimum, standard deviation, Skewness and Kurtosis are reported for each variable used in the current study. Skewness and Kurtosis statistics all suggest that the variables are not normally distributed. To reduce the heteroskedasticity problem arising out of the non-normal distributions, regressions are estimated by the PLS- SEM as the advantage of this regression method. HCE has a significantly greater mean than SCE and RCE.
Appendix 22 also presents Pearson’s correlation coefficient analysis for the dependent and independent variables. Correlation coefficient summarizes the linear relationship between two variables having ranked and provide sufficient information from this study’s point of view. Under the Pearson’s correlation, of particular note is that the correlation coefficients are not of high magnitude between any two of the independent variables to cause concern about multicollinearity problems.
The relationships of HCE-SMA (0.551); SCE-SMA (0.514); RCE-SMA (0.829) are significantly positive, roughly supporting the second hypothesis that each of IC components is positively associated with the practices of strategic management accounting.
The correlation analyses show that, under the Pearson’s correlation, all IC components are positively related with all of the corporate performance indicators at the
significant level. The results for IC components demonstrate that increase in value creation efficiency will increase in market value, profitability and operation efficiency. This supports the third hypothesis that there are significant positive associations between each of IC components and each of corporate performance indicators.
The relationships of SMA-ATO (0.713); SMA-INVEFF (0.776); SMA-ROE (0.930);
SMA-TOBINQ (0.844) are significantly positive, roughly supporting the forth hypothesis that firms with more SMA practices have significantly positive effect on corporate performance.
Furthermore, Appendix 22 also indicates that each group of SMA practices has the significantly positive relationship with each of IC components, except for the outcome related to the relationship between customer accounting and structural capital (0.011 insignificantly). This means that each group of SMA practices is likely to positively associate with most of the IC components as the initial basis of the result prediction before conducting regression models.
According to the results in Appendix 23, VIF values all are uniformly below the threshold value of 2, except for SMA variable under the value of 5. It is concluded, therefore, that collinearity does not reach critical levels in any of constructs and is not an issue for the estimation of the partial least square path models.
SUMMARY OF CHAPTER 5
Chapter 5 presents data collection to construct the variables of SMA practices, the assessment of reflective and formative measurement scales, descriptive statistics and collinearity assessment.
The final sample was composed of 174 valid responses with the dominance of manufacturing sector. 70.11% of the sampled organisations is categorised into the group of high level of SMA practices. 92.62% of the large enterprises has the higher level of SMA implementation. Statistical results of the sample by position and working years in the current position show that the informants are finance managers (27.01%), followed by reporting managers (20.11%), then by the head of the department (14.36%) and general managers (14.36%). All respondents are from senior managers or members of top management team with knowledge about accounting, planning or finance and at least 2 years of working experience in the current organizations.
This chapter also demonstrates the assessment outcomes of reflective and formative measurement scales in terms of strategic management accounting practices, followed the research process illustrated in Chapter 4. The reliability, convergent validity and discriminant validity of the measurement scales related SMA practices are supported by the dataset. The 18 final indicators for 4 constructs of SMA practices are satisfactory for further analyses. The results indicate that there is no measurement scale for exception. The issue of collinearity also does not reach critical levels in any of inner constructs.
The next chapter is going to introduce data analysis and the empirical outcomes of measurement and structural models. It also includes not only testing the direct regression and the mediated path regressions in the relationship between IC and corporate performance via the mediation of SMA practices but also the testing the impact of SMA practices on IC management.