CHAPTER 5 Research Methodology
5.6 Statistical Analysis
The statistical analysis included descriptive statistics, t-tests, Spearman’s correlation and analysis of variance (ANOVA). ANOVA and regression analysis have been used in prior studies to test the relationship between corporate governance practices and firm performance.
The aim in this context is to determine (1) whether a change in the corporate governance variables had taken place in 2010 as a result of the mandated rules introduced in 2009 by the UAE ES&CA and (2) if such mandated rules had an impact on corporate performance. In this study, the analysis used ANOVA to examine the differences o between the two observed periods of the same samples. The analysis compares the differences of the variances between observations and variances in the means. This is an appropriate statistical method in the study to determine if there are statistically significant differences between the observations.
5.6.1 Descriptive Statistics
Descriptive statistics measure the central tendency and dispersion and were used to analyse the basic features of the data in this study. The measures of central tendencies were: mean, mode and median with a particular focus on the mean as it is a significant measure of the central tendency. In addition, the mean was used to indicate the maximum and minimum values of the research variables to show standard deviation or the range (Veal, A.J. & Ticehurst 2005).Descriptive statistics are also useful for making general observations about the data collected. They report on trends and patterns of data and provide a basis for comparison between variables. In this study descriptive statistics provided a comparison of changes in the data between 2008-2009 and 2011-2012. They indicated the extent to which companies had complied with the corporate governance rules which were introduced by the UAE ES&CA and the trends of the firm performance variables.
5.6.2 T-Test
The t-tests are used to determine if there are significant differences between two means (Veal, A.J. & Ticehurst 2005). The SPSS program was used in this study for the analysis of t-test. Two-related sample t-tests were conducted to determine if the differences in corporate governance practices in 2008-2009 and 2011-2012 were significant.
A Wilcoxon Signed Rank Test (two-related-sample tests), which is the non- parametric version of the paired sample t-test (Carver & Nash 2011), was conducted to test the significance of the means of the variables for 2008-2009 and 2011-2012; two-related- sample t-tests are used when measurements are repeated for the same sample (Carver & Nash 2011). Two-related sample t-test was conducted to identify if the differences in the characteristics of corporate governance between 2008-2009 and 2011-2012 were significant.
5.6.3 Spearman’s Correlation
Correlation is used to examine the relationship between two or more ordinal or ratios variables. Correlation can be measured by means of correlation coefficient. The significance of a correlation coefficient depends on its magnitude and the sample size and can be assessed by means of t-test (Veal, A.J. & Ticehurst 2005). Spearman’s Rank correlation coefficient is used to identify and test the strength of a relationship between two sets of data. Furthermore, and for the purpose of this thesis it is worth noting that multicollinearity will not be an issue due to the lack of significant correlation between the independent variables.
Spearman’s rank correlation is calculated as follows: r = (1-6Σd2)/N(n2 - 1)
Where d is the difference between the two ranked variables, n is the number of data pairs and Σ indicates the sum of values
5.6.4 Analysis of Variance
Analysis of variance (ANOVA) is used to compare difference between more than two means at a time. Whether or not the means are from one population (with one mean) or from different sub-populations (with different means) depends not only on the differences between the means but also on how much they are spread out or dispersed(Veal, A.J. & Ticehurst 2005) ANOVA is an exploratory analysis and examines significance in the case of cross- tabulated means and determines whether the differences revealed are within the acceptable significance levels of <.05. The strong point of ANOVA is its capability to distinguish effects in response to many different sources of variations compared simultaneously or in certain cases through time. It has the ability to identify interacting factors and the capability to measure the scale of variation within a hierarchy of effects. This versatility makes it a powerful tool for answering questions about causality (Fitrijanti & Alamanda 2013).
5.6.5 Incremental Regression
The incremental regression analysis was performed to determine the importance of an individual variable in affecting the performance of a firm, by removing the individual variables from the model and capturing the effect on R-squared (Field 2005). These tests highlighted the importance of individual variables in affecting the performance variable in the model.
5.7 Conclusion
This chapter discussed the methodology used to test the hypotheses of the study. Sample size, data collection, the design of the variables, measurement tools and operationalization were discussed. The chapter also discussed the methodology used for data collection and the relevant statistical techniques employed to analyse and interpret the data to examine the relationship between corporate governance variables and their impact on firm performance in the UAE. Next chapter discusses the results of these statistical analyses.