Multiple Regression Analysis
SCHOOL OF BUSINESS AND PUBLIC MANAGEMENTBUSINESS RESEARCH METHODS TERM PAPERNdikwe and Owino (2016) undertook a study entitled, “The Influence of Corporate Governance on Financial Performance of Public Secondary Schools in Kenya” They collected primary data and tested the following research hypothesis;H01: Board composition has no significant influence on the financial performance of secondary schools in Kenya.H02: Board skills have no significant effect on the financial performance of public secondary schools in Kenya.H03: The application of corporate governance principles has no significant impact on the financial performance of public secondary schools in Kenya.H04: Separation of duties between the boards and management has no significant effect on the financial performance of public secondary schools in Kenya.The study was guided by the following analytical framework;FP = β0+β1BC + β2ABS + β3CGP + β4SOD + ԐiWhere FP= financial performance of public secondary schools in Mathira Constituency β0=Constant term
β1, β2, β3, β4 are beta coefficients BC = board composition, ABS = available board skills CGP = corporate governance principles SOD= separation of dutiesԐ 0 = Error TermRequiredPerform a multiple regression analysis based on the attached data set and answer the following questions.Test the data for the following assumptions of regression analysis and interpret your findings: NormalityNormality was tested using a histogram with a superimposed normal curve as shown below. The graph shows a concentration of the variables at the centre of the histogram. This results show that the data was normally distributed and hence adequate for regression analysis. [pic 1]Linearity The study tested the existence of a linear relationship between the financial performance and corporate governance, board skills and board composition. The results in the Table below shows the existence of a significant relation between corporate governance and financial performance (p=0.001), board skills and financial performance (p=0.000) and board composition and financial performance (p=0.000). This meant there existed a linear relationship between the independent and dependent variables.