Global Journal of Management and Business Research, A: Administration and Management, Volume 23 Issue 5

twenty-eight (28) consumer goods companies that were listed on the Nigerian Stock Exchange as of 2019. The entire population was used as a sample; nonetheless, due to insufficient data from some of the listed consumer goods companies (CGC), the sample was reduced to 17 companies using a filter. The data was retrieved from a secondary source: CSR to employee, control variables and financial performance. Data for CSR to the employee was collected using CSR checklist (See appendix B). The CSR checklist was adopted from Jitaree (2015). The annual report of each CGC was perused using content analysis for each practice (item) in the checklist for measurement accordingly. The total CSR to employee index or score for each company for each year was the ratio of the total score obtained to the total score (16) attainable and accordingly computed for the CSR to the employee. This index ranges from 0 to 1. That is, )(1 ... 1 j n i ji j empl n x I CSR ∑ = = where, CSR empl I j = Corporate social responsibility to employee index of the j th firm, where, j =1-17. i=1-16 n j = Total number of CSR to the employees’ items for the j th firm, n = 16, x ij = 1 if the i th item was disclosed, 0 if thei th item was not disclosed So that 0 ≤ CSR empl I j ≤ 1 Data for financial performance was measured using one accounting-based measure of ROA, which has the highestfrequency as a proxy used in the past to measure financial performance in studies on corporate social responsibility and financial performance relationship (Boaventura, Da Silva & De-Mello, 2012). The formula for financial performance is, ) ( 2 ... 100 × = ASSETS TOTAL TAX BEFORE PROFIT ROA Where, ROA=return on assets = dependent variable = Financial performance Data for proxies of control variables (age of the firm, leverage, and company Size) were obtained from the annual reports. The data needed for control variables were: age (number of years since listed on the Nigerian Stock Exchange); leverage (total debt and equity capital); for company size (logarithm of total assets). In this study, sequence charts of bar charts and the parametric inferential statistics of multiple regressions were found more appropriate to depict the trend in the independent and dependent variables and to test the study’s hypothesis respectively. The panel data procedure for regression analysis was adhered to, because multiple regressions conducted without having subjected the data to some diagnostic tests may lead to invalid results. Moreover, the choice of the regression model was robust pls (common effect). a) Model Specification for Multiple Regressions )(3 ........ 6 4 3 2 1 it it it it emplit it CZ LE AG CSR ROA ε β β β β α + + + + + = Where, b) Decision Rule P-value: we reject the null hypothesis, if the p-value is less than 5% (P<5%) otherwise, we do not accept the null hypothesis. Alternatively, at a 5% significance level, we reject the null hypothesis if |t|≥ 1.96 otherwise we accept the null hypothesis. The data was restricted to secondary data, which was extracted from annual reports and accounts of 17 listed consumer goods companies in Nigeria. The study covered a period of eleven years (2009-2019) yielding for the research, 187 observations, that is, 17 x 11. Data was scrutinized for missing values, but all the expected observations per constructs were intact . Followed by, was the detection of outliers. Outliers were detected in constructs (ROA, CSRempl, AG, LE, and CZ), (see Figure 3 below for outliers in ROA). Outliers were removed using the outlier removal code: new_y=@recode((y>@quantile(y,0.95))+(y<@quantile (y,0.05)),NA,y). Also, to get rid of NA (not available) or to retain the total number of observations in the sample data, AsifAhsan@which is the best method to remove outliers in a data set@www.researc hgate-net states that a straight forward way to remove outliers is first to identify the outlying observations and replace them with the median value. The researcher thus obtained the median values in the descriptive statistics for those data Effect of Employee Dimension of Corporate Social Responsibility on Financial Performance of Listed Consumer Goods Companies in Nigeria 6 Global Journal of Management and Business Research Volume XXIII Issue V Version I Year 2023 ( ) A © 2023 Global Journals i= company 1 to 17, t= the year 2009 to 2019 = 11 years, CSRempl it = Corporate Social Responsibility to employeefor company 1 to 17 over 11 years, ROA it = Return on Assets for company 1 to 17 over 11 years, α it = autonomous change in the dependent variable that is not explained by the independent variable for company 1 to 17 over 11 years, ẞ = the proportionate change in the dependent variable arising from a unit change in the independent variable, AG it = Age of the firm = the first control variable for company 1 to 17 over 11 years, LE it = Leverage= the second control variable for company 1 to 17 over 11 years, CZ it = Company Size= the third control variable for company 1 to 17 over 11 years, ε i it = assumed error magnitude of the predictive variable(s) in explaining the criterion variable or the error term that accounts for other factors affecting the dynamics of the dependent variable not captured in the model for company 1 to 17 over 11 years.