In order to assess the effects of CSR on organizational performance, multiple regression analysis was conducted. A total of 179 questionnaires were distributed, and 170 questionnaires were returned. Among the 170 questionnaires returned back, four of them were discarded because they were not fully completed. After removing incomplete questionnaires and missing questionnaires, 166 complete questionnaires were useable (93% effective response rate). The collected data were presented and analyzed using SPSS (version 20) statistical software, and multiple regression analysis was also used to test the hypotheses of the study and the effects of the independent variables on the dependent variable.
To measure the consistency of the scores obtained, the study used Cronbach’s alpha (a measure of the internal consistency of the questionnaire items) by using data from all the respondents, and separate reliability tests for each of the variables were computed. Cronbach’s alpha measures the extent to which item responses obtained at the same time correlate highly with each other, and the widely accepted social science cut-off is that alpha should be greater than 0.70 for a set of items to be considered a scale [
81,
82,
83]. Accordingly, the Cronbach’s alpha test was carried out by using SPSS, and the results are presented in
Table A1 of
Appendix B.
Table A1 in
Appendix B shows the Cronbach’s alpha reliability statistics value of the scale for all predictor and outcome variables. The calculated coefficient alpha for this study was found to be 0.964 for all variables, which is greater than the required threshold of 0.70, confirming the variables to be internally consistent. Accordingly, all variables result in the table suggested that the threshold of the reliability test is > 0.7, which is statistically significant, and the data are reliable. Additionally, it can be seen that all of the four items contribute positively to yield the highest Cronbach’s alpha for the instrument. The total item result is shown in
Table A2 (item total for Cronbach’s alpha) in
Appendix B.
The results of
Table A3 in
Appendix B show the inter-item correlation, which gives the correlation coefficients for items in the study’s scale. Therefore, the study needed to use the inter-item correlation table to check if the items interrelate well, where the inter-item correlation must be greater than 0.3 [
80]. From the result of this study, the correlations between items for all variables are well above 0.3, which is encouraging.
4.2. Regression Analysis
The authors used multiple linear regression analysis to evaluate the effects of corporate social responsibility on organizational performance. To do so, we tested for the classical linear regression model assumptions for the model specified, as presented below. There are several assumptions that are compulsory to provide valid results in multiple linear regressions. Before running the analysis, we tested some of the basic assumptions of multiple linear regression, which are commonly advanced as follows.
Assumption 1. The relationship between the dependent and the independent variables should be linear.
Organizational performance is supposed to be linearly related to CSR dimensions, meaning that the dependent variable (organizational performance) is assumed to be affected by changes in CSR dimensions (the independent variables). The association between the two variables should be linear. This means that the scatter plot of scores should be a straight line, not a curve [
83]. The scatter plots of this study show that there is a virtually linear relationship between the variables, as indicated in
Figure A1 of
Appendix C. The plots do not show any evidence of non-linearity; therefore, the assumption of linearity is satisfied.
Assumption 2. Normality dispersion of the data should be normally distributed.
A normal distribution is a continuous, symmetric, bell-shaped distribution of a variable. The shape and position of a normal distribution curve depend on two parameters, the mean and the standard deviation [
86]. Thus, as indicated in
Figure A2 of
Appendix C, there is no problem of normality in this model, as the data are normally distributed around the mean.
Assumption 3. There should be no multi-collinearity or perfect correlation between independent variables.
Table A4 in
Appendix B shows the test for multi-collinearity, which is relatively easy to detect by calculating the tolerance or VIF (variance inflation factor) values. A tolerance value below 0.10 indicates a collinearity problem. The VIF is just the reciprocal value of the tolerance. Thus, VIF values above ten indicate collinearity issues. According to this measurement, none of the variables’ tolerance level is below 0.10 with their VIF above ten. Therefore, there is not a multi-collinearity problem with the variable.
Assumption 4. Assumption of no auto correlation.
Regression analysis is based on uncorrelated error/residual terms for any two or more observations [
83]. This assumption is tested for each regression procedure with the Durbin–Watson test, which tests for correlation between variables residuals. The test number can vary between 0 and 4, with a value of 2 meaning that the residuals are uncorrelated [
82]. A value greater than 2 indicates a negative correlation between adjacent residuals, whereas a value below 2 indicates a positive correlation.
As a general rule, the residuals are independent (not correlated) if the Durbin–Watson statistic is approximately 2, and an acceptable range is 1.50–2.50 [
87]. In this study, the Durbin–Watson value was 1.830, which is very close to 2; therefore, it can be confirmed that the assumption of independent error has almost certainly been met.
4.3. Regression Results
Regression analysis is one of the most frequently used tools in market research that allows analyzing relationships between independent and dependent variables. In this study, the regression model presents how much of the variance in the measure of organizational performance is explained by the underlying variables of corporate social responsibility, as indicated in
Table 5 below.
The R-value of the model summary in
Table 5 represents correlation and is 0.883 (the R-column), which indicates the high degree of correlation between the independent variables (legal, philanthropic, and ethical) and the dependent variable (organizational performance). That is 88.3%, indicating that there is a strong relationship between the dependent variable (organizational performance) and independent variables (legal, philanthropic, and ethical).
R-square (R-square column) indicates how much the total variation in the dependent variable (organizational performance) can be explained by the independent variables (legal activities, philanthropic activities, and ethical activities); in this case, the value of R-square is 77.9%, which is very strong. This means that about 78% of variance in the data can be explained by the variables; this amount is considerably large.
Adjusted R-square indicates the percentage of variance in the dependent variable or outcome variable explained by the independent variable or predictor variable. In this case, 77.5% of the variance in organizational performance can be explained by legal activities, philanthropic activities, and ethical activities. The three independent variables that were studied explained 77.5% of variance to investigate the effects of corporate social responsibility on organizational performance. This means that other factors that are not included in this research contribute about 22.5% of variance in the dependent variable. Therefore, further research should be conducted to investigate the effects of corporate social responsibility on organizational performance by including other variables that are not addressed in this study.
An ANOVA test shows how well the regression equation fits the data, that is, predicts the dependent variable and also shows an overall fitness of the model used. The last column in
Table A5 of
Appendix B (ANOVA-table) indicates the goodness of the model. From the result of F-test, it is known that the F statistic is 190.225 and the
p-value or the Sig. value is 0.000, which less than 5%. Thus F (3,162) = 190.225,
p < 0.05, and R = 0.779. This indicates that the overall regression model is statistically significant.
Table 6 below showed the regression analysis.
The unstandardized coefficient refers to the direction of the relationship and the amount of change that the different independent variables contribute to organizational performance given that one unit change in the value of the independent variable. The regressions coefficients that are shown in the above table indicate that the intercept, 29.83%, is representing the estimated average value of organizational performance when the philanthropic activities, ethical activities, and legal activities of the firm are zero. Thus, a firm with no philanthropic activities, ethical activities, or legal activities will have a severe influence on the performance of the firm.
The slop of independent variables, which is listed in beta value, also displays useful predictive information about the implication. The slop of philanthropic activities, ethical activities, and legal activities of the firm is 0.633, 1.099, and 0. 881, respectively, meaning that the firm’s performance changed or increased by 0.633, 1.099, and 0.881, respectively, for the three independent variables. This result is also supported by [
88,
89,
90]. Using the regression coefficients for independent variables and the constant term given under the column labeled B, the researchers constructed the regression equation for predicting firm performance as follows:
Y = β0+ β1X1 + β2X2 + β3X3 was interpreted to mean
Y = organizational performance;
X1 = philanthropic activities;
X2 = ethical activities;
X3 = legal activities.
Therefore, Y= 2.983 + 0.365X1 + 0.603X2 + 0.574X3.
According to the above equation, if all factors (philanthropic activities, ethical activities, and legal activities) remain constant at zero, overall organizational performance in the firms will be 2.983. The findings also show that a percentage increase in philanthropic activities will lead to a 36.5% increase in organizational performance of the firms; a percentage increase in ethical activity will lead to a 60.3% increase in organizational performance of the firms; and a percentage increase in legal activities will lead to a 57.4% increase in organizational performance of the firms. This means that the most significant variable is ethical activities, followed by legal activities, among the three independent variables.
4.5. Discussions of the Main Findings
This section presents the findings of the study in line with the objectives of the study based on the analysis made, and the results of this research are also further explained. The main objective of this study is to examine the effects of CSR practices on organizational performance in the case of the Dangote cement factory in Ethiopia. To this end, it is compulsory to know how much the CSR initiatives of the Dangote cement factory are known or observed by the respondents. Among employees’ responses, many of them believe that the Dangote cement factory takes part in CSR practices out of genuine interest in contributing to the wellbeing of society with respect to maximizing profit. The empirical findings are also combined with the theory in order to provide a wider standpoint.
The result indicates that little is known about the demographic profile of the Dangote cement factory that was analyzed through descriptive statistics. Accordingly, regarding the “sex” of the respondents, the study can conclude that most of the respondents (58%) were males. The Dangote cement factory seems dominated by males, and most of the age dispersal of respondents in the factory ranges from 18–30, which is too young. In addition, all of the respondents were educated, and also, they had been working in the factory for more than three years.
The result also shows CSR undertaken by Dangote cement factory through descriptive statistics. Accordingly, it was found that the majority of the respondents believed Dangote cement factory strongly encourages employees to actively participate in corporate social responsibility. The authors also found that the company pays attention to philanthropic issues, ethical issues, and legal issues as CSR to link them with its business performance. In other ways, the majority of respondents believed that the company’s level of awareness about CRS is high relative to other companies. Hence, the researchers can conclude that the majority of the respondents believe that CSR had positive effects on organizational performance.
Regarding the interview questions designed to assess the prior knowledge of the organizations, the manager of the Dangote cement factory suggested that the organization did practice CSR. The factory manager said after some briefings given by the interviewer that today issues related to CSR with respect to quality of the product/service are more commonly heard in the community, and suggested that this topic is very relevant and should be given high priority by organizations to make awareness of CSR practices and their impacts on their organizational performance.
The authors also analyzed the effects of philanthropic activities as CSR on organizational performance through descriptive analysis. The finding shows that most of the respondents do support those philanthropic activities. This means that philanthropy can be better through educational support like supports for private and public educational institutions and scholarships for individuals or groups and through facilitating managers and employees to participate in charitable activities. These findings relate to the findings of [
73]. From the findings of descriptive analysis, it can be noted that the majority of the respondents strongly agreed that ethical responsibilities have positive effects on organizational performance in that in the present context, the subject of organizational ethics has assumed a particular importance. This result is also supported by [
91]. The researcher also analyzes the effects of legal activities as CSR on organizational performance. From the findings, it can be noted that the majority of the respondents strongly agreed that legal responsibilities have positive effects on organizational performance; for example, the society has not only certified businesses as economic entities, but it has also established the minimal ground rules under which businesses are expected to operate and function. Businesses are expected and required to comply with these laws and regulations as a condition of operating. This result is also supported by [
91].
The factory participates in various CSR activities to conduct its business responsibly and dedicates a large amount of money to the cause. From an analysis of variables through the use of multiple regressions, it would be logical to deduct that other things being constant, there is the influence of CSR on the accomplishment of the company’s target performance. The regression analysis further strengthens this claim.
Accordingly, the coefficient of philanthropic responsibility was 0.365, which indicates that keeping other factors constant, a unit change in philanthropic responsibility causes a 36.5% increase in organizational performance and the effects of the independent factor (philanthropic responsibility), whose t-statistic value was found to be 3.860 at
p-value of 0.000, and that of the dependent variable (organizational performance) is found to be positive and statistically significant, which is supported by [
90,
91].
On other hand, ethical responsibility has a positive relationship with the dependent variable (organizational performance). The value of the coefficient for ethical responsibility was 0.603, and its significance level was 0.000, which is less than 0.05. This result indicates the value of b is positive and shows that ethical responsibility has a positive relationship with dependent variables and statistically significant effects on organizational performance. Additionally, this result is supported by [
90,
91,
92,
93].
The coefficient of legal activities was also computed to be 0.574, which means that a unit change in legal activities has the influence to increase organizational performance by 57.4%, assuming all other variables are constant. The calculated t-statistic value of this independent variable is 4.488 at
p-value of 0.000, which proves a positive and significant relationship with the dependent variable (organizational performance). Additionally, this result is supported by [
67,
90,
91].
The regression estimates also showed an adjusted R2 value of 0.775, indicating that about 77.5% of organizational performance of the firms under consideration was explained by the independent variable, while the remaining 22.5% variation in the organizational performance of the firms was caused by other factors not included in the model. This shows that the model has a good fitness of the regression line. The computed F-statistic showed a value of 190.225, while the
p-value was 0.000, which is less than the critical value of 0.05; as a result, the null hypothesis is rejected, while the alternate hypothesis is accepted. This indicates that CSR has a significant effect on the organizational performance of the firms. This result is consistent with the postulation of the stakeholder theory and also in line with the result of previous studies such as [
37,
38], which found that there is a significant relationship between CSR and organizational performance, meaning that CSR has a positive and significant impact on profitability [
94]. In general, based on the findings of this study the researcher tried to address the major aim of investigating how the corporate social responsibility feature would have an effect on organizational performance.