4.1. Descriptive Statistics
Table 3 provides the standard descriptive statistics characterizing the data, as well as the mean value, standard deviation for numerical variables, and minimum and maximum values.
Table 3
Variable | Obs | Mean | Std. Dev. | Min | Max |
Environmental | 308 | .247 | .184 | 0 | .722 |
Social | 308 | .283 | .201 | 0 | .75 |
Governance | 308 | .637 | .254 | 0 | 1 |
LogAsset | 308 | 7.791 | .786 | 5.881 | 9.916 |
FirmAge | 308 | 30.773 | 14.985 | 7 | 78 |
Service | 308 | .205 | .404 | 0 | 1 |
Investment | 308 | .227 | .42 | 0 | 1 |
Industry | 308 | .25 | .434 | 0 | 1 |
Banks | 308 | .136 | .344 | 0 | 1 |
Insurance | 308 | .159 | .366 | 0 | 1 |
FinancialService | 308 | .023 | .149 | 0 | 1 |
ROA | 308 | .21 | .066 | 0 | .515 |
ROE | 308 | .441 | .123 | 0 | .863 |
AuditQ | 308 | .711 | .454 | 0 | 1 |
BordSize | 308 | 0 | 1 | -2.188 | 3.012 |
Bdivers | 308 | 0 | 1 | − .297 | 9.225 |
Table 3 shows the findings of the descriptive analysis of this study's dependent, independent, control, and moderating variables. Based on the above table, all variables consist of 308 observations. The mean values for the environmental, social, and governance variables indicate the average level of disclosure or performance related to these ESG factors among the sample of Palestinian companies. The wide range of values, with minima of 0 and maxima close to 1, suggests considerable variability in ESG practices among the companies studied. This variability could reflect differences in organizational priorities, stakeholder engagement, regulatory compliance, and industry norms. This variation may be relevant to the extent to which companies prioritize environmental sustainability, social responsibility, and effective governance practices. The relatively high standard deviations for both ROE and ROA indicate a significant variability in FP across the sample. The variable (FirmAge) exhibits higher standard deviations than the other variables, indicating diversity among the sample companies in terms of age and possibly experience in the market. The relationship between environmental, social, and governance factors and FP is affected by the outliers, trends, and patterns represented by this dispersion.
4.2. Diagnostic tests
This study attributes an exogenous assessment and authentically does not vouch for endogeneity issues of the variables. Based on these factors, the assumption of exogeneity is that independent variables are orthogonal to the error term this helps to minimize endogeneity concerns. Often, the assumption of being absent is considered in most regression analyses, especially when the independent variable is believed to not be the cause of the error terms and is not affected by them [78]. In Table 4 are the results of endogeneity testing for 6 models.
Table 4
| Model 1 | Model 2 | Model 3 | Model 4 | Model 5 | Model 6 |
Durbin (score) chi2(1) Wu-Hausman F(1,303) | 0.4296 0.4833 | 0.2477 0.2514 | 0.4896 0.4933 | 0.3320 0.3359 | 0.2293 0.2329 | 0.6969 0.6995 |
Table 4 suggest that there is no endogeneity in the models based on the given p-values so the null hypothesis of H0: Variables are exogenous is rejected. Therefore, they are not correlated with the error term and do not suffer from endogeneity issues in this context.
This study used skewness and kurtosis; the models indicate a p-value of 0.000, which is below 0.05, indicating that the normality hypothesis was not satisfied. Gujarati [79] showed that if the number of observations exceeds 30, the normal distribution is not considered a problem with the regression equation and can be ignored; this study has 308 observations.
The study used variable correlation matrix; the Spearman’s correlation test was used to assess the multiple regression model to ensure no strong correlations between the independent variables. A high correlation had a correlation rate of more than 80% between two or more variables. This can distort the link between one of two independent variables [79]. A cross-correlation matrix between the research variables was created to confirm that there were no correlations of this type, as shown in Tables 5 and 6.
The results in Tables 5 and 6 indicate that the highest correlation coefficient was (0.637) between the (social) and (environmental) variables, indicating that there was no interference between these variables in the regression models. The rho values of -0.131 in both tables suggest a lack of substantial correlation between the variables, reinforcing the absence of multi-collinearity concerns.
In addition to the correlation matrix, multi-collinearity (VIF) was performed for the models. A VIF coefficient exceeding 10 suggests multi-collinearity [80]. After conducting this analysis, it was found that the VIF values for the (investment) variable were greater than 10; therefore, they were deleted. Tables addressing VIF and normality were omitted for simplicity.
Table 5
Multi-collinearity Test for Model (ROE)
Variables | (1) | (2) | (3) | (4) | (5) | (6) | (7) | (8) | (9) | (10) | (11) | (12) | (13) | (14) | (15) |
(1) ROE | 1.000 | |
(2)Environmental | 0.243 | 1.000 | |
(3) Social | 0.233 | 0.637 | 1.000 | |
(4) Governance | 0.347 | 0.525 | 0.613 | 1.000 | |
(5) LogAsset | 0.095 | 0.389 | 0.513 | 0.480 | 1.000 | |
(6) FirmAge | 0.156 | 0.046 | 0.040 | -0.021 | -0.058 | 1.000 | |
(7) Service | -0.124 | -0.210 | -0.188 | -0.308 | -0.140 | -0.232 | 1.000 | |
(8) Industry | 0.190 | 0.104 | 0.158 | 0.070 | -0.323 | 0.428 | -0.293 | 1.000 | |
(9) Banks | -0.114 | 0.186 | 0.507 | 0.292 | 0.556 | -0.089 | -0.201 | -0.229 | 1.000 | |
(10) Insurance | 0.260 | 0.055 | 0.136 | 0.167 | 0.142 | -0.025 | -0.221 | -0.251 | -0.173 | 1.000 | |
(11) Financial Serves | -0.036 | -0.086 | 0.012 | -0.017 | -0.156 | -0.030 | -0.077 | -0.088 | -0.061 | -0.066 | 1.000 | |
(12) AuditQ | -0.055 | 0.017 | 0.218 | 0.171 | 0.359 | -0.207 | 0.075 | 0.021 | 0.253 | -0.212 | 0.097 | 1.000 | |
(13) Investment | -0.199 | -0.075 | -0.520 | -0.154 | -0.055 | -0.114 | -0.275 | -0.313 | -0.215 | -0.236 | -0.083 | -0.150 | 1.000 | |
(14) BordSize | 0.051 | 0.289 | 0.367 | 0.344 | 0.533 | 0.045 | -0.071 | -0.111 | 0.377 | -0.059 | -0.136 | 0.232 | -0.026 | 1.000 | |
(15) Bdivers | -0.101 | 0.092 | 0.092 | 0.045 | -0.030 | 0.115 | -0.011 | -0.006 | 0.170 | -0.103 | -0.118 | -0.168 | 0.009 | -0.131 | 1.000 |
Spearman rho = -0.131 |
Table 6
Multi-collinearity Test for Model (ROA)
Variables | (1) | (2) | (3) | (4) | (5) | (6) | (7) | (8) | (9) | (10) | (11) | (12) | (13) | (14) | (15) |
(1) ROA | 1.000 | |
(2)Environmental | 0.180 | 1.000 | |
(3) Social | 0.058 | 0.637 | 1.000 | |
(4) Governance | 0.201 | 0.525 | 0.613 | 1.000 | |
(5) LogAsset | -0.121 | 0.389 | 0.513 | 0.480 | 1.000 | |
(6) FirmAge | 0.210 | 0.046 | 0.040 | -0.021 | -0.058 | 1.000 | |
(7) Service | -0.007 | -0.210 | -0.188 | -0.308 | -0.140 | -0.232 | 1.000 | |
(8) Industry | 0.302 | 0.104 | 0.158 | 0.070 | -0.323 | 0.428 | -0.293 | 1.000 | |
(9) Banks | -0.364 | 0.186 | 0.507 | 0.292 | 0.556 | -0.089 | -0.201 | -0.229 | 1.000 | |
(10) Insurance | 0.079 | 0.055 | 0.136 | 0.167 | 0.142 | -0.025 | -0.221 | -0.251 | -0.173 | 1.000 | |
(11) Financial Serves | 0.065 | -0.086 | 0.012 | -0.017 | -0.156 | -0.030 | -0.077 | -0.088 | -0.061 | -0.066 | 1.000 | |
(12) AuditQ | -0.070 | 0.017 | 0.218 | 0.171 | 0.359 | -0.207 | 0.075 | 0.021 | 0.253 | -0.212 | 0.097 | 1.000 | |
(13) Investment | -0.099 | -0.075 | -0.520 | -0.154 | -0.055 | -0.114 | -0.275 | -0.313 | -0.215 | -0.236 | -0.083 | -0.150 | 1.000 | |
(14) BordSize | -0.049 | 0.289 | 0.367 | 0.344 | 0.533 | 0.045 | -0.071 | -0.111 | 0.377 | -0.059 | -0.136 | 0.232 | -0.026 | 1.000 | |
(15) Bdivers | -0.115 | 0.092 | 0.092 | 0.045 | -0.030 | 0.115 | -0.011 | -0.006 | 0.170 | -0.103 | -0.118 | -0.168 | 0.009 | -0.131 | 1.000 |
Spearman rho = -0.131 |
4.3. Linear regression
As indicated in the tables below, a linear regression model was employed to conduct the regression analysis, demonstrate the impact of the independent factors on the dependent variable, and confer or reject the hypotheses developed earlier.
Table 7
ESG Pillars's Influence on ROE
Model | 1 | | | 3 | | | 5 | | | |
ROE | Coef. | t-value | p-value | Coef. | t-value | p-value | Coef. | t-value | p-value | |
Environmental | .125 | 3.15 | .002*** | | | | | | | |
Social | | | | .164 | 3.09 | .002*** | | | | |
Governance | | | | | | | .135 | 4.29 | 0*** | |
LogAsset | .024 | 1.68 | .095* | .02 | 1.38 | .167 | .019 | 1.42 | .155 | |
FirmAge | 0 | 0.50 | .615 | 0 | 0.73 | .468 | .001 | 1.13 | .258 | |
Service | .023 | 1.21 | .226 | − .006 | -0.28 | .783 | .017 | 0.90 | .367 | |
Industry | .085 | 3.76 | 0*** | .042 | 1.38 | .169 | .06 | 2.51 | .013** | |
Banks | .003 | 0.11 | .914 | − .049 | -1.74 | .083* | − .014 | -0.56 | .576 | |
Insurance | .076 | 3.80 | 0*** | .029 | 1.18 | .238 | .052 | 2.56 | .011** | |
FinancialServie | .052 | 1.19 | .236 | .008 | 0.17 | .868 | .018 | 0.42 | .676 | |
AuditQ | − .04 | -2.41 | .017** | − .043 | -2.61 | .01*** | − .046 | -2.86 | .005*** | |
BordSize | − .005 | -0.68 | .5 | − .006 | -0.79 | .431 | − .01 | -1.37 | .173 | |
Bdivers | − .05 | -7.11 | 0*** | − .053 | -7.29 | 0*** | − .049 | -7.13 | 0*** | |
Constant | .207 | 2.08 | .039** | .248 | 2.37 | .019** | .191 | 2.00 | .047** | |
Mean dependent var | 0.441 | | | 0.441 | | | 0.441 | |
R-squared | 0.302 | | | 0.301 | | | 0.320 | |
F-test | 11.616 | | | 11.570 | | | 12.688 | |
Akaike crit. (AIC) | -503.044 | | | -502.678 | | | -511.498 | |
Prob > F | 0.000 | | | 0.000 | | | 0.000 | |
*** p < .01, ** p < .05, * p < .1 |
Table 7 shows the linear regression results for the study sample of (44) diversified Palestine companies listed on the PEX for the models associated with ROE.
It also shows that Model 1is positive and statistically significant at the (1%) level (F = 11.616), and the variables included in this multiple linear regression model explain 30% of the differences in the FP of firms listed on PEX for environmental disclosure. The regression analysis's findings show that the variables (Environmental, LogAsset, Sectors of Industry, Insurance) have a statistically significant positive influence on FP of firms listed on PEX but are negatively associated with (AuditQ and Bdivers) variables. The coefficient for the environmental pillar variable was 0.125, with a t-value of 3.15, and a statistically significant p-value of 0.002. This indicates that, for every one-unit increase in the environmental score, the ROE is expected to increase by 0.125 units, holding other variables constant. The significant coefficient suggests that environmental performance positively influences ROE, thus supporting the hypothesis that companies with strong environmental practices tend to have higher returns. In contrast, the remaining variables have no statistically significant influence on FP.
Model 3 is statistically significant at the (1%) level (F = 11.570). The variables included in this multiple linear regression model explain 30% of the differences in FP of companies listed on the PEX for social disclosure. The regression analysis findings show that the social disclosure variable statistically positively impacts the FP of companies listed on PEX, as indicated by the coefficient of 0.164, t-value of 3.09, and p-value of 0.002. This implies that companies that focus on social responsibility have a higher ROE. This aligns with expectations and suggests that firms with strong social initiatives may enjoy enhanced FP. By contrast, the variables (sector banks, AuditQ, and divers) had a significant negative influence. The remaining variables have no statistically significant influence on FP.
Model 5 is statistically significant at the (1%) level (F = 12.688), and the variables included in this multiple linear regression model explain 32% of the differences in the FP of companies listed on PEX. The Governance pillar variable demonstrates the strongest influence on ROE among the ESG factors, with a coefficient of 0.135, a t-value of 4.29, and a highly significant p-value of 0.000. This indicates that improvements in governance practices led to substantial increases in ROE. The regression analysis results show that the variables (Governance, Sectors of Industry, Insurance) positively impact the FP of firms listed on PEX, suggesting that firms operating in certain sectors tend to achieve higher returns.
In contrast, the variables (AuditQ and Bdivers) have an adverse statistical influence on FP, implying that factors such as higher audit quality and larger board sizes may exert downward pressure on returns. The remaining variables showed no statistically significant association with FP.
The overall model fit statistics show that the ROE remained constant across all three models at 0.441. This consistency suggests stability in the sample and indicates that on average, companies in the sample exhibit similar ROE values. R-squared values range from 0.301 to 0.320, indicating that approximately 30.1–32.0% of the variance in ROE is explained by the independent variables included in each model. While these percentages are not exceptionally high, they still imply a moderate level of explanatory power for the independent variables over the dependent variable. The F-test statistics were 11.616, 11.570, and 12.688 for Models 1, 3, and 5, respectively. The associated p-values were all below 0.001 (Prob > F = 0.000), indicating that the regression models were statistically significant at a high level of confidence. This suggests that the independent variables collectively have a significant impact on explaining the variance in ROE. The lower values of AIC ranging from − 503.044 to -511.498 indicate better-fitting models and suggest that the models have a good relative fit, with Model 5 having the lowest AIC value, indicating the best fit among the three models.
Table 8
ESG pillar's influence on ROA
Model | 2 | | | 4 | | | 6 | | | |
ROA | Coef. | t-value | p-value | Coef. | t-value | p-value | Coef. | t-value | p-value | |
Environmental | .078 | 3.83 | 0*** | | | | | | | |
Social | | | | .093 | 3.39 | .001*** | | | | |
Governance | | | | | | | .071 | 4.32 | 0*** | |
LogAsset | .006 | 0.80 | .424 | .005 | 0.66 | .51 | .006 | 0.79 | .431 | |
FirmAge | 0 | 1.02 | .31 | 0 | 1.14 | .257 | 0 | 1.45 | .15 | |
Service | .016 | 1.59 | .114 | − .001 | -0.08 | .934 | .012 | 1.26 | .208 | |
Industry | .046 | 3.93 | 0*** | .022 | 1.44 | .151 | .035 | 2.83 | .005*** | |
Banks | 0 | . | . | − .056 | -3.85 | 0*** | − .036 | -2.78 | .006*** | |
Insurance | .01 | 0.95 | .344 | − .017 | -1.36 | .175 | − .003 | -0.30 | .761 | |
FinancialServie | .037 | 1.67 | .096* | .013 | 0.53 | .595 | .021 | 0.91 | .365 | |
AuditQ | − .015 | -1.72 | .087* | − .017 | -2.00 | .047** | − .019 | -2.28 | .024** | |
BordSize | − .002 | -0.58 | .561 | − .003 | -0.66 | .511 | − .005 | -1.19 | .235 | |
Bdivers | − .026 | -7.21 | 0*** | − .028 | -7.33 | 0*** | − .025 | -7.06 | 0*** | |
Constant | .133 | 2.61 | .01*** | .151 | 2.80 | .005*** | .115 | 2.34 | .02** | |
Mean dependent var | 0.210 | | | 0.210 | | | 0.210 | |
R-squared | 0.352 | | | 0.345 | | | 0.360 | |
F-test | 14.611 | | | 14.189 | | | 15.148 | |
Akaike crit. (AIC) | -914.214 | | | -911.066 | | | -918.174 | |
Prob > F | 0.000 | | | 0.000 | | | 0.000 | |
*** p < .01, ** p < .05, * p < .1 |
Table 8 shows the linear regression results when Model 2 used is statistically significant at the (1%) level (F = 14.611). This reflects a significant positive impact on the FP (FP) of companies listed on the PEX, as indicated by the coefficient of 0.078, t-value of 3.83, and p-value of 0.000. The variables included in the multiple linear regression model explain 35% of the differences in the FP of companies listed on the PEX. The regression analysis findings show that the variables (environmental and sectors: Industry, Financial Service) statistically positively impact the FP of companies listed on PEX. Similarly, (AuditQ and Bdiver) has adverse statistical effects on the association between Environmental Disclosure and firm performance. This suggests that while environmental transparency can enhance FP, factors such as poor audit quality and excessive business diversification may hinder positive effects. The remaining variables have no statistically significant influence on FP.
Model 4 is statistically significant at the (1%) level (F = 14.189), and the variables included in the multiple linear regression model explain 35% of the differences in the FP of companies listed on the PEX. The regression analysis results show that the social disclosure variables have a statistically positive impact on the FP of companies listed on PEX. Simultaneously (banks, audit quality, and divers) have a statistically negative impact. This indicates that, while social responsibility initiatives contribute positively to FP, factors such as reliance on bank financing, lower audit quality, and excessive business diversification may undermine these benefits. The remaining variables have no statistically significant influence on FP.
Model 6 is statistically significant at the (1%) level (F = 15.148), and the variables included in the multiple linear regression model explain 36% of the differences in the FP of companies listed on PEX. The regression analysis findings also show that the variables (governance disclosure, Sector: Industry) have a statistically positive impact on the FP of companies listed on PEX. However, (Sector: Banks, Audit Q, and Bdivers) negatively influence firm performance. This underscores the importance of effective governance practices in enhancing FP while also highlighting the detrimental effects of certain industry sectors and governance-related issues. The remaining variables have no statistically significant influence on FP.
The overall model fit statistics show that ROA remains consistent across all models, at 0.210, indicating sample stability. The R-squared values range from 0.352 to 0.360, indicating that approximately 35.2–36.0% of ROA variance is explained by the independent variables in each model, implying good model fit. The F-tests yield statistically significant results with p-values below 0.001, indicating the collective significance of the independent variables in explaining ROA variance. Furthermore, the Akaike Information Criterion (AIC) values, ranging from − 914.214 to -918.174, suggest a good relative fit, with Model 4 exhibiting the lowest AIC value, indicating the best fit among the models.
In conclusion, the above statistics offer a high level of confidence in the accuracy and sufficiency of regression models as tools for demystifying the nature of the connection between different ESG viability aspects that become evident with regard to Palestine companies listed PEX. The results from this overall pattern of tests indicate that ESG exposures are significantly bankable for Palestine companies listed on the PEX. Although all ESG measurements showed a significant relationship with ROA, their power and effectuality varied. Fiscally profitable ESG performance critically depends on the transparency of ESG exposures, but these gains can be attenuated by extortionate bidding and poor audit quality; thus, cautious forceful governance frameworks and enhanced ESG disclosure can increase a company’s endurance by elevating financial stability.
Policymakers and stakeholders can use these findings to advocate for greater ESG transparency and accountability within the Palestine market, thereby promoting long-term value creation and economic stability.
4.4. Moderating effect of governance characteristics between ESG and FP
The primary hypothesis of this research pertains to the moderating role of governance features, specifically board diversity and size, between FP and ESG factors. There are multiple approaches to testing this moderating impact. The first is hierarchical regression, which is primarily utilized in the SPSS environment and is based on r-square changes. However, a more straightforward way to assess the moderating influence of factors is to use structural equation modeling such as STATA. The moderators in this study multiplied the independent factors to produce a moderator effect, which was examined for its moderating effect on FP. Tables 9 and 10 summarize the testing of the moderating effect, in which the code column is used to link the test to the literature.
Table 9
The Moderating Effect of governance characteristics between ESG and FP in (ROA)
| Board Diversity | Board Size |
| Model 7 | Model 8 | Model 9 | Model 10 | Model 11 | Model 12 | |
ROA | p- value | p- value | p- value | p- value | p- value | p- value | |
Environmental | 0 | | | .01 | | | |
Social | | .002 | | | .133 | | |
Governance | | | 0 | | | .001 | |
LogAsset | .438 | .395 | .313 | .852 | .811 | .761 | |
FirmAge | .314 | .205 | .083 | .256 | .262 | .232 | |
Service | .108 | .972 | .178 | .138 | .357 | .074 | |
Industry | 0 | .095 | .002 | 0 | .006 | 0 | |
Banks | .047 | 0 | .004 | 0 | .002 | 0 | |
Insurance | .272 | .276 | .888 | .382 | .924 | .689 | |
FinancialService | .072 | .417 | .2 | .078 | .297 | .147 | |
AuditQ | .086 | .079 | .022 | .182 | .139 | .015 | |
Bdivers | .185 | .801 | .12 | | | | |
BordSize | | | | .07 | .012 | .38 | |
EfSBdiversE | .377 | | | | | | |
EfSBdiversS | | .043 | | | | | |
EfSBdiversG | | | .004 | | | | |
EfSBordSizeE | | | | .093 | | | |
EfSBordSizeS | | | | | .01 | | |
EfSBordSizeG | | | | | | .308 | |
Constant | .004 | . .007 | .016 | .001 | .001 | .002 | |
R-squared | 0.353 | 0.353 | 0.375 | 0.245 | 0.244 | 0.255 | |
*** p < .01, ** p < .05, * p < .1 |
Table 10
The Moderating Effect of Governance Characteristics between ESG and FP in Model (ROE)
| Board Diversity | Board Size | | | Model 13 | Model 14 | Model 15 | Model 16 | Model 17 | Model 18 | | ROE | p- value | p- value | p- value | p- value | p- value | p- value | | Environmental | .004 | | | .057 | | | | Social | | .005 | | | .207 | | | Governance | | | 0 | | | .001 | | LogAsset | .096 | .11 | .105 | .273 | .412 | .375 | | FirmAge | .627 | .389 | .163 | .082 | .15 | .125 | | Service | .217 | .888 | .321 | .208 | .481 | .141 | | Industry | 0 | .1 | .005 | 0 | .01 | 0 | | Banks | .963 | .073 | .491 | .128 | .249 | .077 | | Insurance | 0 | .122 | .005 | 0 | .025 | .002 | | FinancialService | .18 | .63 | .418 | .169 | .512 | .311 | | AuditQ | .016 | .019 | .004 | .036 | .05 | .003 | | BordSize | | | | .179 | .009 | .313 | | Bdivers | .235 | .998 | .114 | | | | | EfSBdiversE | .321 | | | | | | | EfSBdiversS | | .023 | | | | | | EfSBdiversG | | | .004 | | | | | EfSBordSizeE | | | | .244 | | | | EfSBordSizeS | | | | | .005 | | | EfSBordSizeG | | | | | | .271 | | Constant | .021 | .024 | .046 | .005 | .003 | .006 | | R-squared | 0.303 | 0.311 | 0.335 | 0.186 | 0.197 | 0.207 | | *** p < .01, ** p < .05, * p < .1 | | |
Tables 9 and 10 show the moderating effect of Bdivers between the ESG component environmental disclosure and firm performance (ROE and ROA ). It is not a statistically significant model at prob > chi2 = (.377) for ROE and (.321) for ROA. Thus, drivers do not moderate the association between the ESG component of environmental disclosure and FP. Additionally, the moderating effect of Board Size on environmental and firm performance is not statistically significant at prob > chi2 = (.093) ROE and (.244) ROA. Thus, Board Size does not moderate the relationship between the ESG component Environment and FP.
From Tables 9 and 10, the moderating effect of Bdivers between the ESG component social disclosure and firm performance (ROE and ROA) is inspected. It is a statistically significant model at prob > chi2 equal (.043) for ROE, clarifying (0.353%) the variation in the FP and (.023) for ROA, and can clarify (0.311%) the variation in the FP. Thus, divers moderate the relationship between the ESG components of social disclosures and FP.
The moderating effect of board size on social effects and firm performance was also examined. It is a statistically significant model at prob > chi2 = (.01) for ROE, and can clarify (0.244%) of the variation in FP and. at prob > chi2 equals (.005) for ROA, highlighting its role in shaping this relationship. This can clarify the variation in FP (0.197%). Thus, board size moderates the influence of social disclosures and FP.
From Tables 9 and 10, the moderating effect of Bdivers on ESG component governance disclosure and firm performance is inspected. It is a statistically significant model at prob > chi2 equal (.004) and can clarify (0.335%) of the variation in ROE, and a statistically significant model at prob > chi2 equals (.004) and can clarify (0.375%) of the variation in ROA. Thus, Bdivers moderate the association between the disclosure of governance and FP. Additionally, the moderating effect of board size on governance disclosure and firm performance was examined. It is not a statistically significant model at prob > chi2 equal (.271) for ROE, and at prob > chi2 equals (.308) in Model ROA. Thus, BoardSize does not moderate the association between ESG component governance and FP.