Hypothesis Testing

Discuss about the Evaluate The Differences In The Carbon Disclosure Scores.

The main initiative of this research is to evaluate the differences in the carbon disclosure scores between different company sectors and different countries. Further the relationship between the company factors on carbon emission on the disclosure scores are also of interest to evaluate. Thus, in order to find out the stated concerns, the following hypothesis can be framed:

Null Hypothesis (H₀₁): There are no significant differences in the carbon disclosure scores with respect to different sectors of companies.

Alternate Hypothesis (H_A1): There are significant differences in the carbon disclosure scores with respect to different sectors of companies.

Null Hypothesis (H₀₂): There are no significant differences in the carbon disclosure scores with respect to disclosure status.

Alternate Hypothesis (H_A2): There are significant differences in the carbon disclosure scores with respect to disclosure status.

Null Hypothesis (H₀₃): There are no significant relationship between the carbon emission conditions of the companies with their disclosure scores.

Alternate Hypothesis (H_A3): There are significant relationship between the carbon emission conditions of the companies with their disclosure scores.

In order to test the first and the second hypothesis, chi square test of association has been conducted and a correlation and regression analysis has been conducted in order to test the third stated hypothesis.

The results of the chi square test for testing the first hypothesis are provided in table 1. It can be seen from the table that the value of the chi square statistic with 105 degrees of freedom has been obtained as 116.7 with a significance value of 0.205. The significance value of the chi-square test is higher than the estimated level of significance (0.05). This indicates that the test of association is insignificant (Seber 2015). Thus, the null hypothesis (H₀₁) is rejected.

The results of the chi square test for testing the second hypothesis are provided in table 2. It can be seen from the table that the value of the chi square statistic with 21 degrees of freedom has been obtained as 60.000 with a significance value of 0.000. The significance value of the chi-square test is lower than the estimated level of significance (0.05). This indicates that the test of association is significant (Olive 2014). Thus, the null hypothesis (H₀₂) is accepted.

Table 1: Chi-Square Tests for different Sectors
	Value	df	Asymp. Sig. (2-sided)
Pearson Chi-Square	116.700a	105	.205
Likelihood Ratio	96.181	105	.719
Linear-by-Linear Association	1.436	1	.231
N of Valid Cases	60
a. 132 cells (100.0%) have expected count less than 5. The minimum expected count is .10.

Table 2: Chi-Square Tests for Disclosure Status
	Value	df	Asymp. Sig. (2-sided)
Pearson Chi-Square	60.000a	21	.000
Likelihood Ratio	23.822	21	.302
Linear-by-Linear Association	36.604	1	.000
N of Valid Cases	60
a. 41 cells (93.2%) have expected count less than 5. The minimum expected count is .05.

 

In order to test the third hypothesis, which is to establish the relationship between the variables disclosure scores and the carbon emission reduction initiatives of the companies, a correlation and regression analysis has been performed. It can be seen from the correlation matrix in table 3 that the dependent variable disclosure scores have a positive relationship with all the independent variables except with the first independent variable, with which there is a negative relationship (Brandt 2014).

Table 3: Correlations Matrix
	Disclosure_Scores_2015	IV1	IV2	IV3	IV4	IV5	IV6
Disclosure_Scores_2015	Pearson Correlation	1	-.177	.028	.079	.215	.181	.209
Sig. (2-tailed)		.176	.831	.550	.100	.166	.108
N	60	60	60	60	60	60	60
IV1	Pearson Correlation	-.177	1	-.082	-.016	.055	.294*	.055
Sig. (2-tailed)	.176		.534	.900	.677	.023	.674
N	60	60	60	60	60	60	60
IV2	Pearson Correlation	.028	-.082	1	.132	-.349**	-.341**	-.352**
Sig. (2-tailed)	.831	.534		.313	.006	.008	.006
N	60	60	60	60	60	60	60
IV3	Pearson Correlation	.079	-.016	.132	1	.060	.057	.057
Sig. (2-tailed)	.550	.900	.313		.648	.667	.667
N	60	60	60	60	60	60	60
IV4	Pearson Correlation	.215	.055	-.349**	.060	1	.954**	1.000**
Sig. (2-tailed)	.100	.677	.006	.648		.000	.000
N	60	60	60	60	60	60	60
IV5	Pearson Correlation	.181	.294*	-.341**	.057	.954**	1	.954**
Sig. (2-tailed)	.166	.023	.008	.667	.000		.000
N	60	60	60	60	60	60	60
IV6	Pearson Correlation	.209	.055	-.352**	.057	1.000**	.954**	1
Sig. (2-tailed)	.108	.674	.006	.667	.000	.000
N	60	60	60	60	60	60	60
*. Correlation is significant at the 0.05 level (2-tailed).
**. Correlation is significant at the 0.01 level (2-tailed).

Chi Square Test

The effect of the independent variables on the dependent variable cannot be estimated from the correlation analysis and thus, regression analysis has been performed. The results are given in tables 4, 5 and 6. The R Square value and the F statistic indicate the goodness of fit of the model. From the R Square value, it is clear that 17 percent of variability in independent variable can be explained by dependent variables which is very low (Sullivan III 2015).

Table 4: Regression Model Summary
Model	R	R Square	Adjusted R Square	Std. Error of the Estimate
1	.421a	.177	.084	24.171
a. Predictors: (Constant), IV6, IV1, IV3, IV2, IV5, IV4

The value of the F-statistic in table 5 comes with a significance value of 0.097 which indicates insignificance. Thus, the cumulative impact of the independent variables on the dependent variable is insignificant (Draper and Smith 2014).

Table 5: ANOVAa
Model	Sum of Squares	df	Mean Square	F	Sig.
1	Regression	6682.139	6	1113.690	1.906	.097b
Residual	30963.861	53	584.224
Total	37646.000	59
a. Dependent Variable: Disclosure_Scores_2015
b. Predictors: (Constant), IV6, IV1, IV3, IV2, IV5, IV4

The coefficients of the independent variable have been found to be insignificant except for the effect of the independent variables 4 and 6. Thus, the other 4 independent variables can be eliminated from the model (Chatterjee and Hadi 2015).

Table 6: Regression Coefficientsa
Model	Unstandardized Coefficients	Standardized Coefficients	t	Sig.
B	Std. Error	Beta
1	(Constant)	81.291	8.980		9.052	.000
IV1	-5.064E-005	.000	-.416	-1.958	.055
IV2	1.784	5.814	.042	.307	.760
IV3	7.450E-011	.000	.005	.038	.970
IV4	9.576	4.670	13.034	2.051	.045
IV5	.633	.461	.973	1.372	.176
IV6	-10.084	4.755	-13.713	-2.121	.039
a. Dependent Variable: Disclosure_Scores_2015

From the chi square analysis conducted, it can be said that the carbon disclosure scores of the companies does not depend on which type of sector they belong to but the disclosure scores have a significant relationship with countries. Thus, the carbon disclosure scores differ at different countries. This might be due to the difference in the climates at different regions. Thus, it can be said that climatic change is a very significant factor in evaluating the carbon disclosure scores for different companies (Gary, Saunders and Goregaokar 2012). The increase in the disclosure scores will be putting the shareholders of the companies at a lesser risk of investing in the companies. This will in turn be a benefit for the agency. Thus, it can be said that the agency theory is not directly related to the carbon disclosure scores but is related with the help of the stakeholder theory to the carbon disclosure scores.

The correlational analysis has found significant positive relationship between the pre-defined independent variables except for the additional intensity of carbon emission figures, with which negative relationship has been obtained (Macdonald and Headlam 2010). This indicates that the factors that are considered here are helpful in increasing the carbon disclosure scores (Guo 2014). Thus, considering these factors and taking actions accordingly will be beneficial for the companies (Orsato 2017).

The regression analysis has shown that the scope 2 figures and the direction of change in the carbon emission this year from the previous year has been of significant effect in predicting the carbon disclosure scores (Winn et al. 2011). The positive impact of these two factors will be helpful in increasing the carbon disclosure scores and will in turn be of benefit to the companies as well as to the environment in reducing the environmental pollution (Saunders, Lewis and Thornhill 2009).

Correlation Analysis

In this study, the effects of climatic changes have not been considered which has been identified as one of the most important factors responsible for the differences in the carbon disclosure scores of the companies. The research was mainly aimed at finding out the factors that will be useful in reducing the carbon emissions of the companies. The factors that can be responsible for this change has not been put into light properly. Moreover, only 60 companies have been selected for this study which is a very little number as compared to the total number of industries that are present in this whole world. Thus, a sample of this size might not be sufficient in interpreting the correct effects of the selected factors on the carbon disclosure scores and in reducing carbon emissions.

The matter of finding out other important factors has to be evaluated in a much detailed manner to find out the factors that will be responsible for the carbon emissions and so that the company can take necessary measures to have control over all those issues. So far, some companies must have already adopted necessary measures in order to reduce the carbon emissions. The emissions of the companies before and after adopting the measures can also be assessed. The difference in the emissions before and after adoption of the techniques can be tested using appropriate techniques. This will be helpful in indicating whether the factors adopted are sufficient in reducing the emissions or some other factors are still responsible and have not been put into light. This will give a better interpretation to the study conducted in this research.

References

Brandt, S., 2014. Testing Statistical Hypotheses. In Data Analysis (pp. 175-207). Springer, Cham.

Chatterjee, S. and Hadi, A.S., 2015. Regression analysis by example. John Wiley & Sons.

Draper, N.R. and Smith, H., 2014. Applied regression analysis (Vol. 326). John Wiley & Sons.

Gary, D.E., Saunders, M.N. and Goregaokar, H., 2012. Success in challenging times: Key lessons for UK SMEs. Surrey: University of Surrey.

Guo, Y., 2014. Climate Change Disclosure?: Determinants and impact. University of Hawai.

Macdonald, S. and Headlam, N., 2010. Research Methods Handbook. Manchester.

Olive, D.J., 2014. Testing Statistical Hypotheses. In Statistical Theory and Inference (pp. 183-213). Springer, Cham.

Orsato, R. J., 2017. Organizational adaptation to climate change: learning to anticipate energy disruptions. International Journal of Climate Change Strategies and Management, 9(5), 645–665.

Saunders, M., Lewis, P. and Thornhill, A., 2009. Research methods for business students (5th editio). Harlow: Pearson Education Limited.

Seber, G.A., 2015. Testing Several Hypotheses. In The Linear Model and Hypothesis (pp. 73-101). Springer, Cham.

Sullivan III, M., 2015. Fundamentals of statistics. Pearson.

Winn, M., Kirchgeorg, M., Griffiths, A., Linnenluecke, M. K. and Günther, E., 2011. Impacts from climate change on organizations: a conceptual foundation. Business Strategy and the Environment, 20(3), 157–173.