• No results found

Research methods of phase one: Questionnaire survey stage

In document Empirical evidence from Sri Lanka (Page 91-97)

Research methodology and methods

3.5. Research methods

3.5.2. Research methods of phase one: Questionnaire survey stage

This section explains the sample selection strategy for the questionnaire survey followed by the data collection process, questionnaire instrument, and the methods of analysis.

3.5.2.1. Sample selection strategy for the questionnaire survey

Selecting a sample is very important for any research study: “We cannot study everyone everywhere doing everything. Sampling decisions are required not only about which people to interview or which events to observe, but also about settings and processes” (Punch, 1998, p.193). Selecting the organisations for the survey and selecting the people to interview from the selected organisations involves sampling issues. Punch (1998) emphasises that there is no simple way of summarising sampling strategies for research, since it involves a variety of research approaches, purposes, and settings. Some authors have discussed and typologically analysed sampling strategies that can be used in research studies (Bryman, 2012; Johnson, 1990; Miles & Huberman, 1994). However, all these sampling strategies vary considerably and the appropriate approach depends on the purpose and the research objective/questions guiding the study (Bryman, 2012).

Research methods

76

Thus, the sample selection strategy needed to be selected to achieve the objective of this thesis: to explore the nature and extent of CSR practices in Sri Lanka. To achieve this research objective, it is first necessary to identify companies who are likely to engage in CSR practices, both operationally and in disclosure. Therefore, the researcher believes that the sampling strategy for the survey questionnaire has to be purposive sampling because then the researcher can select the companies who are likely to engage in CSR practices.

3.5.2.2. Data collection process and responses for the questionnaire survey There is a general tendency for large companies to carry out CSR activities and the empirical evidence suggests that there is a positive link between the size of companies and level of CSR practice (Adams & Harte, 1998; Andrew, Gul, Guthrie, & Teoh, 1989; Collins, Lawrence, Roper, & Haar, 2010; Fernando &

Pandey, 2012). This link indicates that large companies are more likely to engage in CSR activities and reporting. Therefore, based on market capitalisation, the top 200 companies listed in the CSE were selected for the questionnaire survey for this thesis. From these 200 companies, 51 usable questionnaires were received.

The companies were requested to have the questionnaire completed by a corporate manager who is directly involved in CSR disclosure and/or the decision making process.

The questionnaires (See Appendix 8.) were posted from New Zealand on 26th February 2010 to the registered head offices of the sample 200 companies, together with a cover letter (See Appendix 5.), information sheet (See Appendix 6.), introduction letter from the chief supervisor (See Appendix 7.), and a stamped envelope with a Sri Lankan return address. By 26th March 2010, 20 questionnaires were received by post, 31 questionnaires were collected by visiting companies individually, and 2 questionnaires were received through email as attachments, totalling 53 questionnaires. Two out of 53 questionnaires were considered as unusable as they were incomplete and were removed from the analysis. Therefore,

Research methods

77

the questionnaire survey finally used 51 usable questionnaires for the analysis after achieving a response rate of 25.5 per cent. For a general overview of the respondents, the industry classification is given in Table 3.5-1.

Table 3.5-1: Classification of responded company

Classification Companies Percentage

Bank Finance & Insurance 12 24%

Manufacturing 7 14%

Diversified Holdings 6 12%

Beverage Food & Tobacco 4 8%

Motors 4 8%

Plantations 4 8%

Telecommunication 4 8%

Other 10 20%

Total 51

Note: The ‘Other’ category includes: Hotels and Travels, Construction and Engineering, Chemicals and Pharmaceuticals, Land and Property, Power and Energy, and Stores and Supplies

3.5.2.3. The purpose, nature and scope of the survey instrument

The survey instrument needs to serve the purpose of survey study, which in this case is to explore general CSR practices in the Sri Lankan corporate sector and to compare the results with those from a developed country. In achieving this purpose, two options were considered: constructing a new survey instrument, or adopting a survey instrument, which was used for another study, with some modifications, if needed.

The researcher favoured the latter alternative as he found a CSR study (Collins et al., 2010) for New Zealand the results of which could be used for comparison with Sri Lanka. The researcher evaluated the suitability and possibility of adopting the questionnaire of the New Zealand study (Collins et al., 2010). The survey questions of the New Zealand study were constructed based on a review of literature; the wordings were in a simple and understandable form; and the scope was suited to

Research methods

78

the purpose of this thesis. Thus, the questionnaire, which has been used by the multiyear ongoing New Zealand research project (Collins et al., 2010), was used with the researchers’ permission and with slight modifications to adjust it to the Sri Lankan context (See Appendix 8 for the survey questionnaire.). For example, in obtaining information about company affiliations with accounting institutions, the possible answers are adjusted according to the Sri Lankan professional accounting qualifications and other international accounting qualifications that are popular in Sri Lanka.

Based on a review of the CSR literature, Collins et al. (2010) identified the most common environmental and social practices adopted by CSR reporting companies in constructing the survey questions. The survey questions drew on five areas of CSR practices: environmental practices; social practices; internally and externally perceived pressure towards adopting CSR practices; barriers and drivers; and, management perceptions about the future of CSR practices. To explore each of these areas, simple and close-ended questions were formed. Most of the survey questions were designed to obtain “yes” or “no” answers. For example, to query respondents’ CSR practices the following types of questions were asked: Does your business engage in a recycling programme? Does your company produce a public environmental and/or sustainability report? Does your company develop product and service innovations based on environmental benefits? Does your company provide job training? Does your company give time, money, products or services to local community projects? Does your company have ethical purchasing policies?

In addition to these yes/no questions, two likert scale questions were employed to examine the management perceptions about the future of CSR practices.

Through Collins et al.’s (2007; 2010) research project, past and present data were available for the New Zealand context and they could be used for comparison purposes with the Sri Lankan context, especially in answering research question

Research methods

79

number two: To what extent are the Sri Lankan CSR practices different from CSR practices in New Zealand, a developed country? Chapter Seven of the thesis reports this comparative study.

3.5.2.4. Data analysis methods for the questionnaire survey

Primary data collected through the survey instrument are analysed based on size, ownership, and geographical scale of operations in order to uncover striking phenomena of CSR practices. PASW Statistics v.18 software (previously SPSS) is employed for the statistical analysis. Descriptive statistics are used for overall and sectional analysis and Kruskal-Wallis one-way ANOVA and Mann-Whitney U tests are used for hypotheses testing as the data are not normal. These nonparametric tests are used to identify whether the average CSR practices of investigated variables used in this survey are identical for each respondent category. Therefore, validity of the following null hypotheses is tested using the above nonparametric tests:

H0:The average CSR practices of the issues or variables under investigation are not different for each respondent category, which is size, ownership, and geographical scale of business.

A 95 per cent confidence level/0.05 level of significance has been chosen to test the null hypotheses. This level of confidence is generally regarded as an arbitrarily acceptable standard in social science (Mashat, Ritchie, Lovatt, & Pratten, 2005). If the calculated level of significance in these tests is equal to or less than 0.05, then it indicates that at least one of the categories is different from at least one of the others.

For analysis purposes the respondents are classified into three categories: size, ownership, and geographical scale of operations, in order to test whether there are significant differences between these groups pertaining to CSR practices.

The respondent companies are classified into small, medium, and large, based on number of employees: small 0 to 249; medium 250 to 999; and large 1000 or more.

Research methods

80

There are 13 companies in each of the small and medium categories and in the large category there are 25 companies. The ownership of the respondent companies is classified based on foreign and Sri Lankan ownership, and there are 17 foreign-owned and 34 Sri Lankan-foreign-owned respondents. The respondent companies are classified as to whether they provide their services and/or products only to the local market or not. There were 30 respondent companies with local scale operations and 21 respondent companies with international scale operations.

In the case of questionnaire survey data analysis, environmental practices and social practices are surveyed based on the involvement of different environmental and social activities. In order to get an overall indication about these practices, Environmental Activity Percentage Score (EAPS) value and Socially-related Activity Percentage Score (SAPS) value are calculated to represent overall environmental practices and overall social practices respectively.

3.5.2.5. Environmental Activity Percentage Score (EAPS) value and Socially-related Activity Percentage Score (SAPS) value

The Environmental Activity Percentage Score (EAPS) value is calculated to measure the organisational involvements with environmental activities while the Socially-related Activity Percentage Score (SAPS) value is calculated to measure the organisational involvements with socially-related activities. The higher the score value, the greater would be the engagement with CSR activities by the companies.

Fourteen doable environmental activities are indicated in the questionnaire and respondents are asked whether they are engaged in those activities or not. For each activity the individual percentage of engagement is calculated and also EAPS value is calculated by getting the simple average of these individual percentage values.

Similarly, the SAPS value is calculated based on 10 doable socially-related activities.

EAPS values and SAPS values are calculated for each companies category, such as small, medium, large, foreign-owned, Sri Lankan-owned, international scale, and local scale companies in order to measure the extent of these company categories’

engagement in CSR activities.

Research methods

81

In document Empirical evidence from Sri Lanka (Page 91-97)