The purpose of data collection is to find, identify, and collect data about environmental matters within companies’ corporate reports and official Facebook pages. The basis for data collection is the timescale of the study: the years 2005, 2010-2014. Facebook activity is examined until April 2016, as all the latest entries by the firms are easily available so there was no reason to exclude them from the study. Not all the companies had provided their 2015 reports in time to be included in the study, and that is why the last examined reports are from 2014. That way the same number of reports is examined from all six companies.
In annual reports and sustainability reports, only the chapter or chapters containing environmental disclosure were selected for examination. A total number of 36 annual and sustainability reports were obtained, from which the appropriate chapters were then identified and examined. In some cases, if the report did not have clearly defined chapter/s about environmental assessment, the whole report was scanned through to identify the appropriate segments that were then examined.
The reports were downloaded from the companies’ official corporate websites, and they were in pdf format.
In order to simplify the data collection from social media, data collection units need to be defined. In Facebook, the unit is each post generated by the company. A post is an update of the channel and can take the form of status updates, links, events, discussions, photos, notes, or videos. If a post was added by Facebook, such as a summarizing post about the firm’s recent activity, it was not included. The Facebook entries from all the selected years were scanned, with the selection criteria being the inclusion of environmental themes/keywords, as noted earlier.
Only entries in English were selected, and due to time constraints and the availability of only one researcher, only one round of scanning was conducted. That yielded a total number of 137 entries for examination, a relatively low number considering the vast total number of entries posted by the companies. During the thorough reading of these selected entries, a further criterion was utilized: if the content of an entry is not relevant to the study and does not encompass any of the themes, regardless of it being initially selected after the scanning process, the entry was discarded. After all of them were read through, 16 of them were rejected, leaving a total of 121 entries to be fully examined.
As stated earlier, this study concentrates only on disclosures that fall under the environmental theme. Duff (2014) states that establishing which themes of disclosure are captured is important as it allows clear interpretation of the findings, and also ensures the replicability of the study. The environmental theme can be further classified into more specific categories under which the topics of the companies’ environmentally related activities fall into. That classification is presented in Chapter 4.5.2.
The complete list of themes can be seen in Table 8. The list contains both deductively and inductively derived themes, and their nearest and most often used synonyms in square brackets, which were identified with the help of an online WordNet tool. The list of inductively derived themes was collected while reading and examining the 36 annual and sustainability reports. This yielded a list of themes
that were not among the deductively derived themes. The range of topics covered in the reports is very wide as can be seen from the table.
Table 8: Themes used in the data analysis
Deductively derived themes Inductively derived themes Climate change
It should be noted that the list is by no means exhaustive, but it can be considered comprehensive for the purposes of this study, as it includes themes from both literature and the data itself. The deductive themes were adopted from Tewari’s (2012) and Walker’s and Wan’s (2012) researches. It has to be noted that Tewari examined top 100 companies in ICT industry operating in India, including both multinational and Indian firms. He does point out that the results are not generalizable to other industries or other countries. Walker and Fan researched top hundred Canadian companies in forest, energy, mining, and chemical industries – all visibly polluting industry sectors. They also state that including only Canadian firms in those specific industries is a limitation.
The important point considering this study is that the list is as encompassing as possible, and despite their limitations stated by the authors, the themes from those two articles are used without further discussing those limitations. The particular themes cover two of three industry sectors in which the selected companies in this study operate in: ICT and oil and gas industry, which is a subsector of energy industry. The inductive list can be considered relatively comprehensible, as it covers reports from three different industry sectors and from six different years, during which corporations’ environmental reporting has increased. With the inclusion of both deductively inductively derived themes in the analysis, the complete list can be considered sufficient for this study.