Content Of Csr Communications

Published Date: 02 Nov 2017

Introduction

The general research question of this dissertation is to examine to what extent international corporations have embraced the idea of CSR. This general research question is broken down into distinct aims:

To investigate the value and content of CSR communications.

To evaluate the corporate ethical stance in respect of the CSR communications.

To provide recommendations for the improvement of CSR communications.

The purpose of this chapter is to describe the methods used to collect and analyse the primary data required for this study. The aims require the collection of primary data during primary research to test the secondary research. Burns and Bush (2000) believe that secondary research can both familiarise the researcher with the study and identify concepts and terminology useful in conducting the primary research. However they also identify four possible disadvantages:

Mismatches in units of measurement

Different definitions of classifying data

The relevance of data can differ

There can be a possible inability of verification of data integrity.

These disadvantages have been minimised by only using recognised academic journals and books. Only reputable Internet sources have been used. The secondary research is described in the literature review, which provides an up-to-date comprehensive overview of CSR.

The methodology is defined as ‘the general approach to studying a research topic’ (Silverman, 1995), and it establishes how the research is to be carried out, it also provides terms of reference for other researchers to follow.

Manstead and Semin (1988, see Robson, 2002) point out that the strategies and tactics selected in carrying out a piece of research depend on the type of research question that is asked. It is therefore important to evaluate all of the main methodologies with regard to the aims of this dissertation before choosing one. This chapter will therefore discuss the advantages and disadvantages of the following:

Questionnaire or Survey

Interview

Observation

Case Study

Content Analysis

These methodologies can be separated into fixed or flexible design strategies. Robson (2002) defines fixed design strategies as those that call for a tight pre-specification before the main data collection stage is reached. Data is most often in the form of numbers, so these are often referred to as quantitative studies. Conversely, flexible designs evolve during the data collection stage, data is typically non-numerical, and these are often referred to as qualitative studies.

There are two main traditions of research, positivism and social constructionism. The key idea of positivism is that properties of the social world should be measured by objective measures that result in quantitative data. The opposing view of social constructionism is that it focuses on reality as determined by people rather than objective, external factors (Easterby-Smith et al, 2002). These two traditions can be seen as opposite ends of a spectrum, into which most research methodologies will fit somewhere. There is also a middle ground between these viewpoints, of the relativist point of view. Relativism requires consensus between different viewpoints. Seen as a spectrum, positivism focuses on concrete facts, while relativism takes into account that the facts depend on the viewpoint of the observer, finally social constructionism takes the viewpoint that there is no reality as all humans create their own reality.

Rationale

There are a number of academic principles and views on CSR that have been discussed in the literature review. Academic insights can often be different from real-life examples, this dissertation will therefore add to the existing knowledge by providing a snapshot in-depth analysis and study of CSR policy in ten international corporations as stated in their report and website.

Approaches

The Questionnaire or Survey

Robson (2002) classified surveys as those carried out by self-completion, face-to-face interviews or telephone interviews. All surveys generate statistical data and this eases the process of drawing comparisons and conclusions (Jackson and Bundgard, 2002). Advantages also include that questionnaires are less expensive to administer in terms of time and money when compared to other methods such as interviews (Oppenheim,1992, Pollit and Hungler, 1987).

Robson (2002) identified three central features of surveys:

A fixed, quantitative design.

The collection of a small amount of data in standardized form from a relatively large number of individuals.

The selection of representative samples of individuals from known populations.

According to Jankowicz (2000) a survey is carried out in order to establish people’s views, and this establishes another problem. When dealing with the results of surveys, ‘you are prepared to take people at their word’ (Jankowicz, 2000), CSR specifically deals with questions of corporate social responsibility, not personal responsibility, and so the ethical stance of the respondent may not reflect the ethical stance of the corporation. It is impossible to target the survey to the actual corporation, as although the corporation is made up of individuals, it has been given a number of the rights and statuses of humans (Dobbin, 1998).

It would be difficult to get the survey to a large amount of individuals that are responsible for the formulation of actual CSR policy. They would have to be administered to the person or department that was responsible for formulating CSR, or the appearance of the organisations CSR policy in the report and website. Due to the known low response rate of postal surveys (Owens, 1990) it would be difficult to identify the necessary number of individuals in order to gain a representative sample. This would provide problems in external validity, in that the findings could not be generalised.

This methodology is therefore inconsistent with the aim of evaluating the corporate ethical stance, as it would be difficult for the individual to answer the questions from the ethical stance of the corporation without bias from their own ethical stance.

Interviews

Interviewing involves researchers asking questions and receiving answers (Robson, 2002). Patton (1980, see Punch 1998) distinguished between three types of interviews; the informal conservational interview, the general interview and the open-ended interview. These types of interviews can take place individually or in a group. Transcripts are then used for analysis purposes.

The problems addressed in surveys also apply to the use of interviews for this research project. There are problems in identifying the individuals responsible, again the ethical stance of the individual may not be that of the corporation and the interviewee may introduce bias. It may be difficult for the respondent to clearly distinguish between the two. The words used by the interviewer may trigger certain associations, and there is no guarantee of obtaining ‘truthful’ statements (Alvesson and Deetz, 2001). Ethical issues in interviewing include whether the confidentiality and anonymity of respondents can be guaranteed, or where the respondents will want to tell the interviewer more information than is required (Mason, 1998). Regarding the issue of CSR, respondents may want to tell the interviewer inside information on why the organisation has followed a particular course of CSR; this information may affect the aim of evaluating the corporate ethical stance.

Observation

The purpose of observation is to witness what is going on in a particular setting or set of interactions (Mason, 1998). Silverman (1995) argued that for certain research questions, observation could provide both valid and reliable data. Observation typically provides anthropological data – anthropological fieldwork generally requires immersion in a culture over a long period of time (Silverman, 1995). Observation is typified by recording of language, actions and behaviour. This type of methodology does not lend itself to a ‘snap-shot’ evaluation, where policies – not people, are evaluated at a point in time. It is therefore not suitable for this research topic.

Case Study

Yin (1994) defined a case study as a strategy for doing research that involves an investigation of a particular phenomenon within its real-life context and using multiple sources of evidence. Its main application is to achieve an in-depth understanding and to get a detailed description (McGivern, 2003). This would initially fit with the general research question, as the particular phenomenon to be investigated is to examine to what extent international corporations have embraced the idea of CSR. Collis and Hussey (2003) describe the phenomenon to be investigated as the ‘unit of analysis’, and the multiple sources of evidence would be the report and the website.

Yin (1994) identified three characteristics of case study research:

The research aims to explore and understand certain phenomena within context.

The research does not begin with a set of questions and notions about the limits of the study.

The research uses multiple methods of collecting data, which can be qualitative and quantitative.

These characteristics fit both with the general research question and the more specific aims.

McGivern (2003) warned that if the findings from a case study are to be used to make generalisations about the wider group or population to which the case belongs, then case must be taken in ensuring that the particular case is representative of the wider population. However Yin (1994) described the comparative case study, where the same questions are asked in several related organisations. Jankowicz (2000) argued that in the comparative case study the purpose is not to sample different organisations because the findings are to be generalised, but to replicate – that is to compare the target organisation with others, exploring different stances to issues. This fits with the second aim of the research, which is to evaluate the corporations’ ethical stance.

Disadvantages of case study research include that it can be difficult to understand the events in a period of time without knowledge of what went on before and what may follow, plus progress of case study research can be very time consuming (Collis and Hussey, 2003).

Content Analysis

Krippendorff (1980) defines content analysis as a research technique for making replicable and valid inferences from data to their context. The data can be taken from written documents, films or recorded conversations. During content analysis the documents are analysed by establishing categories and then counting the number of instances when those categories are used in a particular item of text (Silverman, 1995). Fennell (2001) discussed that there are two main styles of content analysis, which can be used singly or combined.

Manifest content analysis includes the visible, surface content of a communication that can be easily counted.

Latent content analysis involves an analysis of the underlying meaning of a communication.

Advantages of content analysis include that it is an unobtrusive technique (Krippendorf, 1980) meaning the data does not change with the introduction of the researcher. Data generated in methodologies such as experiments and interviews can be skewed due to the subjects awareness of being observed or tested (Krippendorf, 1980). Silverman (1995) also discussed that this methodology pays particular attention to reliability, ensuring that different researchers can use the measures in the same way, and validity, through uses of word counts.

Disadvantages include Punch’s (1998) warnings that the researcher must also be aware of the underlying structure of the text, and that the social context of the text must also be studied. This means that categories must be clearly defined, together with a context and that coders should also read the surrounding text. For example, if ‘responsibility’ is defined as elements of the corporations CSR report, a coder reading the sentence ‘Office managers have full responsibility for gritting the car-park in cold weather’ should surmise that the word ‘responsibility’ is not in its correct context, and this word would not be counted.

In conclusion, this methodology fits with general research question and the aims; and can be used to investigate the content of the CSR communications. In any content analysis it must be clear which data is being analysed (Krippendorff, 1980). The data will be made up of the CSR portions of the report and website.

Methodology

Strategy

The chosen strategy is therefore that of the comparative case study, where the unit of analysis is an evaluation of the extent international corporations have embraced the idea of CSR. The characteristics of case study research fit with both the general research question and the more specific aims.

This research comes from a relativist point of view, as small numbers of cases have been chosen for specific reasons and the research will depend (to some extent) on the viewpoint of the observer (Easterby-Smith et al, 2002).

Data Collection

The data collection method is content analysis, which provides a way to collect quantitative data that can be used for comparative purposes, using content analysis together with the case study strategy means that the data can be analysed both in a quantitative and qualitative manner. Case study research can be made up of a variety of data collection methods. However Jankowicz (2000) argued that data in case studies are obtained largely through a review of written records. This will fit with the collection of data through the report and the website.

Both manifest content analysis (in the form of word counts) and latent content analysis will be used in the analysis stage.

Case Studies

Robson (2002) discussed that although a group of cases can be studied, these should not be thought of as a sample. Organisations are chosen for a comparative case study to be representative of different possible positions or stances, regardless of the frequency of these stances in the population. This is opposed to organisations in a study using survey methods, as these organisations are selected into a sample to be representative or some population (Jankowicz , 2000). The case studies are to be chosen by convenience, there is no attempt to use the case studies as a sample; rather they are used to be representative of different CSR stances.

A total of ten corporations will be evaluated. These have been selected purposively rather than randomly on the rationale that all the corporations have been the focus of ethical debate over the last three years. All of the corporations selected have at some time been either been accused of having an unethical approach to business, provide unethical products or manufacture in an unethical manner. These will be:

Marks & Spencer

British American Tobacco

Tetra Pak

BAE

Shell (originally Arthur Anderson)

Corus

Monsanto (originally Huntington Life Sciences)

Diageo

Nike

McDonalds

From the original list two corporations could not be used. Arthur Anderson, as the accountancy firm has now been sold off in numerous countries, although the umbrella corporation still exists it would have been impossible to obtain report or website details. Huntington Life Sciences does not have a website, as it has been targeted so often by animal activists they do not have a web presence. Shell and Monsanto have therefore replaced these.

Appendix 1 contains sources of why these corporations were chosen.

Structure of the Analysis

The analysis will be performed in three phases:

Case study

Manifest content analysis

Latent content analysis

Case Study Phase

Bennett (1986 see Jankowicz, 2000) listed four stages of case study work:

Determining the present situation

Gathering information about the background to the present situation

Gathering more specific data about the important factors in the present situation.

Presenting recommendations for action.

These stages describe how the case study strategy is to be carried out. The case studies will be initially put into context with a brief company profile. Then the report and the website will be analysed to introduce the main themes of CSR that evident in each organisation. This analysis will include a definition of terms, as different organisations have different terms for CSR. This stage will include a brief analysis of the corporations mission/vision or strategy statement to evaluate if CSR has a place at the very highest level.

Pollach (2003) described that text structure takes on a broader meaning in websites, the organisation of text within a website, plus the cohesion of individual pages which is realised through frames, menus and hyper-links in texts. This means that it is important to note the structure of the CSR website communications within the website, as well as the content of the text. During the case study analysis phase the organisation of the website will be described by evaluating the place of CSR within the sitemap.

Manifest content analysis

The report and the ethical pages of the website will then be analysed using content analysis. Manifest content analysis will be used to generate quantitative data:

Report

Word count of CSR information

Page count of CSR information

Percentage word count of CSR information / total information

Percentage page count of CSR information / total information

Limitations of manifest content analysis for the report phase include that the page count may not be valid, as the inclusion of graphics may skew the data. Due to the popularity of external CSR reports to the report this analysis may take place on a linked report.

Website

Appearance of CSR hyperlink from the homepage. This will give an indication of how important the corporation views CSR information.

Site map of website to visualise where CSR information is held

Limitations of manifest content analysis for the website phase include that it is difficult to generate exact word counts for websites due to the nature of hyperlinks, plus website pages can be any length; therefore this method is not valid.

Latent content analysis

The final phase of analysis will be latent content analysis involving an analysis of the underlying meanings of the report and website CSR communications. Due to the popularity of external CSR reports to the report this analysis may take place on a linked report.

Robson (2002) describes how to carry out latent content analysis:

Decide on a sampling strategy.

Define the recording unit.

Construct categories for analysis.

Test the coding on examples of text

Carry out the analysis

The sampling strategy will be the chosen case studies and the recording unit will be a word. The categories for analysis were initially taken from the EIRIS categories as seen in Chapter Two. It is freely admitted that the EIRIS categories are biased towards an ethical stance, but by evaluating the corporations on the categories provided by an ethically-biased, non-profit making organisation it is hoped to achieve a snap-shot of the organisation at its worse. These snap-shots can then be used in a comparative manner to determine the ethical stance as stated in the aims.

These categories will be evaluated on a number of points:

Word count

Is there an objective linked to this issue?

Has an external organisation or group validated the corporation’s claims?

As discussed, the EIRIS categories are used as the basic guide to the CSR measures utilised, however, since this study seeks to test claims by corporations that they have a wider CSR commitment, the specific CSR report is the key document reviewed. This may be included in the report or be a separate document. Since the report may contain some EIRIS CSR categories, such as ‘directors pay’, those companies that produce separate CSR reports and who do not include these items may be disadvantaged in this analysis. Therefore the ‘directors pay’ category has been removed from the list.

Reliability and Validity

Carmines and Zeller (1979 see Neuendorf 2002) define reliability as the extent to which a measuring procedure yields the same result on repeated trials. Triangulation will take place between the two data sources of the report and website, if both the data sources support each then this increase reliability, if the data sources do not support each other then this will be analysed in the results. In the content analysis phase, as a human coder is to be used, there will be a need to establish intercoder reliability, this will be tested during the pilot test. Krippendorff (1980) discussed that data can always be looked at from numerous perspectives; the pilot test and clear definitions of categories should prevent the introduction of errors due to different perspectives if other researchers should follow this methodology. However, the limitations of this study mean that one researcher will complete the coding after the pilot test; this will prevent errors from multiple perspectives but could mean that individual bias is introduced.

Although there is no recognised standard for intercoder reliability Neuendorf’s (2002) review indicated that ‘80% or greater would be acceptable in most situations’.

Validity is defined as the extent to which a measuring procedure represents the intended concept, or is it measuring what it is supposed to measure (Neuendorf, 2002). The study will be valid if it measures to what extent corporations use CSR as indicated in the report and the website.

Pilot Study

The categories were tested during a pilot study using the CSR report of Marks & Spencer. As a good research design should provide for the need to improve recording instructions (Krippendorf, 1980) the pilot may mean that extra categories may be added during this phase, or definitions and context made clearer. Three new categories were added at this point.

Animal welfare

Children

Stakeholder definitions

The additional researcher will test intercoder reliability, any training instructions will be recorded (Krippendorf, 1980) these are contained in Appendix 3. A score over 80% will be acceptable. The results of the pilot are also contained in Appendix 3.

While undertaking the pilot it was realised that one of the original categories needed a specific definition towards the aims of this study, the corporate governance category is therefore defined purely as how the CSR policy is administered and given a strategic direction. Generally this would be in terms of an ethics board or committee. This category has therefore been renamed to ‘Ethics board’.

When undertaking the word counts, it was realised that it was very difficult under the twin limitations of time and money to compute the total word count of the CSR report. Most of the CSR reports use repetition to emphasise key points. As there is no set formula for the CSR report (unlike the annual report), different companies will insert different information into the CSR report. It was therefore decided to use the total word count of the recorded categories as the total word count, rather than the actual total word count.

Other Issues

The websites are dynamic documents that are always changing. A copy of the website will be downloaded in order to carry out the methodology. The website pages may not have the same content as when they were initially evaluated.

Postmodernism

All of the methodologies that have been analysed with regard to CSR have been modernist in nature; this means that they are based on a certain set of assumptions about the nature and behaviour of corporations and CSR (Arias et al, 2001) An alternate view is that of postmodernism, this position is that the world is socially constructed and subjective: the observer is part of what is being observed and that science is driven by human interest (Thomas, 1997). This means that rather than focusing on facts by testing hypothesis in modern research, postmodernism focuses on trying to understand what is happening. Postmodernism offers one way to understand the complexities of life. In terms of CSR, a post modernistic approach would be to analyse CSR with no pre-conceptions, realising that any research results will be taken from the perceived reality of the researcher (Arias et al, 2001). Although the methodology of this dissertation focuses on modern research methods, it is worth mentioning that the postmodernist view that everyone creates their own reality and no one reality is less real that another (Arias et al, 2001) also applies to this study. The results and conclusion of this study, although borne out from empirical research methods, are written from the point of view and perceived reality of the researcher.

Ethics

The data used for this research comes from the report and the website. Both these documents are in the public realm, and so there is no need to obtain permission from the selected corporations. The data is the corporations’ viewpoint, and is published by the corporation.

Scope

Due to the twin constraints of time and money limitations, the primary data is limited to that published by the corporation in its report and website. There would be rich sources of other data, for example press releases or external reports, but these are out of the scope of this study.