28 
Chapter Three 
Methodology 
 
 
 
 
3.1 Introduction 
 
A qualitative research approach was selected for the study. As researcher Merriam 
(1998:5) argues, qualitative research is an umbrella concept covering several forms of 
inquiry that ?help us understand and explain the meaning of social phenomena with as 
little disruption to the natural setting as possible?. Other terms often used interchangeably 
are naturalistic inquiry, interpretative research and inductive research (Merriam, 1998; 
Erickson, 1986: Du Plooy, 2002). 
 
A number of key philosophical assumptions underpin qualitative studies. Firstly, it is 
assumed that ?reality is constructed by individuals interacting with their social worlds? 
(Merriam, 1998:6). Qualitative researchers are interested in understanding the meanings 
people have constructed i.e. how they make sense of their world and their experiences of 
it (Merriam, 1998:6). Secondly, it is assumed that the researcher is the primary instrument 
for data collection and analysis. Thirdly, it is assumed that qualitative studies involve 
fieldwork. Fourthly, it is assumed that an inductive research strategy will be employed. 
?Qualitative researchers build towards theory from observations and intuitive 
understandings gained in the field? (Merriam, 1998:7).  Finally, the product of a 
qualitative study is ?richly descriptive? (Merriam, 1998:8). ?Words and pictures rather 
than numbers are used to convey what the researcher has learnt about a phenomenon? 
(Merriam, 1998:8). 
 
There are also a number of assumptions that underpin most qualitative research. Merriam 
(1998) argues that the design is generally emergent and flexible, and responsive to the 
changing conditions of the study in progress. Further, sample selection is usually non-
 random, purposeful and small. (Merriam, 1998).  
 
A qualitative research framework was seen as appropriate for this policy study. Firstly, 
key to the research were the ?meanings? different stakeholders constructed. It was 
important to ?unpack? how negotiators interpreted decisions taken and their implications. 
29 
Secondly, given the nature of the study an ?inductive research? strategy was most 
appropriate. Theories were not tested - they were tentatively created. Abstractions, 
concepts, hypotheses and theories were being built from the ?bottom up?. Thirdly, in line 
with qualitative research methods, it made sense for a small, purposeful, non-random 
sample to be the focus. Certain key people were involved in the policy process - these 
were the ones that needed to be interviewed. Fourthly, given the nature of the study ?rich 
description? was the most appropriate research product. Words and pictures (rather than 
numbers, statistics etc.) best conveyed the researcher?s learnings. Finally, it was critical 
for the study design to be emergent, flexible and responsive to changing conditions. 
Questions and explanations built on one another as the tentative hypotheses and theories 
emerged. 
 
3.2 Designing the Study and Selecting a Sample 
 
As discussed in greater detail in Chapter Two, the theoretical framework selected was a 
critical political economy of the media approach. This perspective carefully assesses both 
market and state interventions in the media, and their impact. It is a perspective that looks 
at power and the impact of the asymmetries of power on media production. Further, it is 
an approach that not only analyses the way the media operates now, but also looks at how 
best to transform it. It is in terms of this framework that the study?s overarching question 
and sub-questions were derived. To reiterate, the overarching question was constructed as 
follows: 
 
Why and in what ways did the ANC government shift its position regarding 
the promotion of media development and diversity in South Africa, and what 
were the implications?  
 
In terms of the study?s sub-questions, there were two sets. One focused on the MDDA's 
vision and mandate, and the second focused on funding issues. (See Chapter One for a list 
of the questions.)  
 
As discussed a small, non-probability, purposive sample was selected for the study. 
Probability sampling (i.e. simple random sampling) was not selected. Probability 
sampling, allows the investigator to generalise the results of the study from the sample to 
the population from which it was drawn. However as Merriam (1998:61) argues, ?Since 
generalisation in a statistical sense is not a goal of qualitative research, probabilistic 
30 
sampling is not necessary or even justifiable.? As Merriam argues the most common form 
of non-probability sampling is purposive sampling: ?Purposeful sampling is based on the 
assumption that the investigator wants to discover, understand and gain insight and 
therefore must select a sample from which the most must be learned?. (Merriam, 1998:61; 
Also see Du Plooy, 2002) Purposeful sampling includes typical, unique, maximum 
variation, convenience, snowball, chain, and network sampling (Merriam, 1998:62). 
Maximum variation sampling was selected for this study. The latter involves identifying 
and seeking out those who represent the widest possible range of characteristics of 
interest for the study (Merriam, 1998:63). In line with this methodology the following 
criteria were selected. Interviewees needed to be involved in the hands-on, day-to-day 
MDDA policy negotiation process. They needed to operate at a decision-making level. 
Finally, they needed to represent each of the main stakeholder groupings involved ? that 
is, business, government, donors and the community media sector. 
 
Why these criteria? The reason for choosing people involved in the hands-on negotiation 
process was that they would best be able to explain some of the complexities of the 
process as it unfolded. Further, it was important that interviewees played a decision-
 making role. (Some people were involved in the day-to-day negotiation process but were 
not senior enough to take decisions.) Finally, it was important to look at the broad range 
of stakeholder perspectives. The different perspectives of different stakeholders and the 
power of these perspectives were central to the study?s analysis. So in line with these 
criteria the following people were selected for the study: 
 
? Government perspective: 
o Devan Pillay: Director Policy, Government Communication and Information 
Service (now Professor of Sociology at the University of the Witwatersrand). 
Prof Pillay was government's chief MDDA policy negotiator. 
o Joe Mjwara: Senior General Manager, Multimedia Unit, Department of 
Communications (DoC). Mr Mjwara was the key Department of 
Communications negotiator during the MDDA policy process. 
o Andrew Donaldson: Acting Deputy Director General: Budget Office (now 
Deputy Director General: Public Finance). Mr Donaldson was the key 
Department of Finance negotiator during the MDDA policy process. 
 
? Business perspective: 
31 
o Brian Pottinger: Publisher of the Sunday Times and Chair of the Print 
Development Unit Working Group in Print Media South Africa (PMSA)13 (now 
MD Johncom Africa). Mr Pottinger played a senior decision-making role.  
o Natasha Stretton (now Volans): Director, Print Development Unit (now General 
Manager of Print Media South Africa). Ms Stretton played a hands-on, day-to-
 day decision-making role. 
o Lara Kantor: Executive Director, National Association of Broadcasters (NAB) 
(now General Manager: Policy and Regulatory Affairs, SABC). Ms. Kantor 
played a senior decision-making role. 
 
? Community perspective: 
o Jane Duncan: Head of Policy, Freedom of Expression Institute (FXI) (now 
Executive Director, FXI). Ms Duncan, together with Mr. Mfundisi, co-ordinated 
the official community media input. 
o Mabalane Mfundisi: Head, National Community Radio Forum (NCRF) (now 
Manager, South African NGO Coalition [SANGOCO]).  
 
? Donor perspective: 
o Jean Fairburn: Director, Open Society Foundation South Africa (OSF-SA) (now 
studying in the UK). The Open Society Foundation was the largest community 
media donor in the country at the time of the MDDA policy process. 
 
? MDDA 
o Libby Lloyd Chief Executive Officer, MDDA. Ms Lloyd brought an interesting 
early ICASA perspective to her input, as she had co-ordinated MDDA policy 
inputs. Further, she had important insights regarding the final outcomes of the 
MDDA policy process. 
o Khanyi Mkonza: Chairperson, MDDA. Ms. Mkonza had important insights as 
regards the outcomes of the MDDA policy process. 
 
3.3 Collecting Qualitative Data 
 
3.3.1 Interviews 
 
                                                 
13 PMSA is an umbrella body representing large commercial print media companies in South Africa. 
32 
With qualitative research there are generally three types of data produced: interviews, 
documents and field observations (Merriam, 1998; Du Plooy, 2002). The focus of this 
study was on interview data and data gathered through document analysis. 
 
Interview data was a key component. As Weiss (1994:9-10) claims, there are number of 
important reasons to conduct qualitative interviews. These include: to develop detailed 
description, to integrate multiple perspectives, and to bridge inter-subjectivities. (See also 
Seidman, 1998) In terms of developing detailed descriptions, Weiss (1994:9) claims that 
interviews are the perfect tool with which to ?learn as much as we can about an event or 
development that we weren't there to see?. In terms of integrating multiple perspectives, 
qualitative interviews allow us to effectively ?describe an organisation, development or 
event that no single person could have observed in its totality? (Weiss, 1994:9). Finally, 
in terms of bridging inter-subjectivities, qualitative interviews make it possible for 
readers to ?grasp a situation from the inside, as a participant might? (Weiss, 1994:10). 
Developing detailed description, integrating multiple perspectives and bridging inter-
 subjectivities were all critical components of the research. 
 
The most common way of deciding which type of interview to select is to determine the 
amount of structure desired. (Merriam, 1998) ?At the one end of the continuum fall 
highly structured, questionnaire driven interviews; at the other end are unstructured, 
open-ended, conversational formats? (Merriam, 1998:74). Interviewing in qualitative 
investigations is generally more open-ended and less structured (Weiss, 1994; Merriam, 
1998). For this study a mid-point on the continuum was selected -? a semi-structured 
interview. The researcher asked all interviewees a set of similar questions. However, the 
questions were open-ended. Further, the sequence was often shifted to allow interviewees 
to focus on the issues they had particular insights into. Also, interviewees' insights were 
followed up on, and in certain circumstances follow-up interviews were conducted. On 
average each interviewee was interviewed for a three hour period.  
 
Core questions included the following: 
 
? Context: 
o What is the media's role in society? (Does it have any responsibilities to 
anyone?) 
o How would you define media development and diversity? 
33 
o Why is media development and diversity important? 
o What do you think aids (and restricts) media development and diversity? 
 
? The policy process:  
o From your perspective, how did the policy process leading up to the formation of 
the MDDA unfold?  
o What perspective did your particular constituency bring to the policy process? 
o How influential was your constituency in the policy process vis-?-vis other 
constituencies? 
o In what ways did your constituency seek to influence the policy process? 
o How would you assess the final outcome of the policy process? 
o From a policy perspective, what improvements/adjustments do you think could 
be made at this stage? 
 
3.3.2 Documentary evidence 
 
The second set of qualitative data utilised was derived by mining data from documents. A 
number of core government documents ? including position papers, legislation and 
regulations ? were analysed. Core documents included the following: 
 
? Task Group on Government Communications, 1996 
? Media Development Agency: Draft Discussion Document, 25 November 1999. 
? Media Development Agency: Discussion Document, 8 March 2000. 
? Media Development and Diversity Agency Draft Position Paper, November 2000. 
? Media Development and Diversity Agency Final Position Paper, November 2001. 
? Media Development and Diversity Agency Bill (B2A - 2001) and (B2B - 2002). 
? Media Development and Diversity Agency Act No.14 of 2002. 
? Proposed Regulations in Terms of Section 22 of the MDDA Act (Act 14 of 2002) 
July 2003. 
? Regulations in Terms of Section 22 of the MDDA Act (Act 14 of 2002) October 
2003. 
 
These documents were supplement with workshop documents, position papers, 
newspaper articles, publications and so forth. Documents were produced by the MDDA 
itself (once it had been established). They were produced by business and community 
media groupings. However, unfortunately no documentation was available from the 
34 
donor side. The researcher was directed to relevant documentation by interviewees.  
 
3.4 Analysing and Reporting the Data 
 
A number of qualitative data analysis strategies exist. These include ethnographic 
analysis, narrative analysis, phenomenological analysis, and the constant comparative 
method. (Merriam, 1998:156). After careful consideration the constant comparative 
method was selected. The approach entails the following: 
 
The researcher begins with a particular incident from an interview, field 
notes, or document and compares it with another incident in the same set of 
data or in another set. These comparisons lead to tentative categories that are 
then compared to each other and to other instances. Comparisons are 
constantly made within and between levels of conceptualisation until a 
theory can be formulated (Merriam, 1998:159).  
 
Generally there are three levels of analysis. (Merriam 1998:178-87) The most basic 
presentation of a study's findings is a descriptive account. One level deeper is to construct 
categories of themes that capture recurring patterns. These categories are abstractions 
derived from the data, not the data themselves. The third level of analysis involves 
?making inferences, developing models, or generating theory? (Merriam, 1998:187). This 
study attempted to move beyond description into conceptual analysis and further to 
tentative theory generation.