Cluster III. Good governance: where the strategy aims at developing structures and systems of democratic governance that are participatory, representative, accountable and inclusive and
4.7.2 Focus group discussions
Focus group discussions are a means of obtaining information from people in a group. It is an interview with several people on a specific topic or issue (Bryman 2004: 345). Instead of asking questions of each person in turn, focus group researchers encourage participants to talk to one another, asking questions, exchanging anecdotes and commenting on each other‟s experiences and points of view. Focus groups are group discussions exploring a specific set of issues. They usually involve a narrowly focused topic discussed by group members of equal status (Payne and Payne 2004: 103). The group is „focused‟, in that it involves some kind of collective activity, such as debating a set of questions on a specific topic (Barbour and Kitzinger 1999: 4; Robson 2002: 285). Having gained significant popularity among market researchers, focus group discussions are becoming widely used by information science professionals to gather data on group norms and shared experiences (Von Seggern and Young 2003: 272).
Barbour and Kitzinger (1999: 5) explained that focus group discussions are better for exploring how points of view are constructed and expressed, are well suited to the study of attitudes and experiences around specific topics and they are invaluable for examining how knowledge, ideas and story-telling operate within a given cultural context. The aim of focus group discussion is mainly to gain rich and often exploratory information (Tacchi, Slater and Hearn 2003). Focus group discussions are important because the group develops its own conversation, raising issues and ideas that might not emerge in a discussion with the interviewer alone (Cohen, Manion and Morrison 2000: 288). In this study focus group discussions were held with members of the communities surrounding the telecentres (users and non-users of the telecentre) to gain richer understanding of actual practices of use and interaction with technologies in the wider context of people‟s lives and socio-cultural structures.
Focus group discussions are normally used for triangulation purposes, or in conjunction with other data-gathering techniques (Barbour and Kitzinger 1999: 6; Cohen, Manion and Morrison 2007: 376; Denzin and Lincoln 2005: 704). Bloor et al. (2001: 8, 12) pointed out that focus groups have a much large part to play as an ancillary method, alongside and
complementing other methods. The focus group data may as well be compared with other data on the same topic gathered by other methods. For these reasons this study used focus group discussions, in conjunction with other data-collection methods. These discussions were used to gather supplementary data, as well as verifying some points that emanated from some of the responses to the individual interviews.
In conducting focus group discussions, the present researcher had to consider a number of factors. These include deciding on the size and composition of the groups, deciding on the number of focus groups to be conducted and deciding on the criteria to be used to select participants to be included in the discussions (Cohen, Manion and Morrison 2000: 289;
Morgan 1988: 41-8; Morgan and Scannell 1998: 6). In most cases, all these decisions depend on the purpose of the research (Fern 2001: 11; Morgan and Scannell 1998: 6). The following section presents overviews of the decisions that the researcher had to make concerning the focus groups discussions which were used in this study.
4.7.2.1 Group sizes
Different studies have used different group sizes in focus group discussions. Therefore studies are not agreed on the size of the focus groups to be used (Mosia and Ngulube 2005:
178). Barbour and Kitzinger (1999: 8) cautioned that advice about groups‟ size and composition in existing guides to focus group research is often didactic and this can hamper effective application of focus group methods. The ideal number recommended from the literature of between eight and 12 people is coming mainly from market research and might be too large for many sociological studies. Table 4.2 summarises selected optimal groups sizes found in the literature, in order to demonstrate the varying sizes of groups researchers are likely to find in the literature.
Table 4.2 Optimal focus group size found in the literature
Author(s) Optimal focus group size
Morgan 1988: 43 McClellard (1994: 29)
Morgan and Scannell (1998: 71) Barbour and Kitzinger (1999: 8) Greenbaum (2000: 3)
Bless and Higson-Smith (2000: 110) Bloor et al. (2001: 26)
Sekaran (2003: 220)
Von Seggern and Young (2003: 274)
4-12 8-12 6-10 3-5 7-10 4-8 6-8 8-10 4-12
Adopted and modified from Mosia and Ngulube (2005: 178)
When deciding the size of groups for focus group discussions, it is recommended that the group should not be so large as to be unwieldy, or to prevent adequate participation by most members, nor should it be so small that it fails to provide substantial greater coverage than that of an interview with one individual (Bloor et al., 2001: 27; Merton, Fiske and Kendall 1990:137; Morgan 1988: 41). Morgan and Scannell (1998: 71) indicated that deciding on the right numbers of participants for a focus group means striking a balance between having enough people to generate a discussion and having so many people that some feel crowded out. Cohen, Manion and Morrison (2007: 377) stated that with a small group the intra-group dynamics will exert a disproportionate effect, while with too large a group the group becomes unwieldy and hard to manage. The decision on the number of respondents to be included in the group discussion should take into consideration the number of questions (typically 8-12) that the discussion guide have and the duration (typically 1 to 2 hours) the discussion would last. In this case, the discussions lasted for two hours (Morgan and Scannell 1998: 71).
4.7.2.2 Group composition
Closely related to the issue of the size of the focus groups is the composition of these groups.
Interaction between participants is a key feature of the focus group method and therefore careful consideration of group composition is vital (Fern 2001: 16). In determining the
composition of the groups, the present study paid attention to the issues of compatibility of the participants. Cohen, Manion and Morrison (2007: 377) said extreme care should be taken with composition of focus groups, such that every participant is the bearer of the particular characteristic required, or that the group has homogeneity of background in the particular areas, otherwise the discussion will lose focus or become unrepresentative. Morgan and Scannell (1998: 59) advised that when the participants perceive each other as fundamentally similar, they can spend less time explaining themselves to each other and more time discussing the issues at hand. Morgan and Scannell (1998: 59) explained that generating a productive discussion requires good group dynamics and that depends on compatibility of the participants. To achieve compatibility the present study paid attention to the issue of background of the respondents, such as the participants‟ experiences with the use of ICTs.
Different groups were conducted for users of telecentres and those who were relatively experienced with the use of ICTs. Different groups were conducted for non-users of telecentres and those who were relatively less experienced with ICTs. This was important, because different experiences with the use of these technologies produce different perspectives (Morgan and Scannell 1998: 65).
Taking into consideration all the above factors, this study used group sizes of between six and 12 people, depending on the availability of the respondents. The decision to use relatively large groups for this study was influenced by the fact that the level of engagement of members of these communities with these new technologies was generally low. Morgan and Scannell (1998: 69) recommended that small focus groups be used when talking with experts or people who are well experienced with regard to the issues that are being discussed and when each participant has much to say on the topic. They advised that large groups should be used when talking to people who are less experienced with the issue under discussion.
Morgan and Scannell (1998: 75) stressed that non-users and people who have never used a service or a product have little to say about it, yet these non-users are often crucial target audiences.
4.7.2.3 Number of groups
Commenting on the number of focus group discussions to be conducted in a study, Morgan (1988: 41-8) warned that one group is never enough. Morgan and Scannell (1998: 82) agreed
that using one group is often risky. Cohen, Manion and Morrison (2007: 377) stated categorically that one group is insufficient, as the researcher will be unable to know whether the outcome is unique to the behaviour of the group. Bryman (2004: 349) said it is unlikely that one group would satisfy the needs of the researcher, since there is always the possibility that the responses are particular to that one group. The number of focus groups in a study may vary from three or four to over fifty (Barbour and Kitzinger 1999: 7; Bloor et al., 2001: 28;
Morgan and Scannell 1998: 77). However, Morgan and Scannell (1998: 77) said that there is no hard and fast rule about how many groups are sufficient. Dealing with too few groups may result in one missing something, or lead to premature conclusions, but using too many is a waste of time and money. Bloor et al. (2001: 28) pointed out that focus groups are labour intensive in recruitment, transcription and analysis. Therefore, where possible, numbers should be kept down to a bare minimum. The appropriate number of focus groups depends on the research questions, the range of people the researcher wishes to include and time and resource limitations.
Morgan and Scannell (1998: 77) stated that the biggest issue in determining the number of groups is the underlying diversity of what people have to say. If practically everyone has the same thoughts on a topic, this will be evident after a few groups and the “theoretical saturation” will be reached rather soon, whereas when the responses are more diverse, it will take considerably more groups to hear what people have to say. This study conducted two focus group discussions in each of the communities surrounding each telecentre. The whole study had eight focus group discussions in total. (See section 4.10.2 for details)
4.7.2.4 Selection of group members
Statistical representation is not the aim of most focus groups research. Usually, focus groups researchers employ „qualitative sampling‟, in order to encompass diversity and compose a structural rather than a random sample, guided by the particular research questions which they are addressing (Morgan and Scannell 1998: 56) As with most other qualitative methods, the focus group method relies on purposive samples. In purposive sampling respondents or other units are chosen for a particular purpose (Leedy and Ormrod 2005: 206; Robson 2002: 265).
A purposive sampling strategy chooses the focus group participants according to the project‟s
goals. Using purposive sampling the participant should be selected using well-defined purposive selection criteria, as opposed to convenience samples, which emphasised the ease of recruiting the participants (Morgan and Scannell 1998: 56).
In this study, selection of respondents to be involved in the focus group discussions was done in such a way that there was a representation of the strata used in identifying participants for the structured individual interviews. This means the respondents involved in the focus group discussions were people who were involved in various economic activities such as farming, livestock-keeping, small-scale businesses, fishing and small-scale mining activities, based on what the dominant economic activities in each community were. Groups were structured in such a way that users of telecentres and those who were relatively experienced with the use of ICTs were placed in one group. Non-users of telecentres and those who were relatively less experienced with ICTs were in another group. Selection was influenced by the type of information obtained and lessons learned during individual interviews.
The major attraction of focus group discussions over other data collection techniques, as pointed out by Cohen, Manion and Morrison (2000: 287), Cohen, Manion and Morrison (2007: 378), Denzin and Lincoln (2005: 704-5), Leedy and Ormrod (2005: 146), Ratcliffe (2002: 21), Robson (2002: 284-5), were the following:
The potential for discussions to develop, thus yielding a wide range of responses;
Can generate a wider range of responses than individual interviews;
Might be useful for gaining insight into what might be pursued in subsequent individual interviews;
They are often quicker than individual interviews and hence are time-saving and involve minimal disruption;
They are inexpensive to conduct and often produce rich data that are cumulative and elaborative;
Can be stimulating for respondents and aid recall; and
Provide natural quality control on data collection because participants tend to provide checks and balances on each other and extreme views tend to be weeded out.
Focus group discussion has a number of disadvantages, which the researcher was aware of, and tried to reduce their negative effects in this study as much as possible.
Results cannot easily be generalised;
The emerging group culture may interfere with individual expression;
The group may be influenced by one person and group-thinking is a possible outcome;
The requirement for interviewer (moderator) skills are said to be greater than that for individual interviewing, because of the group dynamics that are present;
It is difficult to research sensitive issues with focus group discussion; and
Confidentiality can be a problem between participants when interacting in a group situation.