Contents - Previous - Next

Section 7: Analysis of the results

7.1 Introduction

This section looks at what you do with all the information you collect. The analysis of focus group information can be done at a whole range of levels depending on what you want to do with the information. This Manual will present only the simplest and quickest methods as a complex analysis of information is not usually necessary for health professionals who use this technique simply to solve programme problems or design better programmes. You will need the type of analysis that can easily be performed by you and the project team, while still providing the best way to manage your results as quickly as possible. Those interested in more comprehensive analysis procedures for larger more formal research projects, should refer to other texts (see e.g., Bernard, 1988).

7.2 When do we analyse?

In this type of study, analysis of your information is an on-going process that begins as soon as you enter the field or begin the project, and continues until you write the final report. It is very important to have this idea firmly in your mind before you begin the study. If you leave the analysis to the very end, you could discover large gaps in your results and at that stage it would be too late to correct any problems you have discovered. Early and continuous analysis serves three main purposes:

• to enable the study to focus quickly on the main issues that are important to the participants, and then explore these issues more closely;

• to check that the focus group discussions are being conducted in the best possible way (i.e., natural flowing discussion, participants not forced into answering in a particular way, and so on);

• to examine the results of the discussion early enough to be able to check that the information you require to meet the project object is actually being collected.

7.3 Who does the analysis?

This will depend on the project, but we recommend that all those involved in the collection of information should be included in the process The amount of involvement will of course depend on the time each team member has been given to work on the project. At worst, one team member can be responsible for analysis, but it must be done in consultation with field staff who were present at the focus groups. It is not reasonable to expect to be able to analyse by yourself information that you have had no par collecting.

7.4 How much analysis is necessary?

This must be decided in relation to how much time you have for project, the skills of the team, and how important the decisions are that be based on the information. It also depends on whether or not the stud for research that is intended to be used by persons beyond your programme or department. Just use your common sense.

If your project calls for the absolute minimum analysis, then that simply be a group meeting of those who conducted the focus group discuss what they learnt. It may involve using the notes produced by focus group observer to stimulate a discussion. Small programme decisions can then be made during the same meeting. This will be adequate especially if those involved in the decisions were present at the focus groups. This approach is obviously not enough where larger decisions are being made where results may be used to inform decisions in other locations, or formal research.

The method described below is adequate for small to moderate projects. It can be used for formal research, and is especially useful projects that need to be completed quickly. It does not require any previous research experience and can be performed by people without a university degree.

7.5 What activities are included in the analysis process?

Again, with this type of research, you do not sit down at the end of collection of information and enter numbers into a computer that are then used to perform statistical calculations. There are several activities that part of the analysis process. Some of them you would do without think but it is useful to look at all of these activities and develop some of then that they are more formal and can be included in the final result to enable more comprehensive, professional and useful recommendations. It will also help you think more logically and clearly about what you are doing.

Orientation

At the beginning of the project, you will spend time talking to people about the topic in casual conversation. These might include local political leaders, religious leaders, local medical staff, traditional healers, and so on. Although this information should always be written down, it can be just jotted down in note form and may not need to be expanded upon too much. This information helps you to build up a picture of the topic under study. All staff can discuss their findings in the team meetings that are held in the planning phase to design the project. Make special time for this activity in those meetings.

Debrief

At the end of every focus group, time must be set aside to examine the focus group activities and results. Debriefs are discussed more thoroughly in Part II, Section 14, but for now it is important to realise that this lengthy discussion about each focus group is an important part of the analysis process. It feeds back information quickly that you can act on immediately, and builds upon the developing picture.

Analysis of transcripts

This is probably the most difficult of the analysis activities and is certainly the most formal. As we have already mentioned, the analysis of the transcripts or session notes can be as simple or as complicated as you decide. This process should be carried out as the transcripts become available, not when all the focus groups have been conducted. A detailed description of one possible technique for analysing the information from the transcripts to the final analysis stage follows below.

Analysis of all focus group discussions

Once all the focus groups have been conducted, and you have the results from each session, then it is necessary to look at all the focus group discussions together and begin to describe findings that apply to the study as a whole. Do not forget that if you are using other methods for your study (like individual interviews and observation), as most often you will, it is useful to have a final report on the findings from the focus groups separately. At the end of this activity you should have produced a set of results with a detailed description of what you believe the results tell you in relation to your objectives.

Example analysis: From transcript to final results

STEP 1: ANALYSIS OF INDIVIDUAL TRANSCRIPTS

When the transcripts have been completed you need to read them in several different ways. First of all you need to read them as a whole and to note your general impressions. This should be done just naturally without too much structure. Have your objectives fresh in your mind while you read through the transcripts. Look for major opinions and attitudes that are expressed by the groups.

Next, you need to read the transcripts looking for very specific things. Take out your list of objectives, and the document that lists all the information you require. The list should include all the areas that you are interested in finding out about. This is really only a list of what you think is important information to meet your objectives, but as a result of the focus group discussions, you may discover new areas that are also quite important.

Next, you will need to read through each transcript and remove any responses that were forced from participants by poor moderating skills. This often can occur and does not necessarily mean the moderator is not performing well. Many transcripts need to have responses removed before analysis. Part II of this Manual describes in more detail what may cause these responses, and gives you a better idea about how to identify them in a transcript.

At this stage you are also looking for sections that were poorly transcribed and do not make much sense. Remove these from the analysis process. In addition, there will be some statements that seem to be made simply because others have made them. You cannot really be sure how strongly a participant really believes something they have said, but if you strongly suspect a statement not to be truly accurate (for whatever reason), it does no harm to mark the statement as having less importance in the analysis than other responses. So some responses will be taken with more caution than others.

We recommend that you then code the transcripts. This means marking sections of the transcript in a way that indicates what the participants are talking about. For instance, in our example of the programme of multi-drug therapy for leprosy, every time a participant mentions access to the clinic, you mark the section to indicate this. Use code words to make it faster, e.g., CLINACC (i.e., CLINIC ACCESS). So, in the end your transcript will have a list of code words running down the side of the page. This makes it easier to identify sections of interest later on, as all you will then need to do when looking at the issue of access is run your eyes down the transcripts and take all the responses marked CLINACC. If you are running focus groups on folk diagnosis of malaria, and are interested in the ability of participants to distinguish between fevers, then you might code such discussions as DIFFEVER (i.e., DIFFERENT FEVERS) (see Box 4). For those who have access to computer software packages, the programme The Ethnograph (Siedel, Kjolseth and Seymour 1988) can be used to manage and sort your coded transcript. Note that some responses may refer to more than one issue or idea. These can be given more than one code word so that the response can be included in each relevant area.

You can use the list of information you require to prepare a lot of code words. Try not to break down the information you require into categories or codes that are too small. For example, if you are interested in access to the clinic, you could either code the responses that refer to access to clinic generally (CLINACC), or you can break them down further into categories. These could be: those who find access difficult due to the cost (ACCESS COST), those who have no transport (ACCESS-TRANSPORT), those who have no difficulty at all (ACCESS OK) and so on. Avoid this! You will end up with so many code words and categories of information that you will still be analysing transcripts after the programme plan decisions have already been made! Keep it simple!

Box 4: Coded transcript from focus group on symptoms of malaria

Group discussion, cont.

Not all responses fall into neat categories of the information that you have expected to obtain. In fact, many do not. As you find a response that brings up a new idea or topic that you had not expected, simply code it under a new name and note down that a new idea has been introduced. It is essential to keep a code book. This is simply a list of all the code words you have used and an explanation of what they mean (see Box 5). Sometimes, on larger projects you may have well over 100 codes and it becomes very difficult to remember all the codes and what they mean.

Box 5. Code boon for focus groups on symptoms of malaria

The final step in reading the transcripts involves using your list of required information and checking what information you have actually obtained. This will show you very quickly if the focus group discussion is getting to the point or not, and if you are obtaining the information you set out to gather. If you code information as you go along, you'll be able to alter your question guide quickly to be more effective in the next focus group.

STEP 2: LOG BOOK

The log book is just a place to keep all your responses together according to the topic of interest. In the example of access to the clinic, let's say you have decided to have CLINACC as your code to indicate all responses discussing access to the clinic.

In the log book under CLINACC you will enter every response that is coded as such in your transcripts. You should enter each and every response unless it is exactly the same as another. The idea is to retain the full range of responses. If you are using a large notebook for the log, write all your responses down the left side of the book. The right side should be divided into columns, one for each focus group session. As you find responses that are exactly the same, you can begin to tally the amount of times they occur in the columns.

It is important to keep the tally of responses separate for each focus group. Sometimes, one person will make a statement, and everyone else will agree or say the same type of thing. If eight people say something in only one of ten focus group sessions which you held, this may be less significant than if eight individual people say the same thing in eight different focus groups out of ten. You will want to know how many times an issue was discussed across all the focus groups as well as how many times in total a response was given. SO, keep the results separate for each focus group (see Box 6).

STEP 3: WRITING THE RESULTS

It is important to keep in mind that you will be writing your results not only from the log book, but also from the notes that you made while reading the transcripts as a whole. This is important because when you take the responses out of the transcript, it is possible to misunderstand the circumstances in which a particular response was made.

Writing up the results is really just a matter of deciding which responses are important enough to include. Suppose, for example, that access to the clinic was only mentioned by two participants. You will need to decide whether to include this in your results or not. If your office was concerned about access as a problem, then it would be wise to include this finding: the focus group discussions indicate that access is not a major factor to explain poor clinic attendance and treatment compliance. You will find that you have a surprising amount of information, and you will not want to include it all. Go through the log book and the notes, and decide what to keep and what to leave out, which findings to summarise, and which quotes to include verbatim in the text.

Box 6. Example of log hook from Bohol study (Dawson et al. 1991)

Many reported results from focus groups do not indicate how many focus groups or participants discussed a certain issue. You will read results that say "many respondents said..." or "only a few focus group discussions raised the issue of...." It depends on what you plan to do with the results, but it can be useful for others reading your report to know the frequency of particular issues raised. This allows them to decide how important something is, and people who are familiar especially with quantitative research, such as social surveys and epidemiological studies, rather than qualitative research, find it valuable to assess the significance of findings in these terms. If you choose to indicate numbers of focus groups or number of responses, then it is simply a matter of adding up the tallies in the log book. In addition, your report should include a summary of the method, and a profile of participants derived from the initial registration (for example, the ages, marital status, and occupations of participants) In small studies, this information can easily be hand tallied and then described.

STEP 4: INTERPRETATION

Throughout the study, you should have been thinking about the significance of the information you were collecting, in terms of the problem or question you want to answer. The team should have some well developed ideas about what the respondents are saying. Now is the time to look at the results and discuss them with the rest of the team at length. Do the results really show what our common sense tells us about the community now? What do these results really mean? This careful look at the results will lead you into writing a report and making recommendations.

Contents - Previous - Next