Objective
To explain how baseline studies in PRCA are designed and implemented.
At the end of this chapter you will be able to:
1. Recall why baseline studies are important for communication programme planning.
2. Use PRCA findings to focus baseline studies.
3. Design a baseline research proposal.
4. Carry out a baseline study in the field.
5. Present baseline study findings.
A baseline study is a descriptive cross-sectional survey that mostly provides quantitative information on the current status of a particular situation - on whatever study topic - in a given population. It aims at quantifying the distribution of certain variables in a study population at one point in time. It involves the systematic collection and presentation of data to give a clear picture of a particular situation as it relates the following: What? Who? Where? When? Why? How? A baseline normally covers only a sample of the population. If a baseline study covers the entire population it is called a census.
In PRCA, a baseline study generates information on the levels of awareness, knowledge, attitude and practices (AKAP) of a given population on selected topics in a specific geographic area. The baseline study is repeated at the completion of the communication programme implementation to measure changes that have occurred over time in the characteristics that were studied before the beginning of the programme.
There are many good reasons to conduct a baseline study in addition to a PRCA. While PRCA collects and presents qualitative information about the community, the baseline expresses its findings in a quantitative manner. Below are some of the reasons for conducting baseline studies:
Baseline studies quantify the results of PRCA and serve as means of triangulating or verifying the appraisal results in a statistical manner.
Results of baseline studies, because of their statistical nature, can often convince and provide justification to policy-makers for the necessity of mounting a communication programme for a specific problem or project.
Results of a baseline study, if shared with the community, might galvanise the people to action. If the findings are positive, the community can be shown that they are already doing a good job and should keep it up. When the results are negative, this could also serve as a catalyst for discussions with the community on the most appropriate means of action.
Baseline studies are used to shape the communication strategy by assisting to further segment the priority interaction groups, sharpening communication objectives, and focussing content of media materials.
Baseline study results show the spread and use of various media in the community. They also provide information on the media preferences of the different groups in the community, and hence guide the selection of most appropriate traditional and modern media to be used in a communication programme.
Baseline studies serve as a reference point or benchmark for later comparison or impact studies to assess how well the original communication objectives have been achieved.
If conducted properly, results of baseline surveys can be generalised and used for communities with similar characteristics.
PRCA results are the bases for the design of the baseline study. PRCA results give qualitative information on what people of a given community know and what they perceive as a problem. They give an indication of what needs attention in the particular community and provide guidelines on what a baseline study should focus on.
PRCA is therefore a guide to the most important issues for investigation with a baseline study.
IPRCA can provide a guide, and in some cases, the sampling frame for the selection of households and persons to be interviewed to obtain the most relevant information in a baseline study. This is especially important where no sampling frame can be obtained from the local authorities. In such cases the social maps drawn by the community become the sampling frame.
A preliminary baseline proposal and questionnaire should be prepared with the secondary data available. These will be focused and refined with the findings of the PRCA before they are pre-tested and administered.
Formulate a preliminary baseline study purpose. In general, the purpose of most baseline studies is to determine in a quantitative manner the current situation in the community and assess the people's perceptions, levels of awareness, knowledge, attitudes and practices (AKAP) related to the development issues under qualitative investigation by the PRCA. The purpose of the study assists in the formulation of the study topic or title.
A clear problem statement helps in the further development of the research proposal and the selection of objectives and the methodology for the study. It makes it easier to identify secondary information and reports from which the study under preparation can benefit.
It is important to clearly state why the proposed research should be undertaken. It is often very essential to show the uniqueness of the new study. To achieve this, ensure that the proposed study is not a duplication of other research studies already carried out on the issue and point out the innovative uses of the anticipated results. A good problem statement should assist in the formulation of a convincing justification for a baseline study.
Careful literature review reduces the chances of the proposed study duplicating work that has already been done on the issue under investigation. It helps to discover what others have learned and reported about the issues that the new study proposes to investigate. It familiarises the researchers with other types of methodologies that have been used to study the issues and may provide convincing justification for why the proposed study is needed.
Properly formulated study objectives help to focus the baseline study and reduce the chances of collecting information that is not strictly necessary for understanding and solving the problem under investigation. Objectives also help in the organisation of the study in clearly defined parts or phases. Specific objectives facilitate the development of the research methodology and give a framework for the collection, analysis, interpretation and utilisation of study results.
In PRCA, there are normally two sets of objectives for the baseline study: one set specifies what the study will focus on in terms of the subject-matter under investigation, while the other set deals with communication elements to be measured.
In formulating the study objectives, bear the following in mind:
- Objectives must cover the different aspects of the problem and its contributing factors in a coherent and logical sequence.
- The objectives must be clearly phrased in operational terms, specifying exactly what needs to be done, where and for what purpose.
- Objectives must be realistic considering local conditions.
- Action verbs that are specific and measurable must be used in formulating the objectives. Examples of action verbs include: to determine, to compare, to verify, to calculate, to describe and to establish. Avoid such verbs as to appreciate, to understand or to study.
Set study objectives using the SMART principle
Specific
Measurable
Achievable
Realistic
Time-framed
Sampling methods and sampling size
Sampling
Sampling is essential because populations tend to be large and resources and time available limited with the result that it is usually not possible to study each person. For this reason there is little choice but to select a sample from the population and from it make projections or generalisations regarding the entire population.
Figure 14: A social map showing dwellings
Sample = a selected subset of a population.
Choose the geographic areas and population from which people who will respond to the questionnaire will be selected. It would be ideal to interview everybody in the chosen geographic areas but this is often too costly and time consuming. This is why it is important to select samples from the communities in the affected areas for the baseline study.
Representativeness
To draw conclusions that are valid for the whole population, care must be taken to draw a representative sample from the population. Such a sample must contain all the important characteristics of the population from which it is drawn.
Sampling methods
Simple random sampling: in simple random sampling the first step is to draw up a list of all the units that is, sampling frame. Each unit has an equal chance of being selected out of the entire population.
One way of obtaining sampling frames is to use the social map drawn by the villagers themselves that shows individual dwellings.
When the sampling frame is ready, select the respondents at random. Random selection means that the team has no discretion or choice about which units (areas, groups, homesteads or individuals) actually go into the sample. In random selection each unit has an equal chance of selection that is independent of any other event in the selection process. In other words, this procedure serves as a check on conscious or unconscious biases on the part of the researcher. It prevents the researcher from selecting cases that would support her PRCA findings or her assertions about the issue under study. It is like a lottery. For instance, the researcher can write the names of the people in a community on small pieces of paper, put all the pieces of paper in a hat or box and draw the required number of names from the container.
Systematic random sampling: This is another method of selecting a sample from the population through the use of a sampling interval. It means that all the units are numbered and only units that occur after a certain interval is included in the study. For instance, every fifth hut in the village is selected as a unit to be interviewed.
With simple random sampling and systematic random sampling, it can be assumed that if the people whose names are picked up are interviewed, the study will get a fair representation of the opinion of the people on the whole list. However, this is only possible when all the people in a community are homogeneous: of the same age, gender and so on. To ensure that such differences are taken into account when sampling, other processes must be used.
Stratified random sampling: This involves dividing the population into distinct subgroups according to some important characteristics, e.g. age, gender, race, socio-economic status, geographical distribution, etc. and then selecting a random sample out of each subgroup or strata.
If the proportion of the sample drawn from each strata is the same as the proportion of that strata in the total population then the strata will be fairly represented. For example, if in total population 55 per cent of the households are male-headed and 45 per cent are female-headed. Sample of households stratified according to head of household should have 55 per cent male-headed households and 45 per cent female-headed households.
Cluster sampling: This involves the use of sampling frames that identify groups or clusters of enumeration units. A cluster is a sampling unit with a number of enumeration units. Typically the population is divided into mutually exclusive and exhaustive clusters based on geographical, administrative or political criteria.
Cluster sampling can be performed by following these steps:
Step 1. |
List the clusters |
Step 2. |
Select the clusters |
Step 3. |
List enumeration units in the selected clusters |
Step 4. |
Select a sample of the enumeration units |
For example:
Step 1. |
List villages |
Step 2. |
Select villages |
Step 3. |
List mothers of children between 12-59 months age group |
Step 4. |
Select sample out of the mothers. |
These steps are called stages in sampling terminology and sampling plans are often categorised in terms of the number of stages involved. For example, in a single stage cluster sampling only one stage is involved and in a multi-stage cluster sampling, two or more stages are usually analysed.
Sample size
An essential part of planning any investigation is to decide how many people need to be studied in order to acheive the study objectives. It is a strong belief among researchers that the bigger the sample the more representative and hence the more accurate the results. Too often the number is decided purely on logistic grounds determined by the resources available for the study such as time, manpower, funds, transportation. On the other hand interviewing more persons than is strictly necessary is a waste of time, money and other (usually limited) resources. In general, the table below is a guide on how big a sample to take in relation to total number of people in a population.
Table 5: How big a sample should be adequate?
Total number of group |
Suggested number of sample |
Percentage |
100 |
15 |
15 |
200 |
20 |
10 |
500 |
50 |
10 |
1000 |
50 |
5 |
Data collection tools and procedures
Data collection tools and procedures allow for the systematic collection of information pertaining to the baseline study objective. They include observation and questionnaires administered through face-to-face interviews.
For baseline studies in PRCA, the administered questionnaire and observation are the recommended tools and techniques. Observation involves the collection of information related to the issues under investigation by watching, listening, inspecting and recording of information or activities noticed. An initial guideline on what to observe helps to make this activity systematic.
Use structured or close-ended interview questions and semi-structured or open-ended interview questions for the questionnaire. While structured questions are rigidly phrased with pre-determined answer categories, the semi-structured questions leave room for probing.
Pre-code the questionnaire for easy data collation and analysis.
Table 6: Example of a pre-coded questionnaire.
1. Sex |
1. Female |
|
|
2. Male |
|
2. Age group |
1. 15 - 20 years |
|
|
2. 21 - 30 years |
|
|
3. 31 - 40 years |
|
|
4. 41 and above |
|
3. Place of delivery |
1. Home |
|
|
2. General hospital |
|
|
3. Private hospital |
Questionnaire design
Questionnaires help to obtain accurate information relevant to the purpose of the study. In designing the questionnaire bear the following in mind:
Use simple and direct words that are familiar to all respondents.
Avoid technical jargons, slang or concepts familiar only to those with specialised training.
Ensure the questions are clear and not too general, complex or ambiguous.
Avoid double-barreled questions. It is often tempting to save time and space with questions covering two or more issues at the same time.
Do not use leading or loaded questions. A leading question pushes the respondent in the direction of a certain answer.
Ensure that the questions are applicable to all respondents. For example when a question such as "How many children do you have" is asked, the question assumes that the respondents have children. Inapplicable questions are not only irritating and confusing but potentially misleading. The respondent with no children may give a figure just to save himself from embarrassment or simply to oblige the interviewer: He wants me to give an answer, so why not. To avoid the above, use contingency questions or skip instructions to weed out respondents who do not fit the specific category. See examples below:
Table 7: Contingency question format.
Have you ever fed Soya beans to your family? |
||||||||||||
Yes |
No |
|
||||||||||
If yes: When was the last time you fed soya beans to your family |
||||||||||||
Last night |
Last week |
Last month |
Last year |
cannot remember |
Table7a: A typical Skip Instruction Format
12. Have you ever grown soya beans before? |
Yes (Please answer questions 13-20) |
No (Skip questions 13-20 and go to question 21). |
Constructing the instrument
The most important things to bear in mind when constructing questionnaire instruments are that the questions posed to collect information should not provide answers to the respondent. The instrument must be translated into the local language to ensure that people who administer the questionnaire do not invent their own translations and thereby give their own interpretations to the questions.
Start the questionnaires with less sensitive questions about the respondent's background. This helps to get the respondent relaxed and build rapport. Introduce each section of the questionnaire in such a way that it is clear and coherent to the respondent.
Baseline questionnaires, especially, in on-going projects, contain the following categories of information to be collected from the community. The list below can also be adapted for a project under formulation. In general, this information should come out of the PRCA result analysis. The categories listed below provide the basic topics and concepts to be included in a questionnaire. They cover both the subject matter and issues related to communication.
Demographic information
sex;
age;
occupation;
status in the household;
status in the community;
educational level.
Perception, awareness and knowledge of subject-matter
peoples' perceptions of issues, problems and solutions and opportunities;
levels of awareness of issues, problems and solutions and opportunities;
levels of knowledge and comprehension of issues, problems, solutions and opportunities;
information needs on issues, problems and solutions and opportunities;
Attitude field
peoples' feelings, opinions, beliefs or attitudes towards issues related to the subject matter: needs, problems, solutions and opportunities;
levels of approval of the issues;
Practice field
people's current practices;
peoples' willingness to try the new idea;
people's readiness to adapt or adopt the new ideas;
perceived obstacles to the trial, adaptation or adoption of the new ideas;
drop out rate of people who initially adopted the new idea;
why people stopped using the new ideas.
Communication and information resources and systems
communication and information resources and systems;
common sources of information in the community;
preferred sources;
traditional sources;
modern sources;
frequency of use of source;
languages;
sound, music tune associated with the issue;
visual image-symbols-colour associated with the issue;
animals, insects, reptiles any creatures associated with the issue;
events, history related to the issue;
smell, touch, taste associated with the issue;
positive and negative appeals relevant to the topic;
people's role models and opinion leaders on the issues under investigation;
people's perception of extensionists' level of performance;
what people think of the extensionists;
committees and organisation in community.
Level of people participation
ability to make decisions concerning development issues;
sense of ownership of project;
level of people's involvement in needs assessment;
level of people's involvement in solution identification;
people's involvement in the design and pretesting of communication materials;
people's involvement in committees and associations in the community.
Data analysis
PRCA baseline data analysis can be done manually because the samples are often small 100-150 respondents. However, with bigger samples the use of such computer statistical software as. Epi - info and SPSS is recommended. The popular simple data analysis processes are described below.
Data cleaning and recording
On a daily basis, hold debriefing meetings with interviewers, check if all questions were answered and if all responses were correctly recorded.
Remember the questionnaire is pre-coded. Collected data can, therefore be recorded or transferred directly from the questionnaires to data sheets, or into a computer. A data sheet for recording the age distribution of seven respondents by gender is shown below. Data on other variables such as level of awareness etc should be recorded on the sheet also.
List each variable on the sheet in the same order as it appears on the questionnaire. For a small study all the codes can be entered into a single large sheet and used for data analysis.
Data Sheet
The figures 1 to 7 at the head of the columns represent each respondent while value 1 depicts a respondent in the indicated age group and 0 indicates that there is no respondent in the age group.
Respondents
Table 8: Example of a data sheet.
Variable |
1 |
2 |
3 |
4 |
5 |
6 |
7 |
Total |
Age |
|
|
|
|
|
|
|
|
Less than 20 |
1 |
0 |
0 |
0 |
0 |
0 |
0 |
1 |
21-25 |
0 |
1 |
0 |
0 |
0 |
0 |
0 |
1 |
26-30 |
0 |
0 |
1 |
0 |
0 |
0 |
0 |
1 |
31-35 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
36-40 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
41-45 |
0 |
0 |
0 |
0 |
1 |
1 |
0 |
2 |
46-50 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
51-55 |
0 |
0 |
0 |
0 |
0 |
0 |
1 |
1 |
56-60 |
0 |
0 |
0 |
1 |
0 |
0 |
0 |
1 |
Sex |
|
|
|
|
|
|
|
|
Male |
1 |
0 |
1 |
0 |
1 |
0 |
1 |
4 |
Female |
0 |
1 |
0 |
1 |
0 |
1 |
0 |
3 |
From the sheet above it can easily be seen that there are four men and three women in the sample. In the same vein the number of men or women in each age group can be identified. For instance there is only one respondent in the less than 20 age group and he is male.
Univariate analysis
This is the examination of only one variable at a time for purposes of description and establishing distribution. The most basic format for doing this is to relate all individual respondents under study in terms of the variable in question. For instance, in a study to find out the views of villagers on borehole water, all the respondents' ages (individual cases) can be listed for example, 10, 20, 45, 62, 33, 15, 28. Although this listing will provide the fullest detail of the ages of all the respondents, it is cumbersome. The alternative is to use percentages in expressing this finding. For example, it can be reported that x percent of the respondents were between 10 and 15, y percent were between 16 and 20 and so on.
Table 9: Age of baseline study respondents (hypothetical).
Under 35 |
9% |
|
|
36 - 45 |
21% |
|
46 - 55 |
45% |
|
56 - 65 |
19% |
|
66 and older |
6% |
From the above table you can easily see that the largest age group who participated in the study is actually 46 to 55.
Bivariate analysis
Use a two-way-table or cross-tabulation to analyse two variables for purposes of describing and explaining their relationship to each other. For example, assume that the study in Table 9 above is on respondents' views on borehole water. Assume also that men and women were interviewed.
Notice that the table below could be regarded as an instance of subgroup comparison. While it describes each group independently, it also shows comparatively and descriptively that the women under description view borehole water as safer than the men do. Do you agree that borehole water as safer than the men do.
Table 10: Bivariate relationship between gender and water safety.
|
Men |
Women |
Agree |
63% |
75% |
Disagree |
37% |
25% |
|
100% |
100% |
Total number of responses |
200 |
200 |
Number who failed to respond |
12 |
5 |
The same table, seen as an explanatory bivariate analysis tells a somewhat different story. It suggests that the variable gender has an effect on the variable water safety. In this instance we are no longer talking about men and women but about the independent variable gender as an influence on the samples' and the populations' perception of water safety.
Multivariate analysis
Use this process to analyse more than two variables at a time to find out the relationships among them. For instance, when there is a need to examine simultaneously the effects of gender, age and social class on water safety, Multivariate Analysis can be used.
Let us return to the example of attitude towards water safety described above. Suppose that it is suspected that age has something to do with the answers: that younger people will be more ready to agree that bore hole water is safer than water from other sources than older people. The first step will be to divide the total sample into several subgroups based on the attributes of gender and age, that is, men and women and young men and young women. Then these subgroups will be described in terms of the dependent variable and then a comparison will be made.
Do you agree that borehole water is safer than water from other sources?
Table 11: A multivariate relationship of gender, age and water safety.
|
Under 30 |
30 and above |
||
|
Women |
Men |
Women |
Men |
Agree |
90% |
78% |
60% |
48% |
Disagree |
10% |
22% |
40% |
52% |
Total |
100% |
100% |
100% |
100% |
Total number of responses |
100 |
100 |
100 |
100 |
Number who failed to respond |
2 |
10 |
3 |
2 |
The table above is inefficient but correct. However, to simplify the results, they can be presented as in the table below. Notice that the disagree section has been taken away.
Do you agree that borehole water is safer than water from other sources?
Table 11a: The multivariate relationship expressed in percentages.
|
Men |
Women |
Under 30 |
78% |
90% |
Respondents |
100 |
100 |
30 and over |
48% |
60% |
Respondents |
100 |
100 |
Use the PRCA findings to sharpen the interview questions. It is generally necessary to drop or add questions depending on the outcome of the appraisal.
Use about 7-10 respondents from the community who are similar to the actual study sample for the pre-testing. Pre-testing helps to determine whether or not both the interviewers and the respondents understand the questions, and to reconfirm the relevance of questions and data to be collected. It is used to find out how long it actually takes to administer the questionnaire. Generally, it should not take more than forty-five minutes to one hour to administer a questionnaire. If the questionnaire takes too long to administer, find ways of shortening it without loosing the important aspects. Use the pre-testing exercise to standardise interviewers' interviewing, probing and recording techniques. Revise the questionnaire one more time to incorporate the results of the pre-testing.
Ensure that the study and its purpose are made known in advance to everybody in the community. Notify the relevant public authorities (e.g. village head, farmers' associations, etc.) before the study.
Explain that the study does not have anything to do with land acquisition, taxes or military service, for instance.
Become very familiar with the questionnaire so that when reading it to the respondent it is accurate and yet conversational. It is easier to administer a questionnaire if the respondent thinks that he or she is having a conversation rather than an interrogation.
Whenever possible conduct the interview where activities concerning the purpose of the study are taking place. For instance, If the study is about erosion in farmlands, conduct the interviews on the people's farms.
Approach the respondent in a respectful manner in order to gain his/her confidence and overcome any possible fears before administering the questionnaire. Make the respondent feel important because he or she has the information the study needs. Always assure the respondent about the anonymity of his or her responses. Always explain to each respondent the nature, purpose and time required for "discussing" the questionnaire. To motivate respondents, always stress the value, relevance and benefit of the study to them. The respondents are more likely to co-operate if they think that they will ultimately benefit from the study.
The interviewer is a neutral collector of information. The interviewer's presence should not affect the respondent's perception of a question or the expected answer. The respondent should never think that he is expected to give a certain answer.
Ask only the questions relevant to the affected respondents. In a study about birth control, it would be irrelevant to ask men if they take birth control pills or to stand in front of a man and ask him his sex.
Probing
Sometimes, the interviewer needs to probe for responses, especially, when the respondent is answering a question in a way that would force the interviewer to make a judgment about what the respondent actually means. For instance when multiple categories are presented such as agree, strongly agree, disagree, strongly disagree, don't know, a respondent might give an answer that does not fit into any of the categories by saying something like 'it is true'. It is necessary to prompt the respondent with such words as 'can you say you agree or agree strongly'. At times, it is also necessary to explain why such a 'concrete' answer is required from the respondent.
At the end of the interview, ensure that all relevant questions have been answered and ask if the respondent has any questions about the study. Do not forget to thank the respondent for his or her collaboration and time.
Supervision
The team of enumerators must be supervised daily as they administer the questionnaire in the field to ensure that the raw data or unprocessed information being collected from respondents is of good quality. Random checks on the interviewers should also be carried out while they are in the field to ensure that they are asking the questions correctly.
Interviewers should be debriefed at the end of each day to find out what experiences they had during the day and the difficulties they faced. If they are finding it difficult to collect certain information this should become obvious during the debriefing and a solution found. Keep record of any difficulties in case explanations are needed for certain discrepancies during analysis and presentation of results.
Planning and conducting the baseline study is a great achievement but it might as well have been a waste of the team's time and energy if they cannot present their findings in a persuasive and effective manner. To do a proper presentation it is essential to define the various audiences for the presentations and think through how each presentation can be adjusted to win the sympathy and understanding of each one of them.
The first thing to do after collecting and analysing the baseline study data is to write it up as a report using tables, graphs, anecdotes (or real quotations from respondents) and charts to clarify the findings. Use audio-visual presentations in the form of video, photographs, or even drama to add punch and emphases to the presentation. At the end of preparing the baseline report, it should be integrated with the PRCA findings to produce a single comprehensive report for the communication research.
The Executive Summary
This should not be more than one and a half pages long and should contain a summary of the major findings of the study and their significance as well as a summary of the recommendations. This should be written in such a way that people who don't have time to read the entire report could get the gist as quickly as possible. If the team succeeds in making this section of the report persuasive, it might even make such people eager to read the entire report.
The Introduction
This should contain the following sections:
Background to the study and why the study;
Where and when the study was conducted;
Who paid for the study;
Who conducted the study;
What methods were used, how the sample was selected and why?;
Practical problems or limitations encountered;
Reliability of results.
Presentation of findings
This section is the heart of the report. It should point out the findings of the study and their implications to the purpose of the study. This is where the tables and graphs appear and they should be explained with text. Do not repeat the content of the tables and graphs in the text; instead show the reader the importance of the findings and relate these to the issues under discussion.
Conclusions and recommendations
This is another part of the report that people who are in a hurry tend to read. This means that it should be given a lot of thought before it is written. No unsupported claims should be made since many readers use the quality of this section as a yardstick for measuring the whole work.
Do not assume that the conclusions and recommendations are cast in stone. They should be seen as the starting point of discussions and even debates on the line of action to be taken to address the issues revealed by the study.
Often the veracity or truthfulness of the findings is questioned during presentations. Be ready to defend them in a diplomatic but emphatic manner with information in your possession.
Bibliography
List the books and publications consulted for preparing the study.
At the end of preparing the baseline report, it should be integrated with the PRCA findings to produce a single comprehensive report for the communication research.