Diagnostic Tool


How to Conduct a Survey

Attention: open in a new window. PrintE-mail

Because the cost of implementing surveys is often high, it is important that they be planned and conducted with utmost care. In conducting a farmer survey, the following steps are recommended:

1. Identifying the problem/issues

The first step in planning a farmer survey is to identify the problem and issues that need to be addressed. In pest management, the choice of pest problem to focus on would depend on the needs associated with specific research priorities or the information needs of a given ministry or plant protection organization. Where such priorities have not been articulated, the farmer survey could gather information of value for developing research priorities.

2. Developing survey objectives

Once a priority problem has been identified, the next step is to develop the survey objectives. For instance, the objective might be to determine the likely impact of an extension campaign or the likely adoption of a new IPM practice. Drawing up a list of variables that will help find answers to the survey objectives could put the researcher on the right track in designing the questions to ask in the survey.  Specific questions  aimed at various aspects of the problem could help clarify the research problem.  It is important to remember that the choice of questions should be guided by the survey objectives.

3. Developing the survey instrument

In a farmer survey, a popular instrument used for data collection is a questionnaire, which contains a series of questions designed to gather information from the respondents. Depending on the survey objectives, the survey questionnaire may contain knowledge, attitude and practice questions on an identified pest problem, cropping patterns, demographic and socioeconomic background of respondents, among others.



A questionnaire interview of a farmer in Vietnam


4. Pretesting the questionnaire

When the farmer survey questionnaire has been compiled, it needs to be pre-tested before being copied and implemented in the field survey. Pre-testing involves interviewing a small group of respondents who are similar to the intended target group to determine their reactions to the prototype questionnaire. This is done to determine:

  • the clarity of the wording and translation of the technical terms used,
  • whether the questions are in a logical sequence,
  • the adequacy of the response categories (e.g., where there is a multiple choice),
  • the clarity of questionnaire instructions, and
  • the estimated duration of the interview.

Results of the pretest are used as the basis for revisions in the questionnaire and logistical arrangements for the fieldwork.


Pre-testing a questionnaire in Indonesia

5. Choosing sample respondents

An important concern in survey research is deciding how many and which respondents should be included. A farmer survey uses standard social science methods in selecting the sample, such as multistage sampling, stratified sampling, systematic sampling, cluster sampling and simple random sampling. The choice of sampling technique depends primarily on the nature of the problem, the cost and time factors involved, and the desired precision or reliability of the results. When a project can afford it, a larger sample is preferred to reduce sampling error. In a survey of, say, 300 farmers, it is recommended that the sample be drawn from a cross-section of the sampling population so that this group can be said to represent the larger population. In other situations, the main objective of the survey may be to identify the range of ways in which farmers might use a new practice, in which case selecting farmers from very different farming systems may be more appropriate.

6. Implementing the field survey

Once the questionnaire has been pretested, finalized and reproduced, the next step is to implement the field survey. Resources needed for the field work include personnel, money and time. A field survey team is often composed of a survey coordinator, a supervisor and interviewers. The survey coordinator is responsible for all aspects of the field work: selection, training and deployment of interviewers. The supervisor assists the survey coordinator in spot-checking and monitoring the field interviews.  Before they go to field, interviewers are oriented on the purpose of the survey and trained on interviewing skills and how to conduct the interviews. Guided by the sampling plan and respondent list, the interviewers locate the respondents, conduct the interviews, and check the completed questionnaires after the interview.

See Also Choosing a Field Interviewer and Guidelines for Implementing the Survey

7. Coding and analyzing survey data

Once the completed questionnaires have been edited, the data need to be analyzed. Depending on the main objective of the survey, this analysis phase can be relatively simple such as manually determining the % of respondents giving specific answers or listing the various ways in which farmers said they might use a new practice. For more complex surveys, particularly where the aim is to predict for the entire population from the results of the sample population, it is best to encoded, process and analyze the data using a statistical package. Ease of use, power and cost are some of the important considerations in the choice of computer software for data analysis.

First, the responses to questions need to be coded and tabulated. Coding is the term used to describe the translation of question responses and respondent information into specific categories for analysis. Tabulation is the recording of the numbers of types of responses in the appropriate categories, after which statistical analysis follows: percentages, averages and appropriate tests of significance.

Then the analyzed data are interpreted and the results are reported. The purpose of a survey report is to tell the readers the research problem, data collection methods used, findings and conclusions. Like other research reports, the survey report should consist of an executive summary, introduction, description of the methods, results and discussion, and conclusions.