Surveys
Surveys are used to gather a wide range of information about a population. They can be based on questionnaires and/or interviews (individual/ group/ telephone/email). The emphasis tends to be on quantitative data which is usually analyzed using statistical methods. The qualitative information is used to help interpret the quantitative findings. Surveys generally consist of three distinct stages or steps.
HOW TO DO Surveys:
Survey design
Formulate the questions you want the survey to answer. This can be in the form of a hypothesis which the survey will either prove or disprove. Select variables (any characteristic that can vary) that will allow you to answer the question. Design methods of collecting information about the variables. You can carry out physical measurements, interviews or distribute a questionnaire. The latter asks everyone the same question in the same way so answers can be compared (see M&E Tool n.3 Questionnaires). Think about translating questionnaires/ guidelines for interviewers. You are most likely to be working with a sample of the population (remember representation and random selection). Plan the data collection according to the population sample selected. Think of transport, accommodation and a timetable for your team. Take into account respondents’ daily routines to decide the best time for data gathering.
Data collection
You will need to train people to collect data (enumerators) and have supervisors for them. Produce guidelines to ensure consistence, bias avoidance, etc. Pilot the survey on a small group and amend methods/content if necessary. The data is then collected from the sample by enumerators and supervisors.
Data analysis
After the data is collected, collated and checked, it is analysed. Data collected using quantitative methods can be analysed statistically and this can give averages; ratios; range between minimum and maximum levels; statistically significant differences. This ‘hard’ data is often considered to be accurate and objective however issues of bias, reliability, and validity remain. If data collection methods have not been consistent, the results are not valid.
WHEN TO USE Surveys
Surveys can be used at the analysis, planning and reviewing stages of the development process. Since they provide accurate, precise data, they can be used to establish baseline information at the planning stage, which can later be used for evaluating impact. As part of the evaluation you would revisit the baseline information, and carry out a similar survey to compare ‘then and now’, or you would use the baseline information generated as a starter for discussions on impact and changes using other tools. A survey can be used for evaluation to identify the major differences and relationships in the characteristics of a population, and to find out how different sectors of the population have been affected by the project/programme. If no initial survey was used to generate baseline data, an evaluation survey can still be carried out, and the ‘before’ information can be generated using respondent recall and other techniques. Repeated surveys can be used to measure changes in a population over time – this is useful for monitoring the impact of a programme, or for detecting significant changes in selected populations.