Quantitative surveys are a data collection tool used to gather close-ended responses from individuals and groups. This often takes the form of a Knowledge, Attitude and Practice Survey T.23, a Knowledge, Practice and Coverage survey (both can be done as baselines and endlines), quantitative Doer/Non-Doer Survey T.32 or perception and/or Monitoring M.2 surveys.

Question types primarily include categorical questions (e.g. ‘yes/no ‘), numerical questions (e.g. number of people in the household that use the latrine) and interval/ratio questions (e.g. rating-scale, Likert-scale). They are used to gather information on behaviours, coverage, practices, characteristics, attitudes or demographic information. Surveys are completed using paper questionnaires or, more conveniently, tablets. 

Depending on the methodology used, surveys can be a relatively quick, cost-effective and extensive data-gathering method. Short questionnaires with standardised questions and answers can provide a large data set involving large sample groups in a short amount of time. 

Different types of surveys can be done before, during and after a WASH response, but surveys are commonly used to design (e.g. quantitative Doer/Non-Doer Survey T.32) and measure the effectiveness and impact of a WASH response (M.2 and M.4).  

Process & Good Practice

• Process Steps:

   

   

• Step 1: Define the survey objectives. A well-executed and successful survey begins with well-defined survey goals and objectives. To write effective goals, start with an action verb (such as describe, explain, explore, identify, investigate, gauge, measure, assess or test) followed by the issue under enquiry. For example, ‘to understand the knowledge of the displaced population in the province X regarding the prevention, diagnosis, care and treatment of diarrheal diseases in children under 5 years old ‘. 

• Step 2: Define how the survey will be implemented to ensure high quality and consistently collected data. Consider the sampling frame, method, the protocol for respondent selection, verification of data, ethical and safeguarding measures and a basic plan for data management and analysis. The timeframe and resource (administration, budget and logistics) requirements should be clearly defined at this stage of the process and coordinated with the support departments of implementing agencies, local authorities and/or camp management authorities.

• Step 3: Design the survey questionnaire. No survey can succeed without a well-designed questionnaire. There are many good examples of survey questionnaires that can and should be adapted to the specific situation. In general, the questionnaire includes the questions to be asked, as well as a list of response choices which enumerators use to record the response. The wording of both of these items should be clear, precise and accurately translated into the language(s) spoken by the questionnaire respondents. Once the questionnaire tools have been developed, test each tool in the field to see how well the questions are understood by a small sample of respondents and identify any problems with the administration of the questionnaire. Where electronic devices are used, geo-data and time-stamps can be used to verify sampling and interview protocols.

• Step 4: Conduct the survey using the appropriate protocol for the method chosen. Data quality control measures must be in place during fieldwork, e.g. Quality Improvement Verification Checklists that are performed daily by supervisors. Before field deployment, a clear strategy should be developed to support enumerators during collection to resolve problems that arise and to assess and maintain data quality as data is collected.

• Step 5: Data analysis: survey results must be analysed. Ideally, data entry and cleaning start during or shortly after fieldwork. This saves time and enables data quality issues to be identified while there is still time to rectify them. Verification and validation of the data are important steps in survey quality control. 

• Step 6: Use of the data: the results of the survey must be presented in a final report (using visuals such as graphics where possible) according to the objectives of the survey. The report, however, is only the beginning. It is important to use the findings of the survey and to work as a team to address the challenges identified. The results and action plan should be disseminated to all stakeholders in the WASH sector who are likely to benefit from the information collected, as well as to the populations that participated in the survey.

• The survey team is generally made up of a programme manager, supervisors and teams of enumerators, including men and women who are fluent in the local language.

• The trained enumerators may be hired for between one and two weeks (maximum three) depending on the number of respondents and the number of interviewers recruited – if existing staff capacity is not available.

• The supervisors’ role is to move from one team of interviewers to another to help resolve any difficulties encountered. Supervisors are the guarantors of the survey process, checking that the interviewers follow the correct respondent profiles and fill in the questionnaires correctly. 

• The technical advisors and project managers may be too busy to manage detailed aspects of the survey: the support of a methodology assistant from the beginning and at key moments in the process can help. Collaboration with other individuals or organisations may be necessary to determine the sampling plan (the number of people and areas to be interviewed), create/adapt questionnaires, conduct interviews in the local language and enter or analyse data.

• Ensure that your questionnaire does not take longer than 45 minutes; test and improve the questions based on the feedback. Do not ask questions that will not be analysed and used. 

• The major methodological errors in sampling include insufficient sample size or the number of clusters, failure to sample proportional to population size (for cluster surveys) and failure to weight the sample during analysis.

• Two key factors affect the representativeness of the sample for surveys: (1) accessibility of survey participants and (2) general inequalities which may prevent or reduce the participation of certain groups in surveys.

• Disaggregated data will often be an important component to identify people with a disability. The 6-item Washington Group Short Set of Disability Questions is recommended. The questions assess whether people have difficulty performing basic universal activities such as walking, seeing, hearing, cognition, self-care and communication. 

• Accessibility may be restricted and undermine the survey. Restrictions on the movement of enumerators may be caused by insecurity, logistical challenges, transport problems, remoteness, weather, surveyors being refused access and natural disasters and make it impossible to reach all areas identified through the random sampling process.