Conducting a “field” study (survey): the most common mistakes
I am writing another report and swearing. The fact is that we were ordered only an analysis, and the field (survey) was ordered from other contractors, we decided to save. As a result, the tasks prescribed in the technical specifications cannot be completed. From calculating the elasticity of demand to analyzing the results. I give a short list of recommendations for those who are going to conduct a survey in order to obtain certain data and analyze them.
1. Clearly define the goal. The more specific the goal, the wider the range of questions, the rigidity of the sample, the requirements for those who will conduct the survey.
2. Create a questionnaire. Themselves. And already go with her to request a survey. Do not trust the questionnaire to those who will conduct the survey. They can only be trusted to shovel and optimize the questionnaire. But you have to make the frame yourself.
3. Create a sample structure. Up to the details and instructions for interviewers. Up to the phrases "do not interrogate homeless people and emo." Be sure to indicate the places (areas, points) where the survey will be conducted.
4. If you are working with "field workers" for the first time, be sure to check their work, send "secret buyers / respondents to questions."
5. If you entrust the analysis of the results to third parties / companies, be sure to coordinate the questionnaire and sample design with them before the survey.
I will be glad if you write in the comments your recommendations on the topic, be sure to include them in the body of the topic.
1. Clearly define the goal. The more specific the goal, the wider the range of questions, the rigidity of the sample, the requirements for those who will conduct the survey.
2. Create a questionnaire. Themselves. And already go with her to request a survey. Do not trust the questionnaire to those who will conduct the survey. They can only be trusted to shovel and optimize the questionnaire. But you have to make the frame yourself.
3. Create a sample structure. Up to the details and instructions for interviewers. Up to the phrases "do not interrogate homeless people and emo." Be sure to indicate the places (areas, points) where the survey will be conducted.
4. If you are working with "field workers" for the first time, be sure to check their work, send "secret buyers / respondents to questions."
5. If you entrust the analysis of the results to third parties / companies, be sure to coordinate the questionnaire and sample design with them before the survey.
I will be glad if you write in the comments your recommendations on the topic, be sure to include them in the body of the topic.
