Many times, we have a bunch of data from surveys that we cannot do anything with. To fix this problem, we created a cheat sheet that you can use to create an analyzable survey. We will focus online survey because they are the most popular and the easiest to import data from. First, you have to select a target group and a goal you want to reach with your survey. Let’s say you have everything to start the project: you know your target group, goal, and method. The next step is creating the schema of your survey.
Four Steps of Questions
Start with a generic question and continuously dig deeper. At the end, you can ask for personal or sensitive data.
- Questions about knowledge: “Do you know these brands?”
- Questions about usage: “What kind of wine do you drink at dinner?”
- Questions about opinion or attitude: “What do you think about this brand?”
- Statistical questions: “Data about the Interviewee that you can use to segment the data or quantitative data”
'Most people start with sensitive or statistical data, such as gender, salary, or political affiliation, but this is one of the biggest mistakes they could make'
Six Golden Rules
Remember to follow these rules:
- Start with general parts
- The survey’s introduction affects the fillers, so write it well
- Smooth out the transitions between the questions
- Write and order questions so that they follow one another in a logical way
- Place sensitive questions at the end of the survey
- Place personal questions at the very end of the survey, so that if respondents do not want to answer them, you will still have the rest of the answers
The literature distinguishes two types of questions: open-ended and close-ended. In the case of open-ended questions, the responder has the opportunity to write in his/her answer. In the case of the close-ended questions, there are pre-set answers, and the respondent chooses from them. If we dig deeper, we find that there are two sub-segments in close-ended questions: alternative and selective.
Open-ended question – What is your opinion about global warming?
- Alternative: Do you believe in global warming?
- Selective: Which brand or brands do you know from the following list?
Any type of question can work for your survey it just depends on what kind of information do you want. However, the answers from close-ended questions are much easier to analyze and organize.
|Factor||Reasons for Open-Ended Questions||Reasons for Close-Ended Questions|
|Goal||We would like to collect quotes and written opinions.||We would like to have ordered, scaled, or organized data.|
|Characteristics of the interview||We want the respondents to answer in their own words.||We prefer if the respondents do not use their own words.|
|Method of Questioning||We do not have a list of the possible answers.||We want the respondents to choose from pre-set answers.|
|Analysis||We can handle many different answers and notes.||We would like to have the count or sum of the answers.|
|Reporting||We will provide personal or group information through words.||We are reporting statistical data|
One of the biggest mistakes is analyzing the answers one by one. Therefore, if you want a complex view of your survey then you will have to look for connections between the answers. After doing so, you can be sure that no important information is hidden. Moreover, you will find connections between answers that you would have never thought would be useful.
Survey tools such as Survey-Monkey or Google Survey have limited analysis functions that may mislead you or are good for descriptive statistical analysis only. Do not waste your time with half-solutions. If you really want to understand your data, use a tool like AnswerMiner. You can easily drag and drop your spreadsheet, and you are set. It even works with semi-dirty data. 😉