Use more open ended questions in your quantitative research

Open ended questions are often used only sparingly in online survey questionnaires. A major reason for this is that their results need to be coded after the survey. This can add cost, and cause a delay in preparing the overall survey results.

 

But now it is possible to have many types of open-ended question coded automatically, whilst a survey is underway. This works by the survey software recognizing certain words (set by the programmer on instruction from the questionnaire designer), and then giving them a code if they are posted within an answer.

 

For example, take a question which might usually be programmed as closed single response. This might be “At which store do you do most of your grocery shopping?”

 

This is easy to set up instead as open-ended, because the choice of possible answers is clearly defined. The answer “Aldi” can be given a code 1, “Asda” a code 2, and so on. In addition, the programmer can include a check to ensure that respondents only write one word.

 

Even if this question is intended to be multi response, this can still be programmed as an open ended, and automatically coded by the survey software. For example, this might be: “At which stores do you do most of your grocery shopping?”

 

The advantages

One advantage of this approach is that respondents have to read fewer words within the question, thereby reducing fatigue. It can also be helpful when respondents are answering on a mobile device, because they will not have to scroll through a list of possible answers.

 

Another advantage is that it removes the risk of respondents selecting the first possible pre-coded answer in a closed question. Whilst this first order effect can be mitigated using randomisation, the open-ended approach is superior because it removes the risk of respondents just clicking on one of the first answers they see, and instead thinking about their answer.

 

Finally, perhaps most importantly, the open-ended approach also allows respondents to answer in their own words, without being prompted or influenced by the questionnaire designer.

 

Other types of question

The examples above are relatively simple, because the answers are expected to be just single words. But what about more complex questions? Can these also be made open-ended rather than closed? Absolutely...perhaps not all of them, but many.

 

For example, maybe the questionnaire designer wants to know the reasons why a respondent shops at a particular store. They could create a closed question which includes such pre-coded answers as “price, location, parking, service, etc…”

 

But equally, they could make this open-ended, and this would mean that respondents have to give their answers in their own words.

 

Summary

Online survey software is getting better and better. The ability to automatically code open-ended answers during fieldwork is just one example of this, which, in a sense, allows surveys to produce more qualitative results.  

 

To find out more or for other examples please contact me.

Leave a comment