The whole reason for doing a survey is to collect the opinions of many people and then analyze the results in a way that leads to taking appropriate action. Otherwise why would you do it in the first place?

However, survey data analysis is a bit of a murky field sometimes, especially if the analysis wasn’t considered at the same time as the survey design. Once you have sent a survey out to people and they have answered it, it is extremely difficult to get them to re-do it if the survey wasn’t designed appropriately with the analysis in mind.

So in this article, we will explore a few simple ideas to ensure your survey can be analyzed in a meaningful way.

What is the purpose of your survey?

Generally speaking, surveys can serve a number of different purposes including measurement (how well is something happening or not happening), context (to explain why something is or isn’t happening), or confirmation (is this thing happening or not). Most surveys serve these and more at the same time, however it is important to consider from the outset what your primary goal actually is. This will help you to ensure you are asking relevant questions and are not making the common mistake of adding too many irrelevant questions that just make the survey longer.

Consistency of displaying scale questions

A scale question is simply where you ask people to rate something on a scale of say, 1 to 5, where each end of the scale is assigned a value on a continuum. For example, from 1 Strongly Disagree to 5 Strongly Agree. Notice here how the lowest value is the ‘negative’ option, whilst the highest number is the most favourable option. Often, a number of scale questions are used one after another, and they are a simple way to answer questions, good for the survey taker. Some people would suggest mixing up the scales to ensure people are paying attention, but our recommendation is to keep scale questions displayed consistently so that the unfavourable response is on the left, and the favourable response is on the right. This will ensure that there has been no confusion on behalf of the survey taker, and that your results will be easier to analyze.

A little bit of context is better than none

When you start to design a survey, it is very easy to keep adding questions because you will keep thinking of things to ask. The thought process is usually something like ‘whilst they are answering, we may as well ask them everything at the same time’ which is actually counter-productive. If a survey is much longer than 10 minutes, most people will switch off by then. So instead of asking lots and lots of scale or multiple choice questions, you might ask a question and follow it up with an open ended question. For example, if you were to ask How likely are you to buy from us again? Very unlikely, unlikely, not sure, likely or very likely ( a five point scale), you might simply add a follow up open-ended question like ‘Why did you give that answer?’. The context you get from this question will be invaluable and will reduce the length of the survey.

There are many other factors to consider when designing a successful survey, but if you start with these you will at least put yourself in the position to be able to generate results that are easier to analyze and interpret.

And of course if you need some help, simply get in touch with us.