The data we get from a survey is unhelpful unless you analyze it, and the findings may be inaccurate if you don’t analyze them correctly. Analyzing can make the findings hard to read even when the analysis is correct. The people supposed to make decisions based on the analyzed data should be able to see and understand the findings at a glance; someone should have done all the hard work at the analysis stage.

How best to analyze survey data.

i.                 Consider the four measurement levels.

There are four measurement levels for survey questions. They determine how particular types of survey questions should be measured and the statistical methods used. The levels are as follows with brief explanations.

–        Nominal Scale

This is the level at which data without quantitative value is analyzed. Questions like, do you like your job? You can only have one of two answers, ‘Yes’ or ‘No.’ The only analysis you can do with such data is to count the number of times a response appears and then record the one that appears the most. You can present the outcome in percentages or even graphically, but it is all counting.

–        Ordinal Scale

This scale works when the survey data is presented quantitative value where one value is greater than the other. Questions that start with ‘On a scale of 1-5…’ fall under this scale. For such questions, you can find the median, mean, mode, etc.

–        Interval Scale

An Interval Scale can show both differences between values and the order in which they occur. However, this kind of data has no zero point, and you can find mode, median, and mean and still analyze any notable trends using available tools.

–        Ratio Scale

This pertains to questions like, ‘how much extra time do you spend on average every month?’ There is an order difference between values, and it has zero as part of the possible answers. You can also find the mode, median, and mean for this kind of data.

Other measures to take.

    i.                    Start by analyzing quantitative data.

Quantitative comes from close-ended questions that can be presented as numeric values. One can convert this kind of data into numeric values, and it gives good, crisp insights into the survey findings. Analyzing quantitative data first can enable you to understand your qualitative data more. Qualitative data gives the survey questions, but it is usually subjective. Having the quantitative findings beforehand enables you to make the right deductions.

  ii.                    Cross-tabulate to understand your audience

This is not necessary if your survey is on a closed audience, for example, employees of a single company. However, when you have a customer survey where every walk-in participates, cross-tabulation may help because not everyone who comes is your typical customer. The views of your typical customers should inform your decisions more even as you consider other participants’ views.

   iii.                    Categorize the data in participants’ demographics

The age of respondents always gives good insights into how you should act because people of different ages have different outlooks. Always have people’s ages in mind as you do the analysis.