Coding with context

Introduction

When coding Market Research data, it's often very important for the coder to be able to view context-specific information alongside the verbatim data being coded. In this article, we take a look at why that's true with the aid of a few examples. We'll also have a look at how codeit handles the display of context data to make sure coders have all the information they need while coding verbatims.

Soap? Chocolate or Beer?

Imagine you've been given the task of doing some simple Brand Coding.

You've been given some data gathered from a consumer survey that contains brand mentions including the following:

• Milky Way
• Dove
• Budweiser

Pretty easy, right? Well, a lot might depend on which country the data was collected in. In the US, "Milky Way" is what people in the UK call a "Mars" bar. "Dove" in many countries is a type of chocolate that in the UK would be called "Galaxy" - and just for good measure, "Dove" in the UK is a brand of soap.

Lastly, if you're asking for a "Budweiser" in the US you'll probably get a different beer than if you're in Europe.

So clearly, in this example, you would want to display the relevant country for context alongside each verbatim, for example:

Questions, Questions, Questions...

The Country example above shows the example of displaying demographic information, but it's often more common to display data from other questions from within the survey. For example, imagine you've been given the task of coding some data from an NPS (Net Promoter Score) survey.

In this kind of survey, respondents are asked how likely they would be to recommend a given company, product or service on a scale from 1 to 10.

The respondent is then asked, using an open-end question, to explain why they gave the rating they did.

So, if you're looking at a verbatim like "because of the service", it's hard to know whether to code this as positive or negative without being able to see the corresponding NPS rating score. Context variables to the rescue once again!

Another example occurs when respondents give an answer like "see my previous answer" or "same as before" - the respondent is referring back to other questions in the survey, so it's important to be able to see these responses on the screen when coding.

Loops, Grids and All That

In the Market Research world, there are a few weird and wonderful ways of collecting data that put an additional spin on the "context question" issues above.

Suppose you have a survey as follows;

Q1. What hair care products have you bought in the last 7 days?

Q2. <for each product mentioned in Q1> "What brand of <product> did you buy?"

Question two in our survey is asked multiple times in a loop - it is just one question that generates multiple answers, each with a different context value from Question one.

This might generate some data that looks like this:

As a coder, you will want to code Q2 as a single unit, but it's important that the correct data from Q1 is displayed relative to each iteration.

Fortunately, codeit fully understands these kind of quirks of the Market Research world and makes it easy to import the data and display it in the way you'd expect:

Conclusion

Context is sometimes very important and codeit's powerful data features make it easy to display the right context data when coding Market Research data.