How to Code Open-Ended Responses with Multiple Mentions in Q

Introduction

Coding open-ended responses can often be a tedious task for market researchers. In this post, I will show you how to easily code multiple open-ended text variables into a multiple response question.

In quantitative surveys, there is often a need to ask open-ended questions to get deeper insights into a subject beyond a predetermined list of options. I have already covered the topic of coding a single open-ended variable into a multiple response question in Q. This used a sub-optimal format that mixes multiple brand responses within the same text variable.

The best format for brand awareness is generally considered to collect one brand per text variable. As a result, coding multiple open-ended text variables is a more common scenario for spontaneous awareness. In the following example, I conducted a survey where I included the open-ended question, ‘Please name all the cola brands you are aware of’. I collected this as 3 text variables and now wish to code this as a multiple response or Pick Any question in Q. I will now outline the process below.

Requirements

A data set with an open-ended Pick Any question

In the following example, I conducted a survey where I included the open-ended question, ‘Please name all the cola brands you are aware of’. I collected this as 3 text variables and now wish to code this as a multiple response or Pick Any question in Q. I will now outline the process below.

Method

Setting up your coding

How you set up your open-ended questions in Q can impact how the coding output will appear when coding multiple text variables at once. Selecting multiple variables that are grouped together as one question will create a single coded question, whereas selecting multiple variables that are not grouped together will lead to a separate question per text variable. Below is a table showing the various format combinations:

As I want to create a single Pick Any question, I will group them into one question. We can do this by highlighting these text variables in the Variables and Questions tab, right-clicking, selecting Set Question, and then choosing Text – Multi as the Question Type and OK.

Coding your open-ended question

Now that I have set my question, I can begin coding. Simply right-click the question once again and

Select Insert Variable(s) > Code Text > New Code Frame > Manual Categorization > Mutually Overlapping Categories.

This then brings up the following coding screen:

Steps:

On the right is where you build your code frame. If you right-click, it will bring up a context menu that allows you to add codes. There are other features too for renaming, adding subcodes, and deleting codes. You use Add Category when you want to individually add codes one by one or Import/Export Category Names to paste or type in multiple codes at a time.
Once you’re ready to start allocating responses to the code frame, simply highlight the responses on the left. The grey numbers in brackets represent the number of responses that fit this exact wording, regardless of case and leading or trailing spaces.
Now select one or more code options that you wish to allocate these responses to on the right.
Press Categorize as to apply your coding selection.

When you click Save Categories, your coding will be saved as new variables in your Q project. If you ever need to return to your coding, you can right-click on these variables and select Edit Code Frame. This allows you to continue your coding (if unfinished) or perhaps allocate new responses that may arise when your data file is refreshed.

Checking your coding

It is always good practice to double-check your coding. I will now outline how simple it is to reallocate coded responses.

Steps:

Go to the Show responses from: drop-down and select one of the codes that you created on the right. Here, I’ve selected ‘Coca-Cola’ and can see that ‘max’ has been mis-coded as it should be under ‘Pepsi’.
Next, click the ‘Max’ response and note the 3 buttons at the upper right:

Remove From – this uncodes the response from the selected code
Move to – this recodes the response so that it’s only associated with the new code
Add to – this additionally codes the response into another code

In this case, I will move ‘Max’ to the ‘Pepsi Max’ code by first de-selecting the ‘Coca-Cola’ category, then clicking on the ‘Pepsi Max’ category and selecting the Move to option to correct my coding.

When you make a mistake, it is also possible to press CTRL + Z to undo your previous coding step. Remember to press Save Categories to save any changes to your coding.