Published November 2026 · 8 min read

How to Code Open-Ended Responses From a Qualtrics Export

Qualtrics is great at collecting open-ended answers. Getting them coded is another matter: Text iQ is a paid add-on that keeps the work inside Qualtrics, and a raw export arrives with quirks that trip up most other tools. This guide covers which export to choose, the three things to fix in a Qualtrics file before coding, and how to get your text entry answers back as coded 0/1 columns ready for analysis.

Where the Open-Ended Answers Live in a Qualtrics Export

Open-ended text appears in two kinds of column:

• Text entry questions are exported under the question's export tag, for example Q4, or whatever you renamed it to in the survey editor.
• "Other (please specify)" boxes attached to a multiple-choice option are exported as _TEXT columns. Q2_6_TEXT is the text typed next to choice 6 of Q2. They are easy to overlook, and they are often where the most useful surprises are.

Everything else (StartDate, Duration (in seconds), ResponseId, DistributionChannel and so on) is metadata. You do not need to delete it before coding, because you choose which columns to code. Keep ResponseId: it is your key for joining the coded file to anything else.

Choosing the Export Format

In Qualtrics, go to Data & Analysis → Export & Import → Export data. You will see several formats. For coding, two matter:

If you analyze in SPSS (or can open a .sav in R or Stata), choose SPSS. Our SPSS coding guide covers what the output looks like. If you live in Excel, choose CSV with "Use choice text" so closed questions read "Very satisfied" rather than "5", then apply the fixes below.

Three Fixes Before You Code a Qualtrics CSV

1. Delete header rows 2 and 3

Open the sample file and look at the first three rows:

StartDate,EndDate,...,ResponseId,...,Q4,Q5
Start Date,End Date,...,Response ID,...,What one thing would most improve your library?,Why did you give that rating?
"{""ImportId"":""startDate""}",...,"{""ImportId"":""QID4_TEXT""}","{""ImportId"":""QID5_TEXT""}"

Row 1 is the column names, which is what you want. Row 2 is the question wording and row 3 is internal import metadata. Any tool that expects one header row will treat rows 2 and 3 as two extra respondents. In a coding tool, that means the question text and a JSON string get coded as answers and inflate your counts.

Fix: before deleting row 2, copy the question text for each open-ended column somewhere handy, because you will paste it into the context box later. Then select rows 2 and 3, delete them, and save. Excel will offer to keep the CSV format; say yes, or save as .xlsx.

2. Remove previews and test responses

Survey previews and test runs are exported with everything else. They show up in the Status column (as "Survey Preview" or "Survey Test" with choice text, or 1 and 2 with numeric values). Filter them out, along with incomplete responses (Finished = 0 or False) if your analysis plan excludes partials. Coding junk costs credits and pollutes the category shares.

3. Find every _TEXT column

Search the header row for _TEXT. "Other (please specify)" answers often reveal the options your questionnaire missed. Coding them tells you whether next wave's list needs another choice. They are usually short and sparse, so they cost very little to code.

Coding the Text Entry Questions

1. Upload the cleaned CSV, the .xlsx or the SPSS export (up to 20 MB).
2. Select Q4, Q5 and any _TEXT columns. The picker shows a sample answer and the response count for each column, so you can confirm you picked the right ones.
3. Add context: paste the question wording you saved from row 2, plus a line on what the results are for. With an SPSS export, the wording is already there as the variable label.
4. Review the proposed categories (5 credits per question on packs, included on plans), or skip and type your own codeframe for free. Edit until the list matches your reporting needs, then approve.
5. Classification codes every non-empty answer against your categories, at 1 credit per answer. Answers can get more than one category.
6. Download the same file with 0/1 columns appended: Q4_cat_longer_opening_hours, Q4_cat_quieter_study_space… in CSV or Excel, or labelled Q4_c01… variables in SPSS, plus a _none column for answers that fit no category or were blank.

Because every column you uploaded comes back, ResponseId, the closed questions and any embedded data are already beside the codes. You can cross-tab Q4 themes by Q1 branch straight away, with no merge step.

What About Qualtrics Text iQ?

Text iQ codes inside Qualtrics and is a good choice if your organisation already licenses it and you want topics in Qualtrics dashboards. Exporting and coding outside Qualtrics makes sense when:

• your licence does not include Text iQ;
• you need coded columns in SPSS, Stata or Excel rather than topics in a dashboard;
• you combine Qualtrics data with other sources, or work for clients on different survey platforms;
• you want the same codeframe and process across every project, whatever platform fielded it.

We compare the options in more detail in Qualtrics Text iQ alternatives.