A mixed-method survey carries closed and open-ended questions on one record. A worked questionnaire example, the four formats, and how to design one.
A mixed-method survey is a single questionnaire that asks closed and open-ended questions together — and gathers the documents and transcripts alongside them. The two strands only integrate if every answer lands on the same record. Most surveys lose that link at the export, and the open answers become a column no one reads.
By Unmesh Sheth · Founder & CEO, Sopact · Updated May 2026
A mixed-method survey is a single questionnaire that collects both quantitative and qualitative data — closed-ended items such as ratings and scales, alongside open-ended questions, and often uploaded documents or transcripts. What makes it mixed method, rather than a survey with a comment box, is that both kinds of answer stay attached to the same respondent so they can be analyzed together.
A comment box at the end of a satisfaction survey is not a mixed-method questionnaire. The test is whether the open answer and the closed score can be read on the same record — if they cannot, the survey collected two things and joined neither.
A mixed-method survey is a single questionnaire that collects both quantitative and qualitative data — closed-ended items such as ratings and scales, alongside open-ended questions, and often uploaded documents or transcripts. What makes it mixed method, rather than a survey with a comment box, is that both kinds of answer stay attached to the same respondent so they can be analyzed together.
A comment box at the end of a satisfaction survey is not a mixed-method questionnaire. The test is whether the open answer and the closed score can be read on the same record — if they cannot, the survey collected two things and joined neither.
Four items from a workforce-training questionnaire. Two are closed-ended, two are qualitative — and the point is not the mix of formats. It is that all four write to the same record, so the strands are joined before analysis begins.
A number. Trends across the cohort and across rounds.
A narrative. Explains what the rating cannot.
A 5- or 50-page document, read on arrival, not skimmed later.
A second measure. Anchors the outcome question.
Four items, four formats — one record under one Contact ID. The closed scores and the open answers are already joined, so a joint display can be read the moment the responses arrive.
Q1 and Q4 are the quantitative strand; Q2 and Q3 are the qualitative strand. They are mixed method because they share a record, not because they share a form.
A mixed-method questionnaire collects in four formats. The first is the quantitative strand; the other three are the qualitative strand. What makes the instrument mixed method is not the count of formats — it is that all four land on one record.
Ratings, scales, and multiple choice. Fast to compare across people and rounds, and the spine of the quantitative strand — but they record the level, not the reason.
Free-text questions. They carry the reason behind the rating — the moment, the obstacle, the cause. Paired with a closed item, they make a construct readable on one row.
Uploaded reports, work samples, financials. Qualitative evidence the respondent already has — collected inside the instrument so it lands on the record, not in an inbox.
Interview or check-in recordings, attached to the same record as the ratings. The fullest qualitative source — and the hardest to read at scale without help.
A survey can carry all four formats and still not be mixed method — if the strands part ways at the export. The integration is decided by the record, then read in mixed methods data analysis.
The order matters. A mixed-method questionnaire is designed construct first, not question first — because the integration has to be planned into the instrument before a single response comes in.
List what the study has to measure — confidence, wages, engagement. Each construct, not each question, is the unit the survey is built around.
The rating gives the level; the open answer gives the reason. Pair them construct by construct, so the two can later be read on a single row of a joint display.
A work sample, a report, an interview is qualitative evidence. Build the upload into the instrument so it lands on the respondent's record, not in a separate inbox.
Every item has to write to the same record, so the strands stay joined at the source — and the same ID carries the respondent across every round of the survey.
Decide how the open answers will be coded against the closed measures now, not after collection. Then the joint display assembles as responses arrive, instead of months later.
Steps two and four are the ones a mixed-method survey is usually missing. They are also the ones that decide whether the study integrates — for the design types each questionnaire serves, see mixed methods research design.
The design above — one record, paired items, an aligned codebook — is hard to hold in a general survey builder, where the open text and the ratings part ways the moment the data is exported. The integration then has to be rebuilt by hand, if anyone has the weeks to do it.
Closed ratings, open-ended answers, uploaded documents, interview transcripts — all on the same record under one persistent Contact ID. A versioned rubric reads each narrative against the measures on arrival, construct by construct. The same ID carries the respondent across every round. A mixed-method survey built this way produces a joint display, not two exports that have to be matched by name and email afterward.
Bring your draft instrument — closed items, open questions, document uploads. We will map it onto one record and show you the joint display it produces.
A mixed-method survey is a single questionnaire that collects both quantitative and qualitative data: closed-ended items such as ratings and scales, alongside open-ended questions, and often uploaded documents or transcripts. It counts as mixed method only when both kinds of answer stay attached to the same respondent, so the rating and the narrative can be analyzed together rather than as two separate datasets.
There is no real difference — the terms are used interchangeably. "Questionnaire" names the instrument itself, the set of items a respondent answers; "survey" often names the wider activity of fielding it and gathering responses. A mixed-method questionnaire and a mixed-method survey both mean one instrument carrying closed and open-ended items that stay on one record.
Yes. A survey is mixed methods when it carries both a quantitative strand and a qualitative strand and the two are integrated, not just collected. A survey of ratings with one open comment box is not automatically mixed method; it becomes mixed method when the open answers are analyzed against the closed scores on the same record, construct by construct.
Not on its own. A semi-structured questionnaire mixes fixed and open questions, which gives it both kinds of data — but mixed methods requires that the two strands be integrated into one finding. A semi-structured questionnaire is a mixed-method instrument only when its closed and open answers are read together, not reported in two separate sections.
They are different ideas. Mixed methods means combining qualitative and quantitative data in one study. Mixed mode means fielding one survey through more than one channel — web, phone, mail, in person — to reach more respondents. A survey can be mixed mode without being mixed method, and mixed method without being mixed mode. This page is about mixed methods.
A mixed-method questionnaire pairs closed and open items construct by construct. For example: a 1-to-10 confidence rating (closed), then "describe a moment the training helped or did not" (open), then a work-sample upload (document), then a yes-or-no wage-change question (closed). The four items use four formats, but they are mixed method because all four write to the same record under one ID.
Design it construct first. List what the study must measure, pair every closed item with an open one for the same construct, decide where documents and transcripts attach, write one Contact ID into the instrument so every item lands on the same record, and align the codebook before fielding. Planning the integration into the instrument is what separates a mixed-method survey from a survey with extra questions.
Four. Closed-ended items — ratings, scales, multiple choice — form the quantitative strand. Open-ended items, uploaded documents, and interview transcripts form the qualitative strand. A survey does not need all four to be mixed method, but it does need at least one closed and one qualitative format, both landing on the same record.
Both, by definition. A mixed-method survey carries a quantitative strand (the closed-ended items) and a qualitative strand (the open-ended items, documents, and transcripts). The point of the design is not to choose between them but to integrate them — so the question a mixed-method survey answers is one neither strand could answer alone.
There is no single number. The quantitative strand needs enough responses for the comparisons the study plans to make; the qualitative strand needs enough for themes to recur and stabilize. In practice the closed strand sets the minimum, because statistical comparison needs more cases than thematic reading does. Size each strand for its own job, then field one instrument to the same people.
Yes, and it is stronger when it does. If the same Contact ID carries each respondent from round to round, the closed scores and the open answers can be compared over time, not just within one wave. A mixed-method survey run longitudinally turns a snapshot into a record of how an outcome and its causes moved together.
A comment box collects open text but rarely integrates it — the comments export to a column and are read once, if at all. A mixed-method survey is designed so the open answers are paired with closed items construct by construct and read against them on the same record. The test is whether the open answer and the closed score can be analyzed on one row; a comment box usually fails it.
This page covers the survey instrument. The pillar holds the methodology and the redefinition; the other four guides cover the design types, the analysis step, the worked examples, and the tools.
A working session, not a demo. Bring your draft questionnaire — closed items, open questions, document uploads. We map it onto one record, align the codebook, and show you the joint display the instrument produces. You leave with a mixed-method survey that integrates by design.
Live walkthrough · 30 min · with Unmesh Sheth, Founder & CEO · bring a draft questionnaire
