Quantitative, Qualitative and Mixed: How to Combine Data Collection Techniques

Most discussions of data collection techniques start by splitting them into quantitative and qualitative, then stop. The more useful question is how the two work together: which one leads, how one informs the design of the other, and what that means for the instrument you build. This post looks at quantitative data collection, qualitative data collection and mixed methods research from the point of view of the person who has to design and script the study.

Two kinds of evidence, two kinds of instrument

Quantitative and qualitative techniques differ less in the topics they cover than in the shape of the evidence they produce.

The instrument is where the difference becomes concrete. A quantitative questionnaire has to anticipate the answers in advance and present them in a way every respondent interprets the same way. A qualitative guide has to leave room for answers nobody anticipated. Trying to make one instrument do both jobs usually produces a survey with too many open-ends or an interview that feels like a form.

Quantitative techniques in practice

Online surveys

Online surveys are the most common quantitative technique in commercial research. They are quick to field, handle images and video well, and allow complex routing that would be error-prone on paper. Their success depends on two things outside the questionnaire itself: a suitable sample source and a sound approach to quality. We cover the second in detail in online survey data quality checks.

Interviewer-administered surveys

Telephone and in-person surveys still matter where online coverage is weak, where the population is hard to reach online, or where screening needs a human voice. They use the same structured logic as online surveys but add interviewer effects, which have to be controlled through training and consistent scripts.

Experiments and choice tasks

Many quantitative studies embed experiments: randomly showing different concepts, prices or messages to comparable groups, or asking respondents to choose between bundles of features. These designs depend heavily on correct randomisation and rotation in the script, which is easy to get wrong and hard to repair after fieldwork.

Reporting what happened in the field

Quantitative work also comes with expectations about transparency. Bodies such as AAPOR publish Standard Definitions for classifying the final disposition of every sampled case and calculating outcome rates, which helps others judge how a sample was achieved. Even where a commercial study does not report formal rates, keeping clear records of completes, screen-outs, quota-fulls and drop-offs is good practice.

Qualitative techniques in practice

In-depth interviews

One-to-one interviews are the backbone of qualitative data collection. They give the moderator room to follow a thread, test an explanation and come back to something the respondent said earlier. They are especially valuable for B2B and professional audiences, whose reasoning is specialised.

Groups and communities

Group discussions and online communities add interaction. Participants respond to each other and to stimulus, which surfaces language and objections you would rarely hear one-to-one. Communities extend this over days or weeks.

Observation and ethnography

Observation records behaviour directly, whether in a shop, a home or a workplace, or through screen recordings in usability sessions. Ethnographic work goes further, spending extended time in the respondent's own setting to understand routines and context. These techniques are slow and need careful consent, but they answer questions about actual behaviour that no interview can, because people often do not notice, or cannot describe, what they really do.

Sampling for qualitative work

Qualitative samples are chosen on purpose rather than at random. The aim is to cover the range of views and experiences that matter, for example heavy and light users, or recent switchers and loyal customers, rather than to mirror the population in proportion. That makes the recruitment specification as important as the discussion guide, and it is why qualitative findings describe what exists rather than how common it is.

Open-ended questions inside surveys

Open-ends are a qualitative technique embedded in a quantitative instrument. A few well-placed open questions explain the numbers around them. Too many lengthen the survey and tend to produce thinner answers as respondents tire. They are also among the most useful places to spot low-quality responses.

The question format decides the data

Whatever the technique, the format of each question determines what you can do with the answer. This table is a practical reference for instrument design.

Mixed methods research: three common designs

Mixed methods research combines both kinds of technique in one project, deliberately rather than by accident. Three designs come up again and again.

1. Exploratory sequentialQualitative first to learn the issues and language, then a survey built from what was learned.
2. Explanatory sequentialSurvey first to measure, then interviews to explain surprising or important results.
3. ConvergentBoth run in parallel on the same topic, with findings compared to see where they agree and differ.

Choosing between them

• If you do not yet know what the important attributes or answer options are, use the exploratory design. Writing a questionnaire before you understand the topic is the most expensive mistake in this list.
• If you have a good questionnaire but expect results that need explaining, use the explanatory design, and recruit the follow-up interviewees from survey respondents who agreed to be recontacted.
• If timing is tight and the topic is reasonably well understood, the convergent design saves calendar time, at the cost of losing the chance for one stage to shape the other.

Linking the two samples

In an explanatory design, the follow-up interviews are most useful when they come from the survey itself. Add a recontact question with clear consent wording, store the respondent's survey answers against their ID, and choose interviewees by the answers you want explained. A respondent who rated a product poorly but still recommends it is often more revealing than one whose answers are consistent. Keep personal contact details separate from the survey data, and share them with moderators only as far as the consent allows.

Turning qualitative findings into a scripted survey

The hand-off from qualitative to quantitative is where mixed methods projects most often lose value. A few habits help:

1. Keep respondents' own words. Use the phrases people actually used as answer options and statement wording, not the team's internal jargon.
2. Convert themes into measurable items. Each theme from the interviews should become one or more closed questions, so you can see how widespread it is.
3. Keep an "other, please specify". Qualitative work rarely captures every possible answer. An open option protects the survey against gaps.
4. Pilot with real respondents. A short soft launch shows whether the new questions are understood the way you intended.
5. Write the routing down before scripting. Conditions, quotas and rotations should be agreed in the questionnaire document so the scripter is not guessing.

For the scripting side of this, our guides to survey programming logic and the questionnaire to live survey checklist go step by step.

Quality standards apply to both

It is tempting to think of quality controls as a quantitative concern. In practice, qualitative work needs them too: accurate recruitment, recorded consent, consistent moderation and clear records of who took part. Service standards such as ISO 20252, which sets out vocabulary and service requirements for market, opinion and social research, are written to cover research providers across methods, not just survey work.

Where Sampledge fits

Mixed-methods work puts pressure on the instrument. SampCode is Sampledge's survey scripting and programming product, where the questionnaire becomes a live survey with its routing, quotas and question formats. Once it is in field, SampTrek tracks quotas, suppliers and live fieldwork through to reconciliation, so the quantitative stage stays connected to the rest of the project. For a broader view of the method choices, see our overview of data collection methods in research and our guide to primary data collection techniques.

If you are planning a study that combines both, ask us for a demo using your own questionnaire.