Survey research can be qualitative, quantitative, or both. It depends entirely on how the survey is designed: the types of questions asked, the format of responses collected, and how the data gets analyzed. Most surveys lean quantitative, using numbered scales and preset answer choices, but open-ended questions produce qualitative data. Many researchers combine both approaches in a single survey instrument.
What Makes a Survey Quantitative
A survey becomes quantitative when it uses closed-ended questions that generate structured, numerical data you can count and analyze statistically. This is the most common form of survey research. The hallmark is predetermined response options: yes/no questions, multiple-choice items, rating scales, and ranking questions. These produce data points that can be charted, compared across groups, and tested for statistical significance.
Likert scales are probably the most familiar example. Developed in 1932, a typical Likert scale asks respondents to rate their agreement with a statement across 5 or 7 points, from “strongly disagree” to “strongly agree.” A researcher studying patient satisfaction, for instance, might use 10-point rating scales to measure confidence, distress, and trust, then calculate averages and run statistical comparisons between groups.
Other common quantitative question types include dichotomous questions (yes/no, true/false), matrix grids that apply the same scale across multiple items, and specialized formats like Net Promoter Score (a 0 to 10 scale). All of these produce data you can summarize with percentages, means, or medians and analyze with statistical tests.
What Makes a Survey Qualitative
A survey becomes qualitative when it includes open-ended questions where respondents answer in their own words. Instead of picking from a list, people write short answers, explanations, or longer narratives. The resulting data isn’t numerical. It’s text that captures experiences, opinions, and perspectives in ways that preset categories can’t.
Analyzing qualitative survey data works differently from crunching numbers. Researchers read through responses and identify recurring themes and patterns, a process called thematic analysis or coding. They look for the point at which new responses stop producing new ideas, sometimes called thematic saturation. This kind of analysis reveals the “why” behind people’s answers, the reasoning and context that numbers alone miss. Open-ended responses can range from single-word lists to detailed paragraphs, and each requires a slightly different analytical approach.
How Mixed Methods Surveys Combine Both
Many well-designed surveys use both approaches in the same instrument. A questionnaire might include closed-ended items scored on a rating scale alongside open-ended follow-up questions that ask respondents to explain their answers. This is called mixed methods survey research.
Researchers combine these approaches in several ways. Some start with qualitative open-ended questions to explore a topic, then use those findings to build a structured quantitative survey. Others use a quantitative survey first, then follow up with open-ended questions to dig deeper into surprising results. The most integrated approach puts both question types side by side in one survey, collecting numerical ratings and written explanations at the same time. For example, a satisfaction survey might ask you to rate your experience on a 5-point scale and then ask “What could we improve?” in an open text box right below.
How the Data Type Shapes the Analysis
The distinction between qualitative and quantitative isn’t just academic. It determines what you can do with the results.
Quantitative survey data gets analyzed with descriptive statistics (means, medians, standard deviations, frequency distributions) and inferential statistics (t-tests, chi-square tests, regression analysis). These tools let researchers measure central tendencies, test whether differences between groups are statistically meaningful, and estimate relationships between variables. For scaled questions, analysts often report “top-box” scores, the percentage of respondents who chose the most positive option, alongside distributions showing how answers spread across the scale.
One nuance worth knowing: Likert scale data is technically ordinal, meaning the gaps between “strongly agree” and “agree” aren’t necessarily equal to the gaps between “agree” and “neutral.” Strictly speaking, this means averages aren’t perfectly appropriate, and some researchers prefer reporting medians and percentages instead. In practice, though, the expert consensus is that standard statistical tests work fine with Likert data as long as sample sizes are adequate (at least 5 to 10 observations per group) and the data is roughly normally distributed.
Qualitative survey data, by contrast, gets coded into themes and categories. Researchers look for patterns, count how often certain ideas appear, and sometimes use specialized software to organize and sort free-text responses. The output is descriptive rather than statistical: summaries of what people said, organized by theme, with representative quotes illustrating each finding.
Common Biases That Affect Both Types
Regardless of whether a survey is qualitative or quantitative, certain biases can distort the results. Recall bias occurs when respondents’ memories are colored by their outcomes. Someone who had a bad experience with a product or treatment may remember earlier events more negatively than they actually were. Social desirability bias pushes people to answer in ways they think are more acceptable rather than truthfully, which affects both scaled ratings and open-ended responses.
Non-response bias is another concern. When certain types of people consistently skip the survey or drop out partway through, the remaining data no longer represents the full population. Some researchers try to counter these effects by masking the intent of questions, using validated scales that have been tested for reliability, or designing surveys that minimize the pressure to give “correct” answers.
Which Approach Is Right for Your Purpose
If you need to measure how many, how much, or how often, a quantitative survey with closed-ended questions is the right tool. It gives you numbers you can compare, track over time, and generalize to a larger population. If you need to understand why people feel a certain way or explore a topic you don’t yet know enough about to write good multiple-choice options, qualitative open-ended questions are more useful. If you want both breadth and depth, a mixed methods survey that pairs rating scales with open text fields gives you the most complete picture.
The survey itself is neither inherently qualitative nor quantitative. It’s a data collection method, and the research design you wrap around it determines which category it falls into.