Quantitative vs qualitative research

Quantitative research collects numerical data to measure variables and test hypotheses, while qualitative research collects non-numerical data, words, observations, and meanings, to explore how and why something happens. The simplest way to tell them apart is to ask what the data look like: counts, scores, and measurements point to quantitative work, whereas interview transcripts, field notes, and open text point to qualitative work. Most dissertations lean on one, and a growing number combine both in a mixed-methods design.

How to tell whether a study is qualitative or quantitative

Look at three things: the data, the question, and the analysis. If the study counts or measures, asks "how much" or "how often", and analyses with statistics, it is quantitative. If it gathers words or images, asks "how" or "why", and analyses by coding for themes, it is qualitative. The analytical machinery follows directly from this split, which is why the path from descriptive to inferential statistics belongs to the quantitative side, while coding methods like thematic analysis belong to the qualitative side.

The four types of quantitative data

Quantitative data come in four measurement levels that decide which analysis is legitimate: nominal (unordered categories like blood type), ordinal (ordered categories like a satisfaction rank), interval (equal spacing but no true zero, like temperature in Celsius), and ratio (equal spacing with a true zero, like weight). The level matters because it governs whether you can compute a mean, run a parametric test, or are limited to ranks, a distinction explored in the parametric or nonparametric choice.

Quantitative vs qualitative data, with examples

The research label follows from the data themselves, so it helps to picture each kind. Quantitative data are anything you can count or measure: test scores, reaction times, sales figures, the number of patients readmitted, or a temperature reading. They are summarised with means, percentages, and statistical tests. Qualitative data are anything expressed in words or images: an interview transcript, an open-ended survey comment, a field note, or a photograph. They are summarised by coding for themes rather than by arithmetic. A useful grey area is the Likert scale, where a worded option such as "agree" is given a number so it can be analysed quantitatively, a hybrid handled in analysing Likert-scale survey data. Once you know which kind of data you hold, the legitimate analyses follow from its measurement level and test assumptions.

Examples of qualitative research

Qualitative work covers a family of approaches: interviews and focus groups that elicit lived experience, ethnography that observes a setting over time, case studies that examine one bounded instance in depth, and document or narrative analysis that interprets existing texts. What unites them is the goal of understanding meaning in context rather than measuring frequency. The data are then handled with the coding-based qualitative data analysis methods rather than statistical tests.

Mixed methods research: combining both approaches

Mixed methods research deliberately combines numbers and words in a single study so each strand answers what the other cannot. A common pattern is the explanatory sequential design, where you run a survey first and then interview a subset to explain the surprising results. The reverse, the exploratory sequential design, starts with interviews to surface ideas and then builds a questionnaire to measure them at scale. A convergent design collects both at once and compares them for agreement. The payoff is breadth plus depth, but the cost is that you must run both kinds of analysis rigorously and then integrate them honestly, rather than treating the qualitative part as decoration. On the numerical side you still apply the full pipeline from descriptive to inferential statistics, while the words are handled with the coding-based qualitative data analysis methods.

Which is more difficult, and which should you choose

Neither approach is inherently harder; they are difficult in different ways. Quantitative research demands careful design, adequate sample size, and correct statistical analysis, where a wrong test invalidates the result. Qualitative research demands disciplined coding, reflexivity, and a defensible audit trail, where rigour is harder to demonstrate precisely because it is not reducible to a p-value. Choose based on your research question: if you want to measure, compare, or generalise, go quantitative; if you want to understand a process or meaning in depth, go qualitative. When the question genuinely needs both, a mixed-methods design pairs the breadth of numbers with the depth of words, and the test-selection guide will help you handle the quantitative half once the design is set.