frameworkResearch & behavior

Qualitative and Quantitative Research

A distinction between evidence used to understand patterns and mechanisms in depth and evidence measured to estimate magnitude, frequency, or difference.

Download Markdown (.md)

01What it means

Qualitative research directly examines experiences, behaviors, language, and context to interpret how or why something happens. Quantitative research uses predefined measurements and numerical analysis to estimate how many, how much, how often, or how strongly. Either can study attitudes or behavior, and combining them requires a reason rather than a ritual.[1][2]

Explanation and measurement Choose evidence by the decision, not by a preferred method
Qualitative
How, why, meaning, and contextInterpret experiences, behavior, and language
Quantitative
How many, how much, how often, how stronglyEstimate predefined numerical measures

Either mode can study attitudes or behavior; combine them only when their evidence is complementary.

02Why it matters

Product teams need both explanation and magnitude. Qualitative evidence can reveal mechanisms and unknowns that metrics miss, while quantitative evidence can test prevalence or change beyond a few observed cases.

03When to use it

Use qualitative approaches to explore unfamiliar situations, diagnose problems, or understand meaning. Use quantitative approaches when constructs can be measured and the decision depends on magnitude or comparison. Sequence or combine them when one mode can frame or explain the other.

04How it works

  1. Define the decision and whether it requires explanation, measurement, or both.
  2. Choose a design, sample, and data collection method appropriate to that need.
  3. Analyse qualitative material transparently or quantify predefined measures with suitable statistical reasoning.
  4. Compare conclusions only at the level each sample and method support.
  5. Integrate complementary evidence while reporting disagreements and uncertainty.

05A practical example

Example

A streaming team uses interviews and session observation to learn why household switching is confusing, then measures task success across a larger sample after redesign. The first study identifies mechanisms; the second estimates improvement.

06Where it breaks

Small qualitative samples do not estimate population prevalence, while numerical precision does not guarantee valid constructs or explain causes. Mixed methods add cost and can produce incompatible evidence that requires judgment.

Common mistakes

  • Calling open-text survey responses qualitative depth without contextual probing.
  • Converting counts from a small purposive sample into population percentages.
  • Assuming quantitative evidence is objective regardless of measurement and sampling choices.

07Sources

  1. When to Use Which User-Experience Research MethodsNielsen Norman Group · Professional guidanceWhen to Use Which User-Experience Research Methods
  2. Quantitative User-Research Methodologies: An OverviewNielsen Norman Group · Professional guidanceQuantitative User-Research Methodologies: An Overview
View saved topics

Start typing to search titles, definitions, categories, and frameworks.