Quantitative vs. Qualitative research: What is the difference?

Research is important. Humans have been using it for ages to improve the success of their decisions. Early research was more trial and error, rather than a structured approach, as in this tweet:

Source: Twitter

Nowadays, entire scientific fields are devoted to applying highly structured forms of research to build knowledge. A core element of research is collecting data, which is then analysed to form insights. The collection and analysis of data can be separated into quantitative research and qualitative research.

At a most basic level, quantitative research (or quants) deals with numerical data or statistics. It’s commonly used in studying the natural world and assumes a fixed reality (i.e. what were the average temperatures over the last 100 years).

By contrast, qualitative research is concerned with textual data, used in a descriptive way. It’s commonly used in studying the social world and assumes that reality can be negotiated and is dynamic, based on the perspectives of human subjects (i.e. the participant said that since it’s cold where they are now, global warming is a hoax).

Image source: Pexels

When we plan, design, and build products, it’s important to utilise both qualitative and quantitative research methods. Each has its benefits and drawbacks, and each tells us a different story about our users. The Nielsen Norman Group uses a great example: A survey may tell us that using websites is 122% more difficult for senior citizens than mainstream users. This type of qualitative research gives us an insight that’s easy to grasp and discuss. However, only in-depth and qualitative research will tell us why, and what we can do to address the problem.

The trick then becomes when to use what type of research. Below are some brief examples of the tools I like to use, as well as the benefits and limitations of each type of research method:

Quantitative research tools:

• Surveys with yes/no questions or Likert rating scales (According to mentimeter, Likert rating scales are a type of rating scale that is usually found on survey forms or questionnaires, it measures how people feel about something.)
• Remote usability tests that measure things such as misclick rates, task completion times or task success rates.
• Web analytics that tracks user actions.

Image source: Likert scale

Benefits of quantitative research:

• Data can often be collected quickly.
• Remote collection software can be used.
• Data is generally objective.
• Results can be clearly communicated through numbers and statistics.

The image above is an example of click tracking on Useberry.

Limitations of quantitative research:

• Data is seen as restrictive therefore, you cannot ascertain context.
• Large sample sizes are needed.
• Results can be oversimplified.

Qualitative research tools:

• Surveys with open-ended questions.
• Interviews and focus groups.
• Face-to-face usability tasks with feedback sessions.

The image above is an example of open-ended questions on a survey.

Benefits of qualitative research:

• Results in a better grasp of individual human behaviour.
• A deeper understanding of why and how people respond to phenomena.

Limitations of qualitative research:

• Often subjective in their responses and analysis.
• Personal biases can influence the interpretation of results.
• Extracting themes and actionable insights is more difficult, especially for inexperienced researchers.
• Cost and effort are more, necessitating smaller sample sizes.

Ultimately, when it comes to conducting useful and actionable user research, I find it best to combine both quantitative and qualitative, employing a type of mixed-methods research. Mining secondary data from large qualitative studies (such as Stats SA) is great to understand demographic forces and establish a target market for the product or feature. Interviews or focus groups with small groups of customers can then narrow the data down into a deeper understanding of consumer behaviour. Insights and assumptions from this qualitative research can then be verified, for instance, by sending out a qualitative survey to a larger sample of the population, or any type of relevant subsequent test.

Taking this type of approach will result in findings that are as objective, reliable, and valid as possible. And flexibility, adaptability and iteration will ensure that it stays that way.