What is qualitative research?
Qualitative research is all about language. That covers words, meanings and understanding. It’s used to describe why people feel the way they do, why they act in a certain way, what opinions they have and what motivates them. Qualitative research gives breadth, depth and context to questions, although its linguistic subtleties mean that results are trickier to analyse than quantitative data.
What is quantitative research?
Quantitative research is all about numbers. It gathers information that can be counted, measured, or rated numerically. It’s easy to ‘crunch the numbers’ of quantitative data and produce results visually in graphs, tables and on data analysis dashboards.
Differences between qualitative and quantitative research
Qualitative | Quantitative |
---|---|
Gathered from focus groups, interviews, case studies, expert opinion, observation | Gathered from surveys, questionnaires, polls |
Uses open-ended and open text questions | Uses closed-ended (yes/no) and multiple choice questions |
Uses a ‘human touch’ to uncover and explore an issue (e.g. a customer complaint) | Cannot use a ‘human touch’ to interpret what people are thinking or feeling |
Helps formulate a theory to be researched | Tests and confirms a formulated theory |
Results are categorised, summarised and interpreted linguistically | Results are analysed mathematically and statistically |
Results expressed as text | Results expressed as numbers, tables and graphs |
Fewer respondents needed | Many respondents needed |
Less suitable for scientific research | More suitable for scientific research as it is compatible with most standard statistical analysis methods |
Harder to replicate | Easy to replicate |
Less suitable for sensitive data: respondents may be biased, too familiar, or inclined to leak information | Ideal for sensitive data as it can be anonymised and secured |
Qualitative, quantitative or combined? Choose your research methods
Here’s how to decide which method to use:
- Qualitative research: use this to understand something – experience, problems, ideas. For example, you choose 100 supermarket loyalty card holders and survey them, asking open text questions, e.g. “How could we improve our store?” or “Were you able to find everything you came in for?” This research will pinpoint problems (a lack of trolleys, dirty toilets, poor stock control) that quantitative research will not.
- Quantitative research: use this to test or confirm a theory or hypothesis. For example, you survey 400 loyalty card holders, asking them, “On a scale of 1-5, how happy are you with our store?” You analyse the numerical responses and conclude that the store has scored 4.5.
- Combined method: Use qualitative research to gain insights and propose a theory, then quantitative research to test it. Your surveys can include both multiple choice/closed questions and open text. For example, market research interviews with supermarket focus groups find that customers would like to be able to buy children’s clothes in store. The supermarket pilots a children’s clothing range. Targeted quantitative research reveals that those stores selling children’s clothes achieve higher customer satisfaction scores and a rise in profits for clothing.
eBook: A guide to building agile research functions in-house
Download NowDifferent survey question types
Quantitative data
You have various options for question types. The usual ones are:
Net Promoter Score (NPS)
Likert Scale
Radio buttons (respondents choose just one option)
Check boxes (respondents can choose multiple options)
Drop down
Sliding scale
Star rating
Qualitative data
There are fewer survey question options for collecting qualitative data. But with artificial intelligence programs that analyse open text, and turn qualitative data into quantitative for real-time statistical analysis, they are equally valuable:
Open text ‘Other’ box (can be used with multiple choice questions)
Text box (space for short written answer)
Essay box (space for longer, more detailed written answer)
Analysing survey data
Your survey data is in, now you need to interpret the results. Here’s what to do:
- First, clean your data: you need to sort valueless data such as incomplete surveys, disengaged or inconsistent respondents, and bots.
- Stick to your basic research questions: select the results that answer those questions.
- Make sure your data is representative: and large enough to give an accurate picture of your research sample.
- Cross-tabulate results to create individual tables for specific survey questions.
- Consider filtering out certain respondents: you may wish to exclude some demographics, or locations.
- Work out averages: the mean (average) number of respondents, mode (most common response) and median (the mid-range response), and run statistical analysis.
- Benchmark against previous results: you’ll be able to see improvements, changes and trends emerging.
eBook: A guide to building agile research functions in-house