Skip to content. Skip to navigation

Faculty Development

You are here: Home / Introduction to Educational Research / Quantitative and qualitative research

Quantitative and qualitative research

There is a wealth of information and continuing debate about the nature of research and knowledge and the assumptions that underpin these. The ‘Further Reading’ section includes some accessible articles that describe these approaches and discussions in more depth. In educational research, a current trend is towards the use of mixed methods and here we briefly describe the main components.

Quantitative research relies primarily on numbers as the main unit of analysis. It is not just about numbers though, although research that aims to ask ‘what?’, ‘how much?’ and ‘why (cause)?’ tends to generate data in terms of numbers. It is more commonly used as a primary method in scientific and clinical research, such as drug trials or laboratory experiments where tests may need to be repeated many times, for example to ensure that a new drug is safe. Although quantitative methods, such as surveys, are used a lot in educational research, the vast majority of research is relatively small scale, intensive, focused on change and involves human perceptions. Educational research relies much more heavily on qualitative methods.

One of the most common instruments to gather numerical data in education (particularly in evaluation of programmes) is the questionnaire survey, using a series of closed questions to which responses are given against a Likert or other type of scale. Open questions can also be included to gather richer data. Large amounts of data can be gathered from a wide number of people and the results can be analysed by computer (either by an optical mark reader or through an online survey instrument such as the ‘Bristol Online Survey’ or ‘Survey Monkey’), thus making it fairly straightforward to research a large sample of respondents. Survey questionnaires can be given out and collected face-to-face, sent by post or posted online. If achieving a high response rate is important, then note that the less personal involvement there is with potential respondents, the lower the response rate. So, typically, online surveys may have a response rate of under 20%, whereas if the questionnaires are given out and collected face-to-face, you may achieve a very high response rate.

Qualitative research asks the questions ‘why (explanation)?’ and ‘how?’ and thus tends to generate rich data composed primarily of words as its unit of analysis and its means of understanding. However, it can also use voice tone, loudness, cries, sighs, laughs, and many other forms of human communication. The words may be spoken in individual interviews (face-to-face or on the telephone) or groups, or they may be written, so you may have to analyse the spoken words of an interview, focus group or conversations (for example between a patient and health worker), or the written words of an account or description or diary record. In-depth accounts can also arise from observing behaviours and examining artefacts and other documentation.

On the whole, qualitative research tends to be small-scale, simply because it is hugely labour intensive and often takes a case study approach. Interviews or focus groups will usually need to be transcribed before they can be analysed. In addition, the researcher is often more involved with the person producing the words, and so it is sometimes helpful for others to conduct the analysis; again this can be costly. Having said that, nothing else can provide the same level of richness as qualitative data, and at the very least, adding space for respondents to provide some words to describe what might be otherwise gathered by numbers is immensely useful to the researcher, and may even, in some situations, help the participant.

Qualitative methods range from the classification of themes and interconnections to content analysis, grounded theory and discourse analysis; and reliability and validity are just as important as they are in quantitative analyses. Computer programs can assist in analysis, and although these might not necessarily save time, they often offer more systematic ways of coding data and identifying connections and themes.

Triangulation describes bringing a number of research methods to bear upon a question. For example, to study the effect of threatened closure of a hospital on its staff and the local university that delivers its healthcare programmes there, you might want to:

  • interview a selected few individuals at different levels, in different professions, from stakeholder organisations;
  • review programme evaluation data from the last three years in relation to this hospital;
  • carry out a questionnaire survey of all those involved;
  • analyse personnel data of sickness, absence, turnover, etc.;
  • carry out a network analysis of rumour;
  • undertake an ethnographic study of a long-stay ward looking at staff and patient interactions, etc.

 

A triangulation of methods such as these would provide an exceptional in-depth look at such an event, taking a case study approach, but depending on time and resources simply combining some form of qualitative and quantitative data in a questionnaire survey would be useful. When adopting mixed methods, it is crucial that the sequence of methods and the ways in which the data will be integrated is explicit and justified.

Print module to PDF

Save a PDF of this module, so you can print it and read it in your own time.

Email your comments

Let us know what you think about this module or give us your feedback.

Further information

More information about this module, further reading and a complete list of glossary terms.

Learning activities

Read about the recommended learning activities for this module.