Analysing qualitative data

This page explores the process of qualitative data analysis. Before doing any analysis it is important to be clear about the purpose of the analysis: what are you trying to do, to answer?

Note that qualitative research is generally exploratory. It could, but rarely is, explanatory. In general explanatory research tends to be quantitative as you need large representative samples to be able to generalise any findings. Qualitative research could also be confirmatory, that is testing predetermined ideas. It is unlikely to be comparative research, as comparative research involves comparing, for example two groups of participants. Comparative research requires absolute consistency and tends to be quantitative.

Analysing data involves:

  • Taking apart and reassemble for a purpose, according to some consistent criteria
  • Sorting, organising, reducing, describing
  • Making sense, drawing conclusions, explaining

How this is done depends on the research question – what are you trying to find out, understand – but also on

  • Your research paradigm. Research paradigms and data analysis influences our belief on how we make sense as it indicates what can be made sense of. E.g. Are we looking at establishing cause and effect relations between variables that can be isolated from their background? Or trying to untangle how meanings have been constructed culturally or socially? For more read section theoretical paradigms underpinning research
  • The chosen research design(s) E.g. Narrative research is looking for narratives (on events, on experience), phenomenology for lived experience. See section on research design for more information on designs.
  • The methods chosen for collecting data

The following table summarises how a research design suggests the use of certain data and methods for analysis:

Major research design frameworks overview

Research design Description Possible sources of data Data analysis
Action research Pursues action/change Interviews, observations, documents, images Driven by need to improve
Case study Account & analysis of one or more cases Interviews, observations, documents, images Thematic and cross case analysis
Discourse analysis Analyses how discourses constitute the social world Written or spoken text, images Analysis of text in context to see how understanding is generated
Ethnography Explores culture of a group of people Observation Search for cultural themes
Generic qualitative inquiry  Explores themes and patterns Interviews, observations, documents, images Search and code themes, concepts and context
Grounded theory Development of theory ‘grounded’ in the data Interviews – observations Open coding, axial coding, selective coding to examine concepts
Narrative research Focus on how individual narrate an experience or an event. Interviews, images Search for narratives across data
Phenomenology Focus on how individuals perceive and live a phenomenon In depth interviews Search for descriptions of a phenomenon and determine their meaning

 Types of qualitative data

There are three broad types of data that should be treated separately:

  1. Collected data: naturally occurring (e.g. observation) or elicited (e.g. interviews). How do data collection methods affect data?
  2. Notes or memos on the context, the environment or the process. This may include reflections and summaries. How much you record about the context depend on your topic.
  3. Notes on codes.

You’ll need to think about how much and what needs to be recorded about the context and environment (participants, body language, surroundings). This will depend on your topic and approach. How much extra information/material (news, documents) do you need to bring?

Analysing the data

This is where we make sense out of data.


Meaning in qualitative research 

Qualitative research sees meaning as mediated and always changing, reconstructed in each reading.Therefore the researcher must make clear his/her own perspective about meaning. Reflectivity, as thinking back about practices and bringing these in the open, and reflexivity, as thinking from within oneself, seeing oneself in the experience, are crucial.

  • Meaning resides in the intention of the actor/speaker – requires getting at intentions
  • Meaning resides with the researcher who interprets the data. 
  • Meaning resides with the reader who interprets the interpretation of the researcher.
  • Meaning resides in the wider system of actions (social rules etc. ) – requires an interpretation into the wider system.

Analysis process

Qualitative data generally comes in the form of text, images, notes, sounds…. These are generally categorised into coding frameworks developed

  • a-priori (generated from theory, literature) or
  • a-posteriori (from data gathered)

Note that qualitative information can be converted to numbers or frequencies. This is done for example in content analysis

Before moving on to coding, you may need to decide whether you code data sets separately or together – depends on research question.

Analysing generally goes through a process of coding. Codes should be related to the research question(s).

  • Coding refers to labelling units of text as ‘themes’ or categories.
  • Coding is often done several times.
  • Once units have been labelled, coding requires identifying units of text which are similar or related in meaning, across different data sets.
  • Themes or categories may lead to patterns, which are descriptive findings. Patterns can then be compared, or mapped out to generate an explanation.

The whole process involves a series of stages which are not linear; you may go back to another reading or collect more data. In qualitative research data is often collected while being analysed.

This process involves series of stages (which in qualitative research do not neatly follow):

  1. Preparing data for analysis – depends on research design and question. What are you looking for? Phenomena, themes, patterns, relationships, narratives?
  2. Initial reading – looking at what is in data, eventually identifying themes but remaining descriptive.
  3. Connecting themes, finding links, repetitions, patterns between what was identified at stage 2.
  4. Reflecting on how and why of the links, relations.
  5. If relevant, theorising or developing theory. Most qualitative research doe snot aim to develop theory. Developing theory requires a solid process of looking at data, making tentative conclusions, going back to data,  making tentative conclusions, going back to data….
  6. Articulating findings back into the context of research – how does it fit with existing research?

Preparing the data

At this stage you need to develop a system for setting out data, information on data and information on context. Think about:

  • The nature of your data do you have? How has it been transcribed?
  • Do you need to translate data? Translation is a ‘black box’ in research (Chidlow, Plakoyiannaki & Welch, 2014). Translating text involves much more than the identification of replacement vocabulary. It is a process of intercultural interaction and interpretation. Read also Lapadat (2000) and Wong & Poon (2010) for more about translating research data.
  • How will you analyse it? You will need to develop a system for coding. E.g. on paper in columns next to the data, using excel, using NVivo, Dedoose or a different program.

Initial/first reading of the data

A first reading involves looking at the text without yet relating it to context or other texts. At this stage meaning is sought in:

  • Word choices, expressions
  • Structure, narrative, sequence
  • Possible causes, arguments developed
  • Transitions from one topic to another
  • Contrasts and similarities
  • Metaphors, analogies

First or open coding

During the first reading, first codes are created.  Coding can be done for words, sentence fragments, sentences, blocks of texts. This coding involves the labeling of units of text according to themes or categories. NVivo uses the term ‘nodes’.  The first reading is a first step in the process of breaking down and reconceptualising data.

Coding approaches

Codes should be related to the research question(s) and can be developed in various ways:

  • A priori and content specific coding: codes developed from the topic and the theoretical perspective. Coding categories can be developed as you collect the data (Minichiello, 1995).
  • A priori and non-content specific codes: based on commonsense reasoning, on general categories such as who, what and where.
  • A posteriori: codes developed when going through the data. At a first stage it may be good to code in Vivo, which refers to the use of expressions as they are in the data.

Second, third, fourth….coding

Also referred to as axial coding. Codes are now specified more rigorously and data reassembled by:

  • Re-examining initial coding.
  • Identifying repetitions, uniformities and merge codes.
  • If relevant creating categories and sub categories, signalling hierarchical relationships.
  • Looking for relationships, developments.
  • Assigning meaning at various levels (word, groups of words, sentences).

Selective coding

The last stage sometimes referred to as ‘selective coding’ refers to bringing together all categories of codes around a ‘core’ category (Strauss and Corbin 1990, cited in Rice & Ezzy 1999). This is the result of another look at codes across data sets to:

  • Compare them to look for similarities and differences
  • Compare them with other ideas, premises and assumptions

Data will need to be transformed and generate a new text resulting from:

  • Reconceptualisation
  • Theorisation (insights, concepts, propositions, models)
  • Relations established with existing research

The coding process is not linear. The following example on coding in grounded theory (Charmaz (1991, p. 276) shows this:

  1. Begin by ‘exploring the general research questions’.
  2. Gather data and code for respondents’ meanings, feelings and actions.
  3. Look for processes and relationships between specific events and general processes.
  4. Coding leads to new categories.
  5. Collect more data on the developing categories.
  6. Go back and read earlier data for the new categories and to formulate new questions.
  7. Constantly compare individuals, different events and the categories.
  8. Write memos all the time about categories, processes and ideas.
  9. Move towards memos that are more conceptual and codes that are more abstract.

When should you stop coding?

When nothing new comes up.

watch video icon Graham R Gibbs discusses coding in the following series of videos:

Grounded Theory – Open Coding Part 2



Grounded Theory – Axial coding

Grounded Theory – Line by line coding

Grounded Theory – selective coding


An example of how to perform open coding, axial coding and selective coding

Another useful online resource : How and what to code


Concluding: making meaning

Good analysis goes beyond what people say. Ricoeur (1976) refers to this last stage as the appropriation stage – where new understandings are developed from the data. Codes, as categories of meanings, and concepts within them are compared and contrasted and, where appropriate, linked and examined taking into account the context in which the experiences described in the text took place.

You re-organise all the different codes data to find some new order, logic….you look at all your codes, examine relationships, see what is connected with what, how.

Context is important at this stage as it supports your interpretation.

Note that you are inferring through induction (inferring to observation) or what is referred to as abduction or inference to best explanation of observation. Generally you are not deducing which refers to inferring that something is true if the premises from which it is inferred are true.

You may need to rearrange your coded data various times before coming up with the new text, the explanatory text.

You are expecting to transform the data and generate a new text resulting from

  • Re-conceptualisation
  • Theorisation (insights, concepts, propositions, models)
  • Relations established with existing research

The aim of this is to develop a new understanding, a new meaning which is no longer the private meaning, as experienced by participants, but a public meaning as revealed by the text (Ricoeur, 1976). Whether you go back to participants or not at this stage is up to you, or to your research approach.

Any claims you make have to be the result of a clearly outlined process and have to be based on the data you used.

Potential issues

To ensure solid and good research

  • Be aware of the influence of the researcher on the meaning making at all times. Reflexivity is crucial. Read Le Gallais (2008)
  • Make explicit the relationship between researcher, participant and research data
  • Account for any discrepancies
  • Have a consistent approach to meaning and coding
  • Do not treat codes as fixed and mechanical categories, they have to come from the data
  • Make sure context is taken into account when and as required
  • Take into account that elicited and naturally occurring data are not exactly the same

Reporting qualitative data

Requires a consistent account of the research, the research approach, methods and research purpose/questions. Check guidelines proposed in Pratt 2009.


Charmaz, K. (1991). Good Days, Bad Days: The Self in Chronic Illness and time, Rutgers University Press, New Brunswick.

Chenail, R.J. (1995). Presenting qualitative data. The Qualitative Report, 2/3, 1-9. Retrieved from

Chidlow A., Plakoyiannaki E. and Welch C. (2014). ‘Translation in cross-language international
business research: Beyond equivalence’. Journal of International Business Studies 45, 562–582.

Dey, I. (1993), Qualitative Data Analysis, Routledge & Kegan, London.

Lacey, A. & Luff, D.(2009). Qualitative data analysis, Resource Pack produced by The NIHR RDS for the East Midlands / The NIHR RDS for Yorkshire and the Humber

Lapadat, Judith C. (2000). ‘Problematizing transcription: purpose, paradigm and quality’. International Journal of Social Research Methodology 3/3, 203-219.

Le Gallais, Tricia (2008). ‘Wherever I go there I am: reflections on reflexivity and the research stance’, Reflective Practice: International and Multidisciplinary Perspectives, 9/2, 145-155.

Minichiello, V. (1995). In-depth interviewing: Principles, techniques, analysis (2nd ed). Longman, Melbourne.

Pratt M. G. (2009). ‘For the lack of a boilerplate: tips on writing up (and reviewing) qualitative research’, Academy of Management Journal 52/5, 8.

Rice, P.L. & Ezzy, D. (1999). Qualitative data analysis. Qualitative research methods: A health focus. Oxford University Press.

Ricoeur, P. (1976). Interpretation Theory: Discourse And The Surplus Of Meaning. The Texas Christian University Press: Forth Worth.

Ritchie, J. (2001). ‘Not everything can be reduced to numbers’, pp. 149-173 in Health research. New York: Oxford University Press.

Silverman, D. (2006). Interpreting Qualitative Data: Methods for Analysing Talk, Text and Interaction (3rd ed). Sage, London.

Strauss, A. & Corbin, J.(1990). Basics of Qualitative Research, Sage, Newbury Park.

Wong, J.P.H., & Poon, M.K.L. (2010). ‘Bringing translation out of the shadows: Translation as an issue of methodological significance in cross-cultural qualitative research’. Journal of Transcultural Nursing, 21/2, 151-158. Doi: 10.1177/1043659609357637.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s