from description to analysis – a revision strategy


PhDers are often told by their supervisors that their work needs to move from description to analysis. But what does this mean?  Have you just wasted your time doing all that describing? Well, in short, no.

The good news is that most analysis, whether it is of numbers, words or images, starts with description. You have to write long, in words, what you think you actually have in hand and what you now know.

Description is usually concerned with what some people call “facts and figures” – this is possibly not the most helpful term. Let me try another approach. Description is usually concerned with your data and particular information you’ve selected – information that’s necessary to develop an answer to your research question.

And the description of your selected data/information could be about:

  • what happened, when and to who/what;
  • what something looks or feels like;
  • a series of events;
  • an explanation about how something was or might be done;
  • a summary of a text;
  • a list of something (components, options, methods, theories etc)
  • a timeline;
  • a set of themes or key words;
  • a sketch of important details…

Description of these kinds of data allows you to become clear about what ‘stuff’  you have.  It’s a necessary, but not sufficient, step.  Now you just need to take another one in the process of making sense. Analysis moves on and away from your description.

Here’s a tip that might help you to do that. Imagine your analytic self taking a step back, metaphorically putting on an evaluative hat, and looking again at the description you’ve written to see what can be made of it. Your analytic self picks out the larger shapes in the landscape, moves beyond the detail. ( As per the image at the top of this post.)

In order to find bigger patterns, your analytic self approaches your description with a question – or two or more – in mind.

And, where do these questions come from? Well, you may need to go back to your research aims and objectives, or to your reading, in order to help you work out what you want to know from your description. But I can offer a bit of help.

Common analytic questions include:

  • Who is this person? What makes them the way they are? How have they got to be life this? Why do they do what they do?
  • How does this text work? What is communicated through word, number and image?
  • What claims are made in this text?
  • What reasons and evidence are provided in this text? Are they plausible? Trustworthy? Current? Decontextualised? (etc)
  • What are the key elements of this process? (without these it won’t work)
  • What sequence of events lead to this outcome?
  • What caused this to happen? What was most important and what was of secondary importance?
  • What larger patterns are in this description of data?
  • What is most significant here?
  • How do things (in the description) compare with each other? What seems to be most important about the differences/commonalities?
  • What are the strengths and weaknesses in this/of this?
  • What does this add up to?

You can see that these questions go to bigger and more abstract ideas – core qualities, principles, underpinning processes, causality, key characteristics, evaluation. It’s these larger abstracted patterns that we produce through analysis.

Now, I’ve made that sound simple. But of course it’s not.

Description often feels like a pretty low risk activity. You’re summarising what’s already there. But analysis is jumping into the unknown. No-one has analysed precisely the same set of descriptive data as you. And it’s down to you to make sense of it. Yes, it’s all yours to sort out from here on in.

Asking analytic questions means that your analytic self has to be brave, let go of the certainty of description and make your own judgments. Obviously you can be guided in those judgments by your reading, and discussion with your supervisor. And if you work in a team, then your team is likely to do some analysis together and check out each other’s interpretations.

But ultimately it is you – your analytic self – who moves away from the safety of the descriptive known and moves into a new knowledge territory. You have to decide what you can draw out of your description.

And then it’s revision.

You’ve got your first draft. It took you ages and it’s great that you’ve done it. But you know its not enough. You now want to find any troublesome descriptive places before your reader/supervisor.

Your revising self needs to look for all the spots where you have described something: what happened when and to who/what; what something looks or feels like; a series of events; an explanation about how something was or might be done; a summary of a text; a list of something ( components, options, methods, theories etc); a timeline; a set of themes or key words; a set of important details… Ah, deja vu. it’s the same list as before. But there’s a point to repeating myself .

Ask yourself now, when you see one of these descriptions in your draft, whether it’s OK or not. Is this a place where some further analysis would be helpful.

Ask yourself – Do I need to do more work on this before I go on to offer an explanation? Is this description sufficient for me to answer my research question fully? Are there big patterns here? Principles? Causes? Qualities? Processes? Evaluative judgments? What can I do to this description to get to the point where I can make claims about my results? Should I leave my description in and add analysis, as in ethnography, or change the description to analysis – what would help the reader most?

And if the answer is I can do more, I need to do more? Well, when the time is right, switch from revising self into analytic self and move on….

Photo by Droneflyer Nick on Unsplash

About pat thomson

Pat Thomson is Professor of Education in the School of Education, The University of Nottingham, UK
This entry was posted in academic writing, analysis, crappy first draft, data analysis, description, empirical analysis, revision, revision strategy and tagged , , , , , . Bookmark the permalink.

5 Responses to from description to analysis – a revision strategy

  1. Hunt Family says:

    This is exactly the information I needed today after telling myself last night that my draft wasn’t argumentative enough. Thank you for your timely wisdom! Kind regards Emily Hunt



  2. Pingback: DoctorateHub Blog

  3. Pingback: DoctorateHub Blog

  4. Jean says:

    Hello Pat

    I found your book ‘Helping Doctoral Students Write’ very helpful when I began writing up my doctoral thesis.

    My Research methodology was Action Research employing Collaborative Enquiry. Whilst I have concluded the research and the transcription of the interviews, I am struggling with the writing up and specifically with how to report the data. Can you recommend any writing or finished projects that I may be able to access? If you have written anything on writing up this research methodology I’d be happy to purchase a copy.

    Any guidance you can offer would be much appreciated



    • pat thomson says:

      This is a gap in the literature on this and Helen Kara and I are commissioning a book in our Routledge series about it. Generally though you could think about this kind of structure: introducing the problem, explaining your approach including data generation and analysis , cycle one reconnaissance action and reflection, repeat if a second cycle, discussion and conclusion. Make sure you get an examiner who understands action research and doesn’t expect to see it presented as IMRAD.


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 )

Connecting to %s