do we ‘collect’ data? or – beware the ontological slip …

A post where I have a small rant about one of my least favorite research expressions…

We’ve all heard of a Freudian slip. This is where we inadvertently say something that unintentionally reveals an unconscious, or repressed, feeling, idea or belief. Of course the mis-saying, the slip of the tongue, might just be because we are tired or tense. Or – as in the case of ‘collecting data’ – it’s just a saying that is in common use.

I never use the term ‘collecting data’ myself. And when I hear other people saying it, I always register it somewhere on my inner Richter scale of research-quakes.Why? Well yes, I might be being ridiculous, but on the other hand, maybe not. Let me explain. My concern is about a slip that might be both ontological and epistemological.

(If you feel a bit uncertain about those two terms, I can really recommend listening to Nick Hopwood’s recent podcast on the topic. Nick uses Elgar’s “Pomp and Circumstance” to introduce the various ontological and epistemological choices that we have to make as researchers. It’s not a very long podcast, but does a great job of laying out the landscape of possibilities.)

The first ontological problem I have is with the word collect.

What do we most often collect? Mushrooms in the forest. Letters from the postbox. The children from school. In each of these cases, we are collecting something that actually exists as an external, material Thing in the world – mushrooms, letters, children. BUT, do those people who habitually talk about ‘collecting data’ actually think that data exists out there in the world in the same way as these hypothetical mushrooms, letters and children?

Now, I’m not saying here that nothing in the world exists. I’m not going all Bart Simpson and tree falling in the forest on you. Let me make this clear. If we are doing research on sheep or on/with children, I don’t want to suggest that these sheep and children aren’t ‘real’. What I do want to suggest is that for all researchers who aren’t looking for universal laws which govern a predetermined external reality – and some of us are and that’s OK and you can stop reading now – these things/people aren’t data until WE make them data. They don’t live out there as data. They don’t hang around like mushrooms waiting to be picked – or not. They aren’t data until we make them data.

And as we construct events, objects, people, phenomena as data, we are in the process of constructing a particular set of social/cultural meanings.

If you take a constructivist ontological/epistemological perspective then you will accept that even numbers in a survey are heavily constructed. We define our terms, we decide what to focus on, what to leave out , what to include, how many, who, where, when and so on… There’s not a bit of a survey that the researcher is not all over, including the choice of statistical approach. The same can be said for interviews. As researchers we decide most of what to talk about with interviewees and we certainly control how to record the interview, how it’s edited and how the conversation is represented in and as text. There’s not a bit of an interview that we haven’t actually constructed. The conversation/data wasn’t out there like a mushroom, a letter or a child, waiting to be ‘collected’.

So if you say that you are a constructivist, and you use the term ‘collecting data’ does this matter? Well maybe not at all… unless you have an examiner/reader who wants to focus on the congruence between the stated ontological epistemological position and the methodology/methods. An examiner like me.

But what if there is more to it than fear of patter-like examiners? What if this IS like a Freudian slip, and is actually seepage of a largely unconscious or repressed positioning? Perhaps this is not an individual but a collective repressed postpositivism in which, even though at one level we say we believe knowledge is constructed, deep down we actually believe otherwise? (Well, I accept that that’s probably not the case. I wouldn’t want to argue that because someone is bound to ask for the data which suggests that is what is happening!! I’m just being blog-provocative of course. Although I do remember some empirical research which did argue this position, the old sociology of knowledge work about repressed positivism.)

Putting the question of Freudian slips aside, I do think for those of us that adhere to a broadly constructivist position, it IS important to think about what that means and to consider how that choice is reflected in our terminology. While saying “I’m a constructivist and I collect data” might be commonplace, it does demonstrate a reflexivity blind spot.

I reckon that if we think the truth is not out there, Scully, lying around like a mushroom, or waiting impatiently like the child at the school gates, then it does mean we need think about how we represent that position in and through language – and particularly through metaphors such as ‘collecting’. It’s not good, as my mother would have said, to go about with your slip showing for all to see. But that’s my position.

Rant over – but do watch me flinch next time I hear/read the words ‘collecting data’.

About pat thomson

Pat Thomson is Professor of Education in the School of Education, The University of Nottingham, UK
This entry was posted in data, epistemology, methodology, ontology and tagged , , , , , . Bookmark the permalink.

20 Responses to do we ‘collect’ data? or – beware the ontological slip …

  1. Hi Pat, as an emerging researcher with a “constructivist ontological/epistemological perspective”, my supervisor and I call it “generating” data. Awkies, but there you go. Better than gathering, collecting or other counting terms, it sits on a par with “findings”, not “results”. All semantics but to my supervisor semantics defines one’s ontological/epistemological perspective, and it’s better to get this right, at least, as near to appropriate as our language will allow!


  2. PS just realised your banner photo is nearly identical to my desk, Apple products and worn wooden desk top and everything. Huh.


  3. Simon Bailey says:

    Whether or not such slips are actually representative of some inner conflict, for those of us who are hung up on such ‘quakes’ their appearance in the work we are reading can really bias our opinion. It can be a particular challenge to reviewing: how to write a fair review for a paper which through one or other of these epistemological itches, drives you completely nuts!


  4. Pingback: do we ‘collect’ data? or - beware the ontological slip ... | PhD Research Beginners |

  5. Rabson Mgawi says:

    Thanks for the idea on use of phrases during research. As a constructivist I see the importance of respect for the respondents. Certainly “collecting data” sounds academic but in a way demeaning. Am sure as we read and write on we will discover a better phrase.


  6. Pingback: do we ‘collect’ data? or - beware the ontological slip ... | Metawriting |

  7. Dan LIU says:

    Actually, personally, I would say when you say as a constructivist, we should not use “collect”, you already have the ontologocal belief that the meaning of the word “collect” is predetermined, not constructed. The matter is not so much about which specific word we use, but about how we are going to interpret or construct that word.


    • pat thomson says:

      Constructivism as a broad category usually includes feminism, post colonialism etc and other positions which hold that constructing is a social not an individual practice. In these forms of constructivism then, the argument about collecting refers to a discursive position.


  8. Pingback: Martyn Hammersley’s framework for critical reading of (ethnographic) research: why I like it | Nick Hopwood

  9. Pingback: Becoming a better interviewer – between the lines

  10. PatParslow says:


    Your putative researchers are not like any I have met – nobody actually collects the sheep or people, they make observations or (indeed) generate measurements. Also, data is/are not knowledge – interpretation of data allows the construction of knowledge. Part of that may include how the data is generated (metadata).
    I’m not sure the researcher generates all of the data either. You set up the pre-conditions for its generation (say the survey) and the subjects generate the data. You then collect that data, before assembling it in to a form where it can be analysed.
    I think I might enjoy having you as an examiner – a viva could take quite a while 😉


    • Liz Marsden says:

      I couldn’t have said this better myself. I exactly agree with you. And I really think data are very much like mushrooms, you don’t pick every mushroom for a pie, just the edible ones, just like you choose which participants fit a survey.


  11. macademise says:

    How do you refer to the process commonly talked about as ‘data collection’? Do you say ‘data generation’?


  12. macademise says:

    Sorry, couldn’t figure out how to edit my previous comment – just wanted to add I’m thinking of typical journal article phrases such as ‘Data collection was carried out from October 2012 until March 2013’ or something along those lines.


  13. Gem says:

    Thank you for sharing this – I’m writing up my methodology chapter now and have found myself struggling with the very same term ‘data collection’. I’ve ignored it entirely and called it the ‘interviewing phase’, but also critiqued the concept of ‘data collection’ as one of the implications of social constructionism – hopefully that will cover me!


  14. Pingback: Two out of Three Ain’t Bad | Just words …

  15. Pingback: Managing Data: Transcripts | qualitativephdresearch

  16. Pingback: Ethics Applications – Human Research | Qualitative PhD Research

  17. Pingback: Process notes for an evaluation study | Art of e-learning

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 )

Facebook photo

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

Connecting to %s