Skip to main content
Search form
  • 00:00

    [MUSIC PLAYING]Stephen Gorard, thank you very much for talking to us.My first question today is, how do I choosebetween different methods?

  • 00:24

    The standard answer would be on the basis of whatit is you want to find out.It would be absurd to pick a method in isolationfrom what you're going to be trying to find out.So think of great questions.We know, in reality, that there's an intuitive process.You start with an idea, or a feel, or a themethat you're interested in.And as you discover more about it,in fact, what you want to do changes slightly.

  • 00:48

    So maybe your very last version of the research methodquestions will be towards the end of the project.Of course, if you change completely,it's a very different thing.But for the moment, assume that you're working on something.That you would pick the methods that would be most appropriate.Now, that sounds basic and obvious.And yet, it's quite revolutionary.

  • 01:09

    Because in Social Sciences, that very rarely happens.I'm going to give you an analogy.I've recently moved house.And in moving house, I wanted to buy a new one.In the UK, that basically involves getting a mortgage.And so I've been putting quite a lot of my money,or promising a lot of my money, for 25 years, whatever,to a bank in order to buy a house.

  • 01:32

    So how do I go about buying a house?Well I went to visit it.I looked at the roots.I looked at the neighborhood.I looked at the cost of the loan,and how much the repayments would be,how much the insurance would be, and so on.And I synthesized all that naturally in my head.I came to the decision that this house was, or wasnot, worth this amount of money.And in any skill or task that you or I dowe will tend to behave in the same way.

  • 01:55

    We will naturally collect all relevant data,synthesize them without concern--we might be better or worse doing it--but that's how we'll do it.And come to a decision that this worth it.This is cost effective.This is, whatever it is.So when I'm doing Social Science research,I'm acting exactly the same way.What I find odd is that the same people who, like you and me--if they were buying a house or booking a holiday,or looking after the health of their loved ones,would make rational decisions on the basis of allof the data available to them, naturally synthesized--would suddenly say, Oh, I'm not that kind of researcher.

  • 02:30

    As a researcher, they would do the equivalent of saying,I don't want to know the price of the housebecause I don't do numbers.Or I'm not going to visit the housebecause I don't do that kind of work,and I don't do in-depth work.I think if you think of it in termsof that analogy of buying a house or booking holiday,it's an utterly ridiculous thing to do.So you don't choose your methods beforehand,you choose the methods once you have the questions.

  • 02:51

    And once you have the questions, it's fairly obvious.If you're saying, has the extent of which childrenfrom poor families clustered in particular areasor particular schools, or are dealtwith by particular hospitals, has that changed over time?You would have to collect numeric data about attendanceat hospitals and attendance in the schools.

  • 03:12

    If you want to know, perhaps it has changed,why has it changed?Then you'd presume you'd have to dosome observations, some historical documentary archivalanalysis.So the methods fall naturally, almost like ripe fruit,from a decent question.So how do we come up with decent questions?OK.That's a very interesting one.And it's one of those areas that is, I think,currently underdeveloped in research methods training.

  • 03:39

    There are a range of different ways of doing it.Most people, perhaps when they start with a PhD,perhaps just like yourself, what they will be doingis they'll have and area that they'rereally, really concerned about.We often come with autobiographical reasons.I was particularly concerned with injusticewhich I'd come across in my professional life.

  • 04:00

    One thing, though, is to refine that by reading the literatureand so on.The danger of reading too much literatureis you may end up essentially just doingwhat everyone else does.The beauty of coming at things fresh,as you do when you start your PhD,is that's where innovation happens.

  • 04:20

    That's good.So, once I come up with a research question,then I have to devise a research design.And how am I to go about thinking about doing that?I guess you're going to repeat your answer in some ways, but--I could say more which is that if developing such questions isan immature area of research training at the moment.

  • 04:46

    Research design is almost completely absent.It's the Cinderella of the research methods arena.There are courses being taught, results that's being published,and websites available which purportto be about research design, but are actually are not.So one of the key things for me about research designis that it is independent of any methods of investigation.

  • 05:08

    Take a stereotone example like, longitudinal study?A longitudinal one where I'm goingto pursue a group of cases over a period of time.In my own field of education, I mightbe tracking some student through school or collegeand looking at some changes over time.I could observe.

  • 05:30

    I could use smell, touch, taste, numeric text data about them.The design you're following through timeis completely dependent of the methods of data collectionand analysis.So that what I would expect from the studentwould be the questions lead to the design,and also then to the methods.But the design doesn't determine methods.

  • 05:51

    So if you're going to do a round of a controlled trial,you could that by looking to see whether peoplesay they are happier at the end of the interventionthan they were at the beginning.It doesn't have to be complex, statistical analysisor anything like that.That's a very pragmatic approach to doing research and coming upwith research designs.

  • 06:12

    And I guess that contrasts, doesn't it, with some peoplewho seem to start with theory?And I wonder whether you could talk a little bitabout that relationship.Two reasons for that.One would be the field in which I work.So education, like many in [INAUDIBLE] public policyis automatically an applied area.

  • 06:32

    So if you want in-house studies, or crime, or housingor education, you tend to be coming acrossas being [INAUDIBLE] a bit, genuine problemsthat real people have out there.So in a sense, theory, as I think you mean it,is not really relevant.The government wants to know if wepay adults to go to literacy evening classes,if we give them incentive, will that improve their attendanceand attention?

  • 06:57

    It's a perfectly pragmatic question.And we as investigators will investigate it.There's not a huge amount of theory.Of course, if it does or doesn't work,we might then want to explain whyincentives do or don't work.So that would be the first reason.The second reason is perhaps a bit more controversial.I would say there's lots of methods trainingand a lot of methods advice being given by peoplewho don't do a lot of research.

  • 07:23

    There are people who become de facto methods experts,or deemed experts.And what I would advise all researchersto do is when someone wants to give you advice about research,is look at the research they've done themselves.Because if they know how to do research,they will have done it.And you'll be able to see the success of their approachand what they've done.And I find it very disquieting whenthere are people telling us how we should do research,or if it is possible to do research in a particular way,when they haven't actually tried it themselves.

  • 07:52

    And I'm not going to name names, but thereare some very high profile peoplewho are deemed to be experts in method thattell us we must do it this way and that way and another.And I've never seen an actual paperby them of any research at all.How does that come to be?Those are my two parts to the question.The third angle on it would be, yes of course, the theorycan be important, I've investigatedtheory-driven approaches, but theydon't tend to become pragmatic.

  • 08:21

    Because the theory, if it's a genuine theory,generates capital propositions such that,if you're looking at human capital theory,people will behave in a certain way and you can test that.If you test it and they don't behave in that way,then you've got some evidence supports your theory, but notthe right one.

  • 08:43

    So it's not like you can have a theory which you stick to.So I'm not one of those people whohas a theory like a religion.And I think there are people who look at research like that.I don't really understand how or why,don't really want to understand how or why.But, you do get that venerating some [INAUDIBLE] or something.And actually, and often, enjoying what theycall the conceptual looseness.

  • 09:08

    So they can make it be like an inkblot test.You can make it fit almost anything you wantand call it a lens.I don't do any of that work, because out there peopledon't want it.So theory is what we do at the endor while we're doing our research?Theory will generate propositions to be tested.Obviously, everything is involved in theory.

  • 09:29

    But I suppose I try-- I don't want to go into the details--but I try [INAUDIBLE] three different types of theory.So there's the genuine theory whichis to explore stuff, to explain it in a way that can help us.Or help us to transport solutionsfrom one area to another.Or to generalize findings from one country or one sectorto another.

  • 09:51

    And the big ticket theory-- whichis what I think of with religion, whichis unaffected by data and I thinkwould be not noticed if we simply eliminated it.So by that, you mean formal theory, or formalized theory?I mean when people are using theories as lenses.But they're trying to use the cachethat the word theory has as an explanation of a certain event.

  • 10:17

    But not actually testing it, and notbe concerned whether it works or not.But just using it, essentially just like an inkblot test.You can say, well, I'm going to use the concept of happiness,or whatever it is, to help me explain these interviews.I think what [INAUDIBLE] I've never seen anythingwhere if you eliminated that, it wouldmake any difference to substantive findingsin research.

  • 10:38

    It worries me, I suppose, that it's a red herring.So if I'm at the start of my research career, then perhapsthe better thing I should do is to developmy own repertoire of skills in different forms of datagathering and data analysis so that Ican adapt to different research questions.Would that be your advice?

  • 11:02

    There would be an infinite number of waysthat you could research anything.And you can't possibly in your last year, or six months,whatever it is, to develop skills in all of them.I think you should-- I think we should all be hungry,to learn more.I mean, that's what we do.We start learning when we are finished.So I said yes, we should be looking for more.The main reason that I think we shouldbe looking wide at a repertoire of skills,is as consumers of research, the key thingis you're going to be reading and making critical judgmentsabout existing research.

  • 11:34

    How can you do that if you've gotno idea what people have done?I mean they don't necessarily make it easy for youby explaining it well.But even if they do, if you've got no idea whatan ethnography is, or you've got no idea how you cando a [INAUDIBLE] test or whatever it's going to be,how are you going to know whether or notwhat they've said makes sense?

  • 11:55

    I can't understand a researcher whowould suggest that they were mono-method or duo-method.We all have to have some understanding of all of them.Because otherwise, what you're going to do with the stuffthat you were reading about which you don't know?Are you going to say, I will reject all of itbecause I don't do that.Or I will accept all of it, which is really dangerous,isn't it?So as a critical consumer, I thinkwe have to have all of that range,but when you come to do your more specific project,you are bound to do that skills in a particular area,or one or two things.

  • 12:26

    But then you might have a new project.What I'd like to see from researchersis that they build on the skills that they've got,but perhaps each new project, they'regoing to invent a new method.I very rarely use an off-the-shelf designor an off-the-shelf method for any project.I mean, I very rarely use the same method twicebecause the situations don't recur.

  • 12:48

    What you do is, you look at what's availableand, perhaps, redesign, or redevelop stuff, or combinetwo or more things to make an approach belongto this particular question.[MUSIC PLAYING]

Video Info

Publisher: SAGE Publications Ltd.

Publication Year: 2011

Video Type:Interview

Methods: Research questions, Research design, Philosophy of research

Keywords: cost benefit analysis; critical thinking; expertise; generalization; innovation; pragmatism; rational choice; Skills development ... Show More

Segment Info

Segment Num.: 1

Persons Discussed:

Events Discussed:



Professor Stephen Gorard explains that researchers should start projects by defining what they want to learn, then choosing research methods to best match the question. He says that too few academics have good research design skills, and he discusses the relationship between research and theory.

Looks like you do not have access to this content.

How do I choose between different research methods?

Professor Stephen Gorard explains that researchers should start projects by defining what they want to learn, then choosing research methods to best match the question. He says that too few academics have good research design skills, and he discusses the relationship between research and theory.

Copy and paste the following HTML into your website