Master’s Series on Field Research
A series of interviews with major figures in field research conducted in the early 1980s by Peter Blanck
Transcript of an interview with Dexter Dunphy
Dexter Dunphy:  Well, I’d like to talk a little about how I got involved in field research.  And I got involved as a grad. student in Sydney.  I was taking a course in education and was particularly interested in adolescent development or child development, and within that, adolescent development.  And there was an emphasis on the literature on adolescence about the importance of the adolescent peer group in socialization.  Up to the adolescent stage, the family was seen as the most important group.  And then in adolescence there was a consensus among all the theoreticians and writers on adolescence that the peer group, the adolescent peer group-* became a critical factor in taking over the family in socializing the child.  The only problem was there was very little real data on how the peer group influenced the child.  They obviously did so, dramatically.  I mean, if you look at what happens in adolescence, the, the early end of adolescence, the adolescent isn’t associating with members of the opposite sex at all, to take one critical dimension, and comes out the other end getting ready to get married, or actually being married, so it was suggested that the adolescent peer group was a vital link in that change in attitude along with hormones .  But it seemed that the hormones weren’t enough because it was primarily those kids who went through the adolescent peer groups who were the ones who mainly made that change, and made it dramatically and very clearly.  In fact, that represented about 80 percent of young people in modern urban cultures.  Now, it seemed that most of the evidence, they were just anecdotal.  No one had actually got into adolescent peer groups to see how they worked, particularly out of school context, so I decided what I’d like to do for Masters thesis was actually get out there in the field, and study some of these groups at first hand, and look at the dynamics, study them in small groups.  They were, not even the most elementary things had been written about them.  The various textbooks had conflicting evidence about how big those groups were, just taking a very, sort of simple example.  Some said the adolescent went around in crowds; and others said they went around in cliques; and others said they went around in gangs, and so on.  There was little literature on delinquent gangs.  So I talked to some of my fellow grad students and also to the faculty and so on, and said, “I want to do this.”  The general reaction was, “Well, you can’t do this.  How on earth would an adult get to wander around with teenage groups.”  However, I decided this is where I’d set my heart; this is what I’d like to do, and so I set off to try and do it.  My first attempt was a failure.  I thought I wanted some representative sample and it did so happen that one of my professors had taken a sort of stratified random sample of adolescents in Sydney, and had done a survey of them.  That was a good place to start.  I’ll just take a random subset from this group, and I’ll call up these adolescents and I’ll arrange to come out and observe their groups.  And I, in fact, did take this sample and I called the first adolescent, and he told me “No” and told me he didn’t have any friends, and he seemed a little bit embarrassed, and I was embarrassed by the end of the conversation, too.  So I then, I called another adolescent.  He said he had some friends, but that he didn’t want to be bothered by me.  And so it went on.  After about half a dozen phone calls, I was more embarrassed than the people I called.  It seemed it was fairly clear that they were not very happy about my coming out and wandering around with them.  So that was a dead end that I thought maybe my friends and supervisors were right, and that I probably couldn’t do this.  I’m not quite sure now how it happened, but somehow or other, I think somebody suggested to me that maybe I could start with a teenage club, that there were some clubs around Sydney, and maybe, as those are run by adults, I could get access to those situations and then move out into the community from there.  So this is what I began to do.  So I had to locate clubs.  So I located clubs around Sydney, called up the adults who were in charge of them--  I used for, mainly friends to get introductions to some of the ones that were around there.  They were mainly church groups, I remember, at first, and then I got into one club at first, and I was introduced to the teenagers in that club.  And I found that, indeed, it was fairly easy to move into club setting where there are both adults and teenagers involved.  I was a bit worried about the representative-ness of the group I was studying and so on, but at this time I was desperate to get in touch with any teenagers at all.  I found indeed, the crowd setting was a very good, that the club setting was a very good way to make contact, and that one could then naturally, say, when the formal procedures were over, wander out of the building with some of the kids as they were walking out; and almost automatically, they, as they got to know me, they would sort of say, “Oh, we’re going on to ,” what I was trying to say, “the Milk Bar,” or somewhere, or “We’re going to hang over here on this street corner for a while.  Would you like to join us?”, or they’d move over and just stand around, and I’d be with the group, and so it’d be natural to go on talking to people.  So step by step, it would only take 2 or 3 weeks.  I found that I was being drawn out into other activities that they were involved in, or sit with them somewhere, or going down, walk down to the beach with them somewhere, or whatever.  And so, in fact, it was possible from that setting, using as a launching pad, if you like, to get out into their regular, some of their regular activities, not all of them.  I was--, it soon became clear that they didn’t want me to get involved with some of them.  For example, I spent some time studying a delinquent gang.  It was really a group of delinquent gangs.  They called themselves “Rockers” because it was the Rock and Roll age of the 50’s, and the “real thing.”  Rock and Roll music was their thing.  And that was an interesting story in a way because I’d come into their community group which had been set up to do some good for these youngsters.  Who were roaming the streets and getting into delinquent activities and criminal activities and so on.  And so I came into this club situation, which has been set up by middle class people for, to do good for these adolescents.  And I found the businessmen arriving along with their suits, and teenage Rockers with their leather jackets and so on.  And in the end, that group didn’t work, in the sense that the teenagers ended up tearing the club apart.  They came back to do that after they’d left, but when they walked out one night, they just simply walked out, leaving all the adult leaders behind.  And I thought, that’s the end of this part of my study.  I wandered out after them, and they were standing around complaining about the broken promises that they felt had been made and broken by the adult leaders.  And I sort of pulled a long face and I said, “Well, I’m just this guy who’s trying to write this book about teenagers, and this makes it difficult for me, guys.  What am I going to do?  I mean, I’ve spent a few weeks getting to know you guys and so on.  Now you’ve walked out of the club.”  They said, “Oh, that’s all right.  You come out to the wharf (This is where they hung out, in Sydney Harbor).  You come out to the wharf.  We’ll introduce you to our friends.  You’re all right.  You’re not like those sons of bitches from that group back there.  And so I went, was invited, at that stage, to come down with them.  And they were looking for someone to talk to.  I think one of the issues in field research is what can you exchange with the people that you’re working with in the field, and with a delinquent group, for example, I found that one of the things I’ve exchanged for them, that they valued was acceptance by an adult.  This was a group of kids who had dropped out of school, many of them.  Some of them had been asked to leave by teachers.  They, in some cases, they had been thrown out of their homes by their parents.  They had been taken away to, as they say, “sent up the river,” by a judge.  They had small periods in reform homes and so on.  So their experiences with adults was that adults were non-accepting of their way of life.  Something I found I could give them was the feeling of being accepted by an adult.  I was quite happy to sit there and listen to them talk about their school experiences, or their experiences with various other agencies of the community, or with their parents, and not–, I deliberately withheld judgment on their activities.  And that kind of acceptance was something that they were searching for.  So we formed–, in that particular case, for example, and I found this to be true of most of the teenagers I've dealt with– very often they were in some kind of rebellion against adults, or felt that adults were judging their life.  To have an adult who was sympathetic, and who would listen, was a resource that they wanted.  And so, in exchange for being there and them giving me the information, I was giving them the acceptance of an adult, and a non-judgmental acceptance at that, which they seem to value.  And there were other things that I could offer.  I had a car, and for some of the younger teenagers, they didn't have transport, so if they wanted to get somewhere, they were very happy to use me as a form of transport, and that was another simple thing that I could exchange.  And I guess I had more knowledge of life, so if they wanted to do something, or achieve something, then sometimes I could help them on some advice on how to do that.  But basically, I think it was the acceptance of them as people and the willingness to listen and it turns out that most people want to talk about themselves.  They enjoy talking about themselves.  And if they can get a captive audience, then that's great, a great thing.  And that's one of the things that a field researcher has to exchange: attention, and interest in other people.  And very few of us find opportunities to have people sit there and listen to us for hours on end.  But that's something that a field researcher can offer, and most of us like to talk about ourselves, I guess.

I think that when you go into other kinds of field situations, there are other kinds of exchanges.  For example, I do a lot of work on consulting, with businesses and with governmental departments, and one of the interesting things I find is that if I want to do some research, and I go, hat in hand, to the senior manager and say, "I want to do some research."  He'll put his hand on his heart and say, "Well, I'm–.  Of course we're interested in knowledge, the advancement of knowledge and so on.  Yes, we might grudgingly help you but we'll limit your access to this data, and we won't give you this information and so on and so forth."  Because I'm not really exchanging anything very much, I have nothing to offer.  In fact, I'm asking for something and give nothing very much in return except some– maybe the prestige of being associated with the university, but that's fairly marginal, and it's usually not very public.  On the other hand, if I go as a consultant, and very often, for example, the firm will approach me and say, "We have this problem.  Will you help us solve it?"  And they are in fact paying me.  I am exchanging expertise in solving their problems or helping them solve their problems, and for that, they will give me all kinds of information.  They'll bring me information.  I don't even have to ask for it.  They'll beg me to go and look at a certain site, to go and interview people, to run a survey and so on.  So the situation changes dramatically.  What I do have to be able to produce, however, is some relevant skills in terms of helping them sort out what their problem is, and come up with solutions themselves, or to be able to provide some alternatives, or some skills, or some information that is critical to them, to the solution of the problem from myself.  Usually, it's a combination of those 2 things.  Although in the home, my philosophy would be to try and get them to solve their own problems themselves.  So the kinds of–, I see myself offering the skills that allow them to, allow them to categorize the information, come up with some options, and then make judgments about how those options might be pursued.

Peter Blanck:  Of course, the more you get involved consulting-wise, with the firm, the finer the line is between the objectiveness of what you can do as an observer and the goals that you have to fulfill for them.  I think you'll talk about that later on.  Maybe you want to continue on.

Dexter Dunphy:  Maybe I'll just take that sort of issue up here.  It seems to me that social scientists, when they're doing research, have a notions that somehow or other, a notion of objectivity.  The objectivity of a social scientist.  I think it's a lot of bull-shit, frankly.  None of us are objective.  We are all interested.  We all have special interests that we're pursuing as we go into a field situation.  I think the important thing is not so much what interests we have, or whether we have interests or not, because I think we always haves interests of our own, or they're interests of someone else's, or are being paid to do some result.  All of this biases our data.  Every social scientist is in fact biased and subjective in terms of his analysis.  The initial set that he brings to a situation that may be a theoretical bias for example.  So it seems to me, the search for coming in objectively is a search that is never going to be fulfilled.  The important for me is to know the sources of your bias and understand, as far as possible, the kinds of bias he is bringing in, and to try to offset these biases in some way, by trying to get insights from various sources.  And one of the things I think I've learned from field research is that you are not doing it yourself if you do it intelligently.  I learned, for example, in studying adolescent peer groups, not just to use my own observations as a source of data, but to employ everybody I could in the peer group itself as another set of eyes and hands, another set of observers.  So, for example, I would not only go and observe what other people were doing, but I’d ask other people what they had observed in these groups, the participants themselves, and I'd find that they observed, very often, different things to the things I observed, because they observe from a different point of view.  I also ask them to keep diaries of whom they contacted, and when they contacted, and why they contacted; and who contacted them, and for what reasons.  So they became my research assistants, if you like.  Looking at their world through their own eyes, and they had access to situations which no way I would have access to those situations.  I couldn't sit in every one of their homes for 2 weeks consecutively, but I could ask them to do that, and then take those diaries, which each individual kept, and check the diaries for accuracy against each other.  And then interview them separately if there were discrepancies and so on.  I could go to adult club leaders, for example, who had access to many of the same settings, and ask them how they saw the situation.  I could go into the kids' homes and talk to their parents and ask them how they saw the situations.  So, I learned early, I guess, to offset my particular biases by collecting the biases of every other interest group in that area.  And this has come out into working as a consultant in industry.  One of the first things that I do is try to locate the critical interest groups in any organization that I'm studying, or any social setting that I'm studying.  And then I try to make sure that I use the biases of those interest groups to do a kind of cross-cutting analysis of the organization.  It's rather, I think, like a surveyor trying to survey a terrain.  He might do a magnificent and very finely tuned survey from one particular vantage point, but that's nowhere near as good as getting 3 rough surveys from 3 different points, and being able to triangulate on various points in the environment.  It seems to me social research, or field research, at its best, is like that.  If you can use observation, direct observation; if you can use secondary observation by people in the groups, you can supplement that by interviewing.  You can supplement that by the use of diaries, surveys, questionnaires, and so on.  The more of those methods you get, to look at the same basic set of data, even if those methods are rough; the more of those methods you have, the more likely you are to arrive at some understanding of what that camel looks like, or I should say, elephant, I think the analogy is.  Somebody will be holding its trunk somebody will be holding its tail; somebody holding its legs.  And you're able to put the elephant together by getting the viewpoints of the various people who are standing around the elephant.

Peter Blanck:  Do you want to continue on the chronology of the gang study?

Dexter Dunphy:  Sure.  Just trying to think where I was there.  Yes, I went into the study of adolescent groups, trying to solve a problem, and that is, how do these groups transform so effectively the social attitudes and role behaviors and so on of teenagers.  And this is, this is interesting in terms of using the different kinds of methods because on of the first things that I did was to either interview or give out a very brief questionnaire to, to the members of, let's say, a club, associated with the club, asking them who they saw as members of their group, and I collected that data and I put that to one side, and then I observed who seemed to hang around together, who seemed to talk to each other and interact with each other.  Now, the reason I started on the second road, and started to do it more and more systematically was when I first got that questionnaire data, it didn't make any sense.  I found what Coleman discovered when he, in fact, did a study on teenage groups in, in school settings.  He read the group adolescent literature, too, and he found that there seem to be a contradiction between the adolescent literature, and what he'd got from his survey data.  He found, instead of nice little cliques, which is all the adolescent literature talked about, he found chains of association, and they didn't seem–, if you've got a clique, you would expect them to find everybody in that clique choosing each other, or most people choosing each other.  And that would give you a nice boundary around the group.  When I asked my group, I got chains, too– no groups.  But when I observed teenagers interacting, I found groups, clear groups.  And if I ask people in one group about who somebody else belonged to who wasn't in their group, they'd also talk about groups.  They'd say, "Oh, Joanne West has a group up there, in that particular area."  And there'd be a name, and there'd be a lot of correspondence, regardless who I was talking about , in who they'd name belong to that group.  But this didn't seem to fit with their data, so with the data I gathered through the questionnaire, or if I asked an individual, "Which group to you belong to?"  He didn't seem to fit with the group he is associated with.  And I put that data away form quite some time.  I thought maybe there's something wrong here.  They didn't understand the question: That didn't seem like it made any sense.  They're lying: That didn't seem to make any sense.  I had the feeling these kids were being very sincere.  Finally, when I had a lot of data on the actual associations, the real associations that I had observed, and were reported to me by others, then I came back to these other data.  And I tried to put together these choices with a map, a sort of social map I made, of who was associated with whom.  And suddenly, it fell into place.  What I had gained, I began to realize, from asking people which groups they belonged to, was their reference group; and there was a shift, a strong shift, between reference group and the membership group, and that shift was systematic in one direction– upwards.  If you came into a local region, you found there were a group of small–, cluster of small groups, which are called cliques, that as you got up to the middle adolescence, those cliques join together.  They still remain cliques but they became a crowd, a much larger association of cliques.  Maybe 30 or so young people in each clique might have only 5-10, averaging 6 teenagers or so, and then together would make up a crowd.  And then the older adolescents, they move–.  The cliques are starting to break up again, but now, maybe heterosexual instead of unisexual, as they were in the early stages of adolescence.  You had boys' cliques and girls' cliques in the early stages.  So it seemed as though the crowd was the key in middle adolescence, somewhere around middle adolescence is, one-sex groups came together, started mixing with each other, and then changed composition.  And if you ask people who are the members of their group, they would name some of the upper status members in their membership group, plus some of the lower status members, or sometimes, some of the leaders of the crowd group, or the next group up in the hierarchy of these groups; so each locality had a hierarchy, a status hierarchy, roughly associated with chronological age.  But really more associated with heterosexual advancement, if you like, that is, the more heterosexually mature an individual was, the higher he was in that hierarchy.  And heterosexual advancement is relative.  I mean, that might be just at the lower age levels that might be–.  this kid is really ahead of his peers because he is able to run around, chases the girls and put snails down their backs or something like that.  This is really advanced heterosexual behavior at age 13; but at age 16, it might be dating and actually getting to the stage of kissing, and doing some of those strange things that older adolescents get into.  So this became–, this apparent discrepancy between what I observed and what was reported to me by individuals about which groups they belonged to.  It became in fact to be the clue about how–, the power of these adolescent groups.  It turned out that in order to get into this whole process of socialization, you have to get admitted into one of these cliques.  There was no other way in.  You couldn't come in as an individual.  You can only come in as a clique.  Getting into one of these cliques as an achievement-type thing.  You had to fight to get in.  Everybody talked about that process.  They were very tight, very exclusive, and very norm-controlled, that is, they had clear ideas of what you did, or what you didn't do, and you had to conform to those things to be accepted.  Once you're accepted there, the most advanced boy or girl in that clique was the group leader, and you identify very strongly with that group leader.  That group leader associated with members of the next crowd up, and identified with them.  And it was this upward identification process which gave the whole system its power.  If I'm down here in the hierarchy, at this age level, then I identify with my peer group leader  He identifies with the lower status members or the leaders of the next crowd.  They identify with the next crowd up, and so on.  So that each person is referring himself to a–, to people who are more mature heterosexually than he or she is, and this way, there's a sort of, a set of hooks, if you like, going down in through this hierarchy pulling people up, and as they're getting older, and getting more mature, they are learning new skills.  The cost involved for the individuals of staying in those groups is to keep acquiring the next level of heterosexual skills.  So one can only find this out, in fact, by getting into a field situation.  To get into a field situation, you had to establish some rapport with these groups, and start to move with them, and get these 2 kinds of evidence: the observational evidence, and the interior psychological representation of that external reality that people carry around in their head and in their hearts.  And so, to me, one of the lessons that I learned from that, in doing field research, even today, is that there are 2 levels of reality.  There is what the social scientist sees as the external reality.  What he calls, usually, "objective observation," although I don't believe it's fully objective, s I've said.  And there is the interior representation of that reality, which may be something very different, again, which you can only get by, in some way, reaching inside of someone, and having him tell you how he sees the world, in some way.  Now, you can do that through questionnaire.  You can do it through surveys; you can do it through interviews; you can do it through just hanging around listening to what people say and maybe you need to do it on 2 or 3 or all of those ways and get an accurate representation of the internal constructs of the people in that situation.  But you certainly need to get it from the key interest groups in whatever that situation is.

Peter Blanck:  Now the methods for collecting data– Let me just see if I've got this straight– are accessible to both types of issues.  The objective, as you say, and the more subjective, the interior.  But it's more difficult to get at the interior issues.

Dexter Dunphy:  Yes.  I think as a social scientist, the observational methods get at, generally get at the, the exterior reality, if you like.  We, we can all see who interacts with whom, for example, if we hang around with a group out in the field, although it may be more difficult to do that in between group meetings.  And the case when I was there, on of the problems was, how do I find out what kids are doing during the week, when they're scattered around their homes.  I can't be in all of those places at once.  And I found the answer to that by getting the kids to become observers for me and keep diaries.  So I couldn't observe them directly, so I used them as observers for me.  But I used all of them as observers, so that I could check, cross-check, the accuracy of their observations.  But to get to that interior psychological representation, it seems to be the most natural methods are interview, interview and simply listening to conversations between people by just being around.  Now I think Street Corner Society, for example, gives a very good model of how to do that, particularly the research methodological section at the back of that book.

Peter Blanck:  You had–.  This is switching gears a little bit.  When we first talked, you had mention quick and dirty methods of survey research, and how often you have to balance the need for practical feedback with less representative sampling techniques.
Maybe you'd want to talk a little about that.

Dexter Dunphy:  I think along with this notion of using multi-methods to zero in on some data, I'd like to put the notion of using what I call "quick and dirty: methods and get quick results.”  I say, I use the term "dirty" – I use it with inverted commas, or in quotes – because I think many social scientists see these–, Strict social scientists see these methods as "dirty."  As social scientists, I think we're trained to be very, very particular about the kinds of methods that we use: To be very particular about sampling approaches, the representative-ness of the kinds of samples we get, to spend a long time on questionnaire construction and so on.  My own feeling is that sometimes, we have trade-offs that we have to do between perfecting those kinds of factors, and some of the real exigencies of working in situations where you need get feedback back in time, or get data back and analyzed in time for it to be useful or relevant.  Even as a social scientist, I think doing studies, quite often you – if you spend more than 6 months doing a study – the data that you're getting back is not useful from the point of your own research, particularly if you're doing consulting with groups, or action research.  Those groups want data and they want hot data.  They don't–, it's no use going to management groups that want to know about morale in their organizations, for example, and saying, "Yes, I'm going to run you a perfect survey.  It's going to take us 6 months to write the questionnaire.  Its going to take us another 6 months to administer it, do the pilot studies and so on.  And it's going to take me another 3 or 4 months, or 6 months to analyze it."  And I'm not exaggerating.  I know of many cases where social scientists, although they give a shorter term in which they are going to do that, in fact take 18 months to plan, carry out, and analyze the results of a survey an organization wants surveyed.  My experience with that is that by the time that survey has been completed, management thinks it's a big yawn, and they just throw it into their filing cabinet, and forget it, because 18 months later, that data is useless to a manager who wants to make decisions.  So, one of the things that I've been working with is methods that are reliable enough to give us data that is relevant, is a pretty good approximation of the reality out there, and yet is going quickly enough for it to be relevant to actual decisions which have to be made in the pragmatic fashion.

Peter Blanck:  Did you develop actual methods?

Dexter Dunphy:  Yes, the kinds, the kinds of methods–.  I think, for example, of a large survey I did for an international, in a subsidiary of a large international corporation.  The executive group wanted to consider its future strategies for the next 10 years.  In order to do that, they wanted an audit, if you like, of the current state of their organization, and where people within the organization at all levels felt the organization might be going.  What were the critical problems would the organization face in the next 10 years?  And what aspect of the organizational life needed some real attention because of this?  Now, they didn't want this data in 6 months; time; they didn't want it in 12 months' time; they wanted it in 2 months' time because that was when they needed the data to–, for planning.  And so, in fact, it proved to be possible to plan that piece of research, to do a sample set of interviews and questionnaire, questionnaire combined, a sort of combined interview-questionnaire approach through the whole organization, a very large organization, Australia-wide and Australia is as large as the continental U.S., and to get those results, and to analyze them all within 2 months.  Now, most survey researchers would balk at the prospect of doing something like that.  How did we do it?  We simply picked up an existing morale questionnaire, or organizational survey questionnaire, which had been used 2 years before in the organization, which had taken a long time to do.  We picked out key questions that had been used and which had seemed to be significant that had come out of that survey.  This allowed us to make up a short questionnaire very quickly, which had already in a sense been pilot-tested.  We then also devised a set of pilot interview questions that could go along with that, and as interviewers, we went, we took a number of managers who had retired early in their early 50's, who knew the organization but were no longer part of the power structure, and we used about 6 of them and we, in 2 days, trained them to go out and do these interviews, and to administer these questionnaires.  They knew the organization thoroughly.  They helped me choose an appropriate sample, which was a kind of stratified random sample, but over-representing some of the critical interest groups in the organization.  If you have, for example, as you have in this industry, a dozen people in one area of the organization who can stop the organization dead in its tracks by going on strike, you're wise to see that they're represented in your sample, although their numbers might not be particularly significant, because politically, they're significant.  You may see that that raises some sampling issues most social scientists don't look at.

Peter Blanck:  I want to press you on that sampling power structure issue.  Maybe you want to talk about that.

Dexter Dunphy:  Sure.  Let me talk a little about sampling the power structure.  Most social scientist, when they're going into an organization to do a survey, don't sample the power structure.  They get an analysis of the number of employees in the organization.  They ask what categories they're in-- by age, by skill level, and so on, by status in the organization, formal status.  But they very seldom ask for a representation of the political structure, if you like, of the organization, the power structure of the organization.  From my point of view, I think one of the most important things we're doing in organizational change, and that's usually, that's what managers is about, we're making some kind of impact on the power structure, and if we're going to study a power structure, then we have to sample the power structure.  And you might have a group of 12 people, or of 8 people, in an organization of thousands who are in a critical situation.  If it's in an oil company, for example, those people who connect up the oil pipes in off-shore oil rigs-- there may be a relatively small number of them, but if they don't connect those oil pipes, then the oil doesn't flow, and that organization does not operate.  So you need to have, to make sure that they're represented in that their views are represented in terms of any kinds of survey.  So one of the things that I do is to move around the organization, interviewing people at various levels, asking them about the power structure.  And they tell me about the formal power structure and they tell me about the informal power structure, and they tell me about the industrial relations power structure.  And out of that, I try to create some kind of power map of the organization.  Then, I sit down with people who know the organization well.  Again, I like to get a group that doesn't represent just the management group and I ask them, "What sort of number, if we are to take a sample of 100 or 150 for interview, how, how do you think we should sample from this group?  And out of that, we come up with a new sampling design that is a sample that represents the relative power of those groups.  Now, I don’t know.  I've never seen social scientists sample in that way before, and it seems to make a lot of sense to managers in the organizations and union representatives and others within organizations.

Peter Blanck:  That's good.  You had used the term "action research" a couple of times.  Maybe you could define that as you see it.  What action research is.

Dexter Dunphy:  Action research, as I see it, is sort of research on the run.  It's research designed for--, it can be designed for social science purposes, and it can be designed for the purposes of managers or others who want to change organizations.  And sometimes, it's designed for both.  In other words, I, as a social researcher, might be interested in understanding characteristics of certain--, of large scale organizations, which have highly sophisticated technology.  I'm not interested in studying those organizations and doing a study that is going to last 3 years because in fact, if we look at the introduction to new office technologies, I want to know how those technologies are being introduced now, the problems that are being faced now, the kinds of human problems that immerge a result of those technologies now.  I don't want to know how the--, in 2 years' time.  I want to know in 6 months' time, so that I can write an article soon enough so that those people who are studying the field making recommendations about it, trying to understand it, to know in time enough so that they are able to influence the course of events.  Similarly, managers who are introducing new technologies don’t want to know what's happening, where the problems are occurring in 2 years' time.  They want to, they want to know where those problems are occurring now so that they can take actions to try and overcome those problems, and they want to know that in a relatively short time perspective.  So very often, my interest as a social researcher, and interests as a manager, a practical man who has to deal with the real situation, coincide, and that's what I call action research.  When we go in with a short time perspective, we want to know how that process is operating at this time, and we want to know--, we want to be able to understand that quickly, so we can intervene, or we can tell others how that is operating before it changes.  One of the things that is happening in organizations now is that the rate of change is building up so fast that we don't have time to do long-term social scientific studies.  If we take that sort of time, that situation is changed.  It's no longer what it was, and the research that we did is no longer relevant.  My experience, once when I did some research on rapid growth organizations, I studied 100 organizations.  I did the traditional kinds of research-- long time perspective.  It took me 18 months to do it.  By the time I finished getting, by the time I got to the point where I could write up my data, it was a no economic growth situation.  Nobody was interested in it any longer.  (I wasn't interested in it any longer.) In the problems that, of rapidly going organizations, everybody wanted to know if that organizational attrition and how do we organize for that?

Peter Blanck:  The ideas that you were talking about is related to the exchange idea in that this whole thing is tied up with getting feedback, which is something that many people haven't talked about.  A lot of people have talked about intensive observation that you talk about with your work with delinquent gangs, and yet the feedback idea is very interesting.

Dexter Dunphy:  Can I pursue that one a bit?  I'd like to pursue the notion of feedback in action research.  Let me give you an example from real, from real life.  I was brought in to work with a group of managers who were planning a large scale resource development project in western Australia.  This was a new mine.  There was nothing there but desert.  There were only 6 managers.  Right at the very beginning of this project, they had to design a new town.  They had to design a new organization to get this particular ore out of the ground, and to process it.  And I was brought in to work with them on the human resource aspects, and that meant all sorts of things about how, how--, what sort of work force should we have?  It led us into the business of when I asked, for example, "What sort of work force do you want to have?"  They said they want to have a stable work force.  I said, "Does that mean married people?"  And they said, "Yes, we want to have couples, not necessarily married, but couples.  That'll give us more stability, we believe."  Then, I was led to ask, "Okay, where are the jobs for women?"  Now, they're tradition Australian mining managers, and they put up their hands in horror.  "My God, jobs for women?  We don't give women jobs in mines."  And that led us on to a whole discussion on values and so on, and we ended up with jobs for women; offering, hiring couples in dual career programs and so on.  Now, when we designed the town, working with architects, town planners and so on, and we designed the organization, and we in fact began to get construction on site; we got the first few houses; we had a shaft dug for the mine.  So we had an elementary organization, and we had an elementary town community, and we had a few people at this stage.  I then came in to actually go through and interview people as they moved into the houses.  We had designed these houses.  We designed living quarters for single people.  We designed them according to certain principles.  I then went and asked the first people who were able to test these how they found them.  Did they find them satisfactory?  Where did they fall short of their expectations?  What other things could be done?  And I did the same thing for the organization as it began to develop at various critical stages of its development.  For example, as we moved from, began to look forward to moving from the construction phase of the organization into the production phase, I was brought in to ask people what problems they anticipated, and then to work with them to try to eliminate those problems, as we moved across into that phase.  In other words, I was providing important feedback loop to management, and to all of the planners who were working with management, including myself, on how effective the plans were.  And that feedback loop turned out to be absolutely vital because some of the policies that we developed, in practice didn’t work out.  But we were able to pick up the bugs in the policies or in the implementations of those policies early enough to be able to modify the implementation or modify the policies in time to stop those problems from becoming large problems, and I believe that that feedback cycle is extremely important.  Now, as a result of that, I've changed my model of social research.  The old model was, you make a pilot study, and then you go and do a massive, one piece of research.  I'm a strong believer now in doing progressive studies with--, which ask some questions, go in and get some data, refine the questions, then ask the next set questions with another smaller set of studies, and then refine those questions even further, coming into another set of studies and so on, which I think is a different model of social research.  But it builds a feedback, a set of feedback loops in there which I think is very important.

Peter Blanck:  Are there any other issues that you have written down that you want to talk about?  

Dexter Dunphy:  I think I'd like to talk a little about the role of the field researcher.  There's a--, an important question, I think, as a field researcher, is to--.  Once you've gained access to a situation, how do you define your role in relationship to those people who are in the group?  I think one of the mistakes that's always made by early, young researchers, or researchers who are new to the field research, is to over-identify with the group they're researching, and I think most field researchers have anecdotes of this kind.  For example, I remember when I was moving around with one of these delinquent gangs; in fact their language was really very strong.  They used a lot of four-letter words, and this wasn't my normal vocabulary, but that was part of the culture of the group.  And so, after a while, I decided that that was the way they talked; maybe, not very consciously thinking about it, I thought I might be accepted better if I started to use their language.  And so, one day, I came out with a long string of profane words, like the ones they were using; and they looked at me, and they were really shocked.  And they were also very resentful, and I didn't understand this at first.  They told me I shouldn’t talk like that.  And basically, I guess, as I listened to them, what they were saying to me was, "You're being untrue to yourself.  You're not being for real,"  and I think they were also saying to me, "You're pretending to be one of us, and you're not one of us, and keep that in mind.  You're a guest in our group."  And I realized that as I moved into a group, one of the resources I had to exchange for them was to be a representative of some outside reality that they could test their ideas against.  The more I identified with the group, the more I become one of them, the less I had to offer in that regard.  And I think many field researchers of my experience have stories to tell like that.  In this exchange that goes on, it seems to me quite often, they'll try to strike a hard bargain, and they'll test you.  I had one experience, for example, where I had the same delinquent group in my home.  Unknown to me, I've driven them in my car, but they brought a whole bagful of beer bottles, full beer bottles, and which they proceeded to bring out in my flat.  Now, already the neighbors--.  I had about 15 of them, and they were a pretty rough group, and they'd also brought a couple of group molls with them, and the neighbors were already threatening to call the police because they kept taking these young girls down to the garden-- in their terms-- for a "quickie," and the neighbors didn’t appreciate this sort of exhibitionistic performance in the back yard of the block of apartments I lived in.  So I was a bit concerned about this, and I was also concerned as this gang would--, sometimes, when they got heavily into the liquor, they get very violent, and I didn't particularly want that to happen in my, my apartment.  They had already torn one youth club apart.  So I was, I was a bit upset about this, and I tried to persuade them not to drink this.  They, the leader of the group proceeded to open one of these bottles despite my attempt to stop him, and I realized that this was a test, and that I had to do something very quickly, so I decided to take a risk.  And it was a very scary moment, actually, because they were carrying knives and other things.  And I jumped the leader of the group and took the bottle away from him.  I, the thing that was scary was I didn’t know which way the rest of the group was going to go.  I had a struggle with the leader.  Only one person came to his aide, and I was able to kick him swiftly in a painful place, but it was a real, a real fight.  I realized afterwards that they were aware of the fact that they were guests in my home, and in a sense, they had some sensitivity to that, and they were just testing me.  But it was a rough scene for a moment.  Subsequently, we were able to work out an arrangement whereby they had some liquor, and I deal with them.  They had some liquor to drink, but they didn't get drunk.  Now, again, most field researchers would tell some sort of story like that, where the group will test you for how far you can go, and it doesn't matter whether you're doing field research in an organization, a company, whatever.  They will test how hard they can push you.  They'll also test whether, in fact, you would treat their information as confidential and so on.  So, one has to work out as a field researcher, a set of norms about how you are going to behave.  You have to make sure that you understand your role, and you work that role out as you go, but you don't allow yourself to be manipulated because you'll only be respected as you respect yourself, as you place value on what you have to offer them, as you act as a person with integrity, and that is true whether you're dealing with a delinquent gang, which may not seem to respect integrity, but they have their notions of integrity, too; and they expect you to be for real, and they'll be for real with you, if you're real with them.

Peter Blanck:  What about the issue of confidentiality?  How do you handle that?

Dexter Dunphy:  The issue of confidentiality is, I think, an absolute, number 1 priority issue.  I handle it by saying by making some kind of contract with the group.  If it's a management group, for example, and I'm going to be asked to give a report, I will interview people, or get information from various sources, from informants, but I make it quite clear that I will never reveal anything that will identify any particular informant.  But, however, I will make an overall summary of what people tell me, and I'm very careful to modify any information in such a way that no information can be traced, and never to say anything to anybody in that organization about anything that anyone else has told me on a one-to-one basis.  And that's an absolute rule, which must not be broken under any circumstances, and people within that organization will always test you to see whether you'll do that, and if you do it with one person, everybody in that organization will know the next day that you're not to be trusted.  So, if you want that information, if you want access to that information, you have no alternative but to respect the integrity, the confidentiality of everybody in that organization, no matter if it's the top of the organization, or the bottom.  And I think that's very relevant for consulting work, because one of the things I make clear in my initial contract with the chief executive, if it's the chief executive that brings me in, is that my contract, as I see it, is to the organization as a whole; it is not to him as an individual, that is, my responsibility is to that total organization.  I will not reveal anything to the chief executive, even though he is signing the checks, that is confidential to any particular individual, nor will I represent his interest as against any other--, the interest of other groups in the organization, including unions, blue collar workers, and so on.  In other words, I see myself in some way, as most valued to that organization if I reflect back to them the reality of their situation, the reality of the interests of the various interest groups within the organization.  The moment I identify with one particular subgroup, or become a member of that organization, in a psychological sense, by using, say, the work "we", meaning you and me together, then I'm losing my value to the organization.

Peter Blanck:  In your field work, with gangs, for instance, how did you keep notes?  Did you keep a diary for yourself at the end of the day or--.  You certainly didn't use a tape recorder, I wouldn't think.

Dexter Dunphy:  Sometimes I did use a tape recorder, but always asking the group if they were happy for me to use that.  Sometimes they were; sometimes they weren't.  I'm amazed at how a lot of people don't mind being tape-recorded.  What I did with the teenagers was--, a lot of them, at that stage, tape recorders weren't so common.  A lot of them hadn't heard their voices on tape recorders.  It was a new kind of toy.  So I let them play with it, and listen to their own voices.  And as they become familiar with it, then they were pretty happy to have it around, and they didn't bother about it.  The situation's somewhat different now.  But on the other hand, people are very used to having tape recorders.  I kept a diary.  I normally didn't make notes when I was wandering around because that's a bit conspicuous, and people become very attentive to the what it is you're writing down.  But I was not above sort of wandering into a john somewhere and making a few notes when nobody was looking, or finding an opportunity to just slip into a corner, and note one or two key things down, like interaction pattern and so on, that I could then amplify on.  But mostly, after I've been with the group, then I'll go somewhere and make some notes afterwards, but I still use that kind of practice when I'm doing field work in organizations.  If I'm interviewing an individual, I've no hesitation in making notes on what they're telling me, and in fact, I find that often is a good listening device.  Because I'm writing, there's a pull there that pulls them into talking.  So it's quite useful in some--, in putting pressure on the informant to give you information.
	(Break in taping)

	It's interesting to look at the traditional model of social research that we inherit, which said--.  This is the model we learned in graduate school, I guess.  You have theory, then you have hypotheses; and then you set up a pilot study to really test your method; and then you do your field research, which might be a survey or main research, whatever; and then you have your results, and you write it up; and that comes back to your theory.  Now that's all very well as long as you really know what your hypotheses are.  But it seems to me that mostly in social science, and certainly in a lot of field research, you don't have a clear theory; you certainly don't have clear hypotheses.  One of the things you want to do is generate some interesting hypotheses.  So it seems to me that you start out with theory, or-- that may be too dignified a word-- a set of hunches, that relate perhaps to some theories.  Maybe not one theory, maybe 2 or 3 theories.  Out of that, you then derive some ideas-- again, lets call them hypotheses, but that's again, maybe too dignified a word.  You then get into doing some field research, but this should be small scale.  You then refine your hypotheses, and you generate some new hypotheses.  So you've got new ones, and then you go into your field research, and maybe you expand in your field research.  And you test them, and refine them again, and so on, and so on.  Now what you're doing here is building in some feedback loops here, so that you're feeding back here; you're feeding around, and then feeding back to your theory again.  You keep recycling.  Now, it seems to me that this is a useful model not only for field researchers, but it's also a useful model for consultants who are working with management.  Management is setting up policies.  They're implementing those policies.  They want to know whether those policies are succeeded before they get right down to the end of the road, because if they fail then, the whole organization has failed.  So it's both a model for action research from a researcher's point of view, and from the point of view of consultant.
Dexter Dunphy:  I really enjoy field research.  The thing that fascinates me about it is, it is that's, it's live.  I like to look at real situations rather than, I guess, artificial situations.  I like to understand the real world, the way it really is, and I like, in a funny way, the difficulty of field research.  It's challenging.  It's to me a much more exciting and sensitive process than the kind of research that's done sitting in laboratories, behind desks, and so on.  It brings me into contact with a lot of people.  I guess I'm curious-- nosey, if you like.  I like to just wander around the world poking my way into sort of odd corners, finding out how the world looks from this perspective-- blue collar worker on the line, manager up here, or engineer over here.  And it's exciting also because it's concerned with, generally concerned with change in some kind of way.  That's the kind of field research that I enjoy.  So I like to get involved in research that matters, that's going to have some practical consequences, where I'm going to see some results in a fairly immediate sense, for what I do.  And maybe that's just a personality thing.  I like, I'm the sort of person who likes fairly immediate feedback.  I don't like to wait 3 years to see results.

Peter Blanck:  Good.  Okay.  
Ends at [00:04:58] in movie 5311ShortsBalesDenphyLawre.mov
Master's Series on Field Research 		16