4.1.2 Questions About the Constructs

К оглавлению
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 
17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 
34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 
51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 
68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 
85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 
102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 
119 120 121 122 123 124 125 126 127 128 129 130 131 

(3) The interviewee took a long time to get the idea that I wasn’t looking for ‘right

answers’, either saying so directly, or making comments about ‘oh, I wonder what

you’ll think of me for saying this, but . . .’ Does this matter?

Only if it gets in the way of the constructs. These should, indeed, be the

person’s own, uncontaminated with socially or organisationally desirable

ways of viewing the topic.

In this age of psychometric testing, pub quizzes, and selection interviews, where

‘the right answer’ matters, I suppose it’s natural for people to wonder about this.

You need to check that your reasons for conducting the interview are understood

(see Section 3.1.1), and to stress that your purpose is to discover what the

interviewee thinks for its own sake, regardless of whether people might agree or disagree.

Point out that, in the topic you’re exploring, there are no right or wrong

answers and that, while there may (or may not) be preferred ways of looking

at things, the only reason you’re doing it this way is to ‘avoid all that:

otherwise, I’d have used a standard questionnaire or a different kind of

interview’, or words to that effect.

Emphasise the confidentiality and anonymity arrangements which have been

made, and reiterate point 4 in the basic procedure (Section 3.1.2).

(4) The interviewee got stuck and couldn’t offer any construct for a particular triad of

elements at step 10. What to do?

No worries. Drop that triad and offer another one. Or try one of the alternative

elicitation techniques listed in Section 4.2.

(5) The interviewee got stuck on the very first construct. Help!

Try each of these, in order until you find one that works.

Acknowledge that, ‘Okay, that particular comparison doesn’t suggest

anything to you; here’s another one’, and offer another triad.


Illustrate the procedure by offering a construct of your own if the nature of the

elements makes it feasible. (If they’re names of people known only to your

interviewee, it won’t work.)

For example, if the topic is ‘politicians’ and the elements are the names of

politicians, you might say something like, ‘Well, if you were to look at Churchill,

Hitler, and Stalin, one sort of contrast that occurs to me is, ‘‘Two of them were

rather unpleasant, Hitler and Stalin; but, in contrast, Churchill, for all his faults,

didn’t rule through terror and fear.’’ Do you see what I mean by looking for

something that two of them have in common which the third one doesn’t share?

Now, ignore me, and tell me what contrast occurs to you’, or words to that effect.

At this point, it’s essential for you to emphasise that that was just one of your

constructs; that you’re interested in his/hers; to discard your own; and to

invite the interviewee to consider the triad and find one of his or her own. The

example you offer shouldn’t be factual (‘Churchill and Stalin died a natural

death while Hitler committed suicide’) because, while illustrating the idea of

contrast very clearly, this may confuse the interviewee: you are not, after all,

looking for correct answers. A straightforward but uncontentious opinion

would seem to fit the bill here.


Consider some of the alternative ways of eliciting constructs (see Section 4.2


(6) It was a real struggle to obtain more than four different constructs! Am I doing

something wrong?

No, not at all. It’s remarkable how few, genuinely different constructs a person

has on any one topic! (Alter the topic, of course, and you realise that we have

many more constructs than the ones previously offered; but that’s not the

point, since a grid is always done on a single topic.)

The following are usually helpful in eliciting further constructs. Firstly, you

might share this observation with your interviewee, and reassure him or her

on this point. Secondly, provide a short break, and have a brief chat about

something else for a minute or two. Particularly, try one of the alternative

construct-elicitation methods (see Section 4.2).

It all depends, of course, on the individual and the topic involved. In my own

work with managers on occupational questions, I find that I usually get

between 7 and 12 constructs during a 1-hour session with each manager.

On very rare occasions (one grid in a hundred or thereabouts), I’ve met

someone who blocks completely. This wasn’t their fault, and they weren’t

being dim! For them, the grid is a very contrived and artificial way of looking

at things. They simply don’t see that more than one or two constructs are

relevant to the topic, and are rather uncomfortable with the notion that I’m

obviously expecting more. In that situation, abandon the grid. It’s useless for

your purpose of understanding the other person in his or her terms. Find some

other way.

If you think about Kelly’s idea of ‘constructive alternativism’, you’ll recognise this is

bound to happen at some stage! Personal construct theory asserts that we have our

own personal theories ^ ways of making sense of experience ^ and that these may

well differ fromperson to person.This is not said lightly. It’s a fundamental epistemological

position, and hiswhole theory depends on it. It follows that personal construct

theory and approaches based on it may themselves be found useful by somepeople,

but not by others ^ in this respect, there is no difference between a personal theory

and an ‘official’ theory; between a theory held by a ‘scientist’ and one held by an

ordinary Joe or Jane.

Some of these other people (who find constructivist theories implausible) are called

positivists.Theirepistemologicalpositionisthat thereisa truth‘out there’whichexists

independently of the people who search for it, and therefore one should feel uncomfortableif

someone doesn’t see thingsthe samewayasagroup of scientistswhohave

spent yearsofdedicated effort discoveringwhat’s‘really’goingon.There’sanarticleby

Rom Harre¤ which puts this, and the constructivist alternatives, very well; see Harre¤


Whatever the reason for it, if your own interviewee doesn’t see things your way, as a

constructivist you have to treat this with respect. Drop the grid and seek some other

way of understanding the person.

(7) Halfway through eliciting a construct, it emerges that the interviewee is being

repetitive, offering a construct which we’ve already elicited and rated earlier; what do I


Share this impression with the interviewee, ask him or her if your impression

is correct – drop the construct and move on to a new one if it is, but continue if

s/he says that it isn’t.

Listen to what s/he says, and be particularly alert to the possibility that it isn’t

quite the same. In any case, the ratings will show whether the construct is

identical or not; if it’s the same, the ratings will be the same. If in doubt,

include it.

(8) Earlier on, you said that I should aim to get around 7–10 constructs, ensuring that

each one represented a clearly different way of construing the topic. But the

interviewee mentioned so many slightly different things that I could easily have

written down 20 or 30 constructs! You know: we were talking about a friend being

‘trustworthy’ as opposed to ‘unreliable’, and the interviewee said that this involved

being able to keep a secret, and never lying to you, and supporting you when people

unjustly criticised you, and, and, and . . .’ Should I turn each of these into a construct

and then rate each element on these different constructs? How do you decide when a

construct is clearly different?

It depends on your purpose. If the idea is to understand the interviewee in an

overall sense, or to see how s/he understands the topic as a whole, you’d be

inclined to aim for fewer, very distinct constructs. If, on the other hand, your

purpose is to obtain a detailed view (for example, if you’re studying a person’s

special expertise, in depth, as part of an attempt to capture their tacit

knowledge in a knowledge-capture application), you’d aim at more, and live

with any potential overlaps. Think of the first situation as one in which you’re

using a map of the whole country, where you’re interested in the main rivers

and mountain chains; and the latter as one in which you’re looking at a region

in much greater detail. Both kinds of map are useful, depending on your


One way of deciding in any one instance is to ask whether the ratings of

elements on the various slightly different constructs are likely to be the same,

or very different. You’d certainly include the different aspects if they lead to

very different ratings, since the interviewee is clearly offering you different


If in doubt, ask the interviewee: ‘Is this a different way of looking at things;

how important to you is it that I should include it?’

As always, if in doubt, ask.Kelly would say: ‘If youwant to know what the other fellow

thinks, why not ask them? They might just tell you.’ (What Kelly actually said was, ‘If

you do not know what is wrong with a person, ask him; he may tell you’; Kelly,1955/

1991: 322).You should know that George Kelly was a clinical psychologist, concerned

to understand his patients’distress in their own terms.This proved to be a very useful

ideawhen dealingwith other people in general, not just when they’re ill, or in distress,

which, of course, iswhy you’re reading this book!

(9) What happens if a construct being offered seems strange, unlikely, or plain

factually wrong?

For someone used to other forms of interviewing, or to psychological testing

even more so, the idea of accepting the interviewee’s own terms may seem a

little peculiar at first. Some of the constructs may seem very idiosyncratic to

you, weird and strange. The implications s/he’s drawing may seem muddled,

wrong-headed, or indeed, factually incorrect. Never mind; accept them.

This idea of credulous listening (e.g., Jones,1998: 346) is central to grid technique.

There may be situations in which (if you’re using the grid as part of a counselling,

guidance, personal development, or team-building programme) you might want to

come back to the matter later, to discuss and perhaps challenge the reasoning


But this is not the place for it.During a grid, accept what the interviewee tells you.Any

debate or negotiation should be about clarifying what the individual means,

regardless of whether or not you agreewith it.

(10) Okay, but what if some of the constructs are a bit trivial, even when taken with

respect to the interviewee’s position? For instance, in the practice grid, which was

about constructs to do with friends, my interviewee half-heartedly offered ‘male as

opposed to female’ among a large set of fairly detailed constructs about what they’re

like as people; should I accept it?

It depends on three things:

(a) the purpose of the grid

(b) the context set by the other constructs

(c) whether the interviewee feels that it’s trivial.

There will be situations in which you’ll both agree that a person’s sex is

probably irrelevant and uninformative – for example, topic: job activities;

elements: named employees. So ignore it and move on to another construct.

There will be situations in which you’ll both agree that it’s to do with gender

role as well as sex, and so the person’s sex, as a construct, is relevant and

informative with respect to that other construct, gender! – for example, topic:

attitudes to housework; elements: married friends known to me.

And there will be situations in which sex is socially and politically very

relevant indeed – for example, topic: how a boss construes the work of

employees in the department, carried out in the context of an appraisal for

promotion purposes.

There’s more on this in Section 5.3.3, under the heading of ‘Propositional

Versus Constellatory Constructs’.

So, ask the interviewee! Certainly, if you both feel the construct is a bit clicheґd,

remember to use a qualifying phrase (Section 3.2.1): ask ‘in what way?’

Seeking a more operationally defined expression for a construct is an excellent

way of going beyond clicheґ.