﻿

# 4.1.3 Questions About the Rating Procedure

К оглавлению

(11) Does it matter if I combine steps 8 and 9 of the basic procedure? In other words,

should I always get the ratings of the elements in the triad first, before getting the

ratings of the remaining elements, on a particular construct? Wouldn’t it be simpler

just to run along the row each time, from the first element to the last element, rating as

I go?

If that feels more comfortable, fine. I do find that, for the first few constructs,

until the interviewee gets used to the procedure, it helps to record the ratings

of the particular triad first, before completing the rest. It seems more logical to

the interviewee. And it makes it easier for you to check that the interviewee

hasn’t reversed the ratings unintentionally.

Let’s do a deal. Until you get used to grids, always do step 8 before step 9 for

the first two constructs you elicit. Then for the remainder, try it both ways and

see how it feels, before deciding for yourself which suits you better.

(12) I got a bit tangled up with the ratings. Do you insist on doing step 5 that way

round? What I mean is, should I always put the thing the two elements have in

common on the left of the scale, and the opposite, which characterises the odd one out,

on the right?

When my interviewee says, ‘John and Mary go together because they’re both nervous,

while Alan is the opposite: he’s very self-confident’, it seems natural to put ‘nervous’

down on the left and ‘self-confident’ on the right – fair enough.

But if he were then to say, ‘John is a bit rigid in his thinking; in contrast, Mary and

Alan are alike because they’re very liberal’, it’s more natural to put ‘rigid’ on the left,

where it gets a rating of ‘1’, and ‘liberal’ on the right, where it gets a ‘5’. Shouldn’t the

rule be, ‘give the higher rating, ‘‘5’’ not ‘‘1’’, to the end of the construct which is

‘‘better’’, or ‘‘positive, more valuable’’ ’, d’you see what I mean? It was certainly more

natural for the interviewee!

That may feel right when a construct has a clear positively evaluated, and a

clear negatively evaluated, end: ‘nasty versus nice’, ‘evil versus good’. But not

all constructs do. For example, if you were to characterise a person as ‘socially

skilled’ and another, in contrast, as ‘technically skilled’, then, other things

being equal, which is better? Neither, surely. It all depends.

Then there will be times when one end of a construct is slightly better, but only

a little, and vice versa; and this, believe me, can lead to enormous confusion

when you start to analyse constructs, and the way in which they relate to each

other.

And so, trust me! I’m a psychologist! Write down the end of the construct

which describes the ‘two that go together’ on the left; and the contrasting end

of the construct, which describes the ‘odd one out’, on the right.

Always.

Regardless of ‘which way round is positively evaluated’.

By the way, for future reference, the end of the construct which describes the

two elements that are alike is called the ‘emergent pole’ of the construct; and

the end of the construct that describes the odd one out is called the ‘implicit

pole’ of the construct. So the rule is ‘emergent on the left, and implicit on the

right’.

(13) Earlier on, you talked about ‘credulous listening’. But what happens if you

suspect that the interviewee has made a genuine mistake in trying to express him- or

herself? In doing the practice grid, I had a situation in which, bearing in mind

everything else that the interviewee had said, I was sure she’d got the ratings the

wrong way round, saying ‘5’ when she really intended to say ‘1’. Do you really mean

I’ve got to accept this, ‘credulously’, as you put it?

No, not at all. There’s credulity, and there’s carelessness.

As step 8 of the procedure tells you, if you suspect this is the case, just check

what’s intended. Point out that the ratings on this construct seem to be ‘the

wrong way round’. In the vast number of cases, your interviewee will

recognise that they’re not communicating their intention. Take their word for

it either way!

Just to drive the point home. If the elements in a grid on politicians included

‘Churchill, Hitler, and Stalin’ and the construct was

1 5

‘Used indisputably evil means – ‘May have dealt with opposition

of dealing with opposition’ robustly, but not by evil means’

and you were offered ratings of Churchill: ‘1’, Hitler: ‘5’, and Stalin ‘4’, you’d

usually have grounds for checking that the interviewee hadn’t reversed the

directionality of the rating scale, intending to say Churchill: ‘5’, Hitler: ‘1’,

Stalin: ‘2’. All the more so if the construct was

1 5

Took the initiative in 1939 – Reacted to others’ initiatives

and the ratings are Churchill: ‘1’, Hitler: ‘5’, and Stalin ‘4’, where the factual

incorrectness, rather than attitudinal idiosyncrasy, of the response alerts you

to the possibility of an error of intention.

But if, after checking, the interviewee replied that the ratings were correct as

s/he originally stated them, you’d have to accept them, as we said earlier.

As Kelly puts it,‘People differ fromeach other in their construction of events’, and that

may includematterswhichweusually think ofas‘facts’.To be an effective grid user, it’s

best to accept allfactsasbeliefsabout nature, and to seeaccuracyassolely to dowith

a faithful recording of the interviewee’s beliefs. For example, should an interviewee

choose to deny the Holocaust, asserting it to be a historical fiction, it is your professionalobligationto

grit your teeth and accept this as an accurate statement of theway

s/he expresses a construction of events.You can always explore the intention later.

This lies at the heart of any constructivist technique.

(14) What happens when a construct applies to some of the elements but not to the

others? Working with the grid you suggested as Exercise 3.1, my interviewee said that

although the construct, ‘their family always welcomes me – their family’s a bit distant

with me’, was suggested by a particular triad of elements (friends), that construct

doesn’t actually apply to one of the other elements. He said that one of his friends lives

by himself, the parents being deceased – so how his parents get on with the interviewee

isn’t in question. What do I do? I need to have numbers to be able to analyse the grid

later on! So should I provide a rating or not?

as precisely as possible, and you’d be distorting the intended meaning by

having a rating here. Leave it blank.

If you really, really, really have to have a rating in each cell of the grid (this

may be the case with some analysis software you plan to use, though good

software will be able to cope with missing ratings), then the least damage

would be to put in a ‘neutral’, mid-point rating (‘3’ on a 5-point scale, or ‘4’ on

a 7-point scale).

In that example, there was an obvious indication that the construct didn’t apply:

one of the friend’s parents was no longer alive! But it isn’t always so clear-cut.

Sometimes you can encourage the interviewee to reword the construct in a way

which applies to all the elements while expressing the idea s/he has in mind.

For example, you could ask whether the construct

Raised in a sociable and – Raised in a family which was

welcoming family cautious with outsiders

might be a satisfactory substitute for the construct, ‘their family always

welcomes me – their family’s a bit distant with me’. It would certainly allow

the interviewee to provide a rating for all the elements! But it mightn’t be what

s/he wishes to say at that point. Always ask.

Finally, when there’s no obvious indication that the construct doesn’t apply, it

might be that your interviewee is struggling to express a complex idea which

should really be split up into two constructs. Check that your interviewee’s

construct expresses the contrast s/he has in mind, rather than, in fact, being

made up of two separate contrasts. For example, the construct

Their family always welcomes me – Their family’s a bit distant

could, actually, stand for two distinct things your interviewee wants to tell

you, needing to be decomposed into

Their family always welcomes me – Their family is a bit distant with me

and

Their family is very sociable and – Their family’s a bit distant (with

welcoming other people)

for the interviewee’s meaning to be captured completely. Someone who was

seen by your interviewee as having parents who were unwelcoming to the

interviewee though they were very sociable towards other people would be

difficult to rate on the construct as it was first offered, but could easily receive

a rating on each of the two component constructs.

So you’d check with your interviewee whether the original construct needed

to be decomposed into two and, if so, you’d write both into the grid, instead of

the original, and rate all the elements on each construct.

Underlying this bit of practical advice there are two fairly deep theoretical issues.The

first has to do with the structure of a person’s construct system, and the relative

centrality of the particular constructs being offered. As we’ll see in Chapter 8,

people’s constructs are hierarchically organised: some are central, and superordinate

to others; other constructs are subordinate to others. Now, constructs

which are central tend to be applicable to awiderange of elements, andwhen decomposed

intomore detailed aspects, their relevancemay bemore specific.More on this

later, in Section 8.1.

The second issue has to do with how clearly and precisely a person is helped to

express the distinctions and contrasts they have inmind, when choosing thewording

ofa construct.Thisparticular issuewasexploredin somedetail in a famousexchange