Teach Yourself Repertory Grid: Part the Seventh, Analysis (One).

Teach Yourself Repertory Grid, Part the Seventh: Analysis (One).

I hope you enjoyed your homework … you should have generated some constructs (and maybe some insights) about careers, relationships, and motivation. I deliberately teased you a little, by leaving some loose ends – I hoped that some questions would form in your head, and in this posting I hope I can answer some of them.

One of the most important considerations when you are planning a Grid-based project is the question of where in the process you are going to find your most useful and insightful information – where’s the beef? I’ve already lamented the trend for rushing through an ill-thought-out ‘construct grab’ in haste to get some data that can be put through a computer program; you can actually do a good deal of work with Grid without going near a computer, and I personally wish that this stage played a larger part in the training of novice Grid practitioners – then I wouldn’t get so many questions saying ‘I’ve put it all through the computer but I still don’t know what it means,’ and I wouldn’t have to spend so much time thinking of variations on ‘If you don’t know where you’re going, you’ll probably end up somewhere else.’

So, and taking a guess at the experiences you may have had while doing the three suggested homework topics, let’s take a look at where in a Grid interview the important information is to be found. Of course, it depends on the purpose and therefore on the elements and the kind of constructs you elicit, but here are some typical – and simple – analyses:

1. The number of constructs generated.

Did you by any chance try the career choice interview on a passing teenager? Or on someone else who hasn’t much experience of the world of work? Because if you did, you might have been surprised at how few constructs – especially constructs that weren’t propositional – your interviewee could produce.

Or, if you missed that particular opportunity, try giving me some constructs about the following elements:

PRINT OF THE WAVE

TREE OF LIFE

TRINITY

Bet you can’t – unless you’re interested in the history of old Shetland knitting patterns, because they’re the names of three different patterns that womenfolk would knit into their men’s sweaters.

You can’t produce constructs about a subject about which you know nothing; and if someone can only produce a limited number of constructs – especially if they produce mainly propositional constructs – then you may, just may, be justified in coming to the conclusion that they haven’t much experience of the subject.

Obviously, this is a very coarse measure indeed, and you’d only be justified in taking any notice of it when you can be absolutely certain that your subject was comfortable with the process, that you’d tried several approaches to eliciting constructs, and that there really was nothing else left to say. But it can be a useful measure, especially when you want to compare and contrast – for example, I used a simple construct count as a before-and-after evaluation of a training course, because the least you’d expect after a successful training event is that the person has more constructs than they started out with; another example is a study of the construct systems of venture capitalists, where the least successful of the sample had less than half the number of constructs of anyone else.

Be careful with this measure. It’s more useful when used to measure difference than to draw conclusions from one single measure, and if you have to use statistics to ask whether the difference matters, you shouldn’t be asking the question – the contrast should hit you in the back of the neck. But in the appropriate circumstances, it’s a useful measure in its own right and can point you in the next direction you want to examine.

2. Content analysis of constructs.

The practice session on relationships ought to have given you an insight into the importance of analysing the actual content of the constructs, especially if you made the opportunity to interview one or two other people as well as yourself. Did you notice any particular themes emerging? and did you extract different themes from different people? I’d be very surprised if you didn’t – if not in the first round of constructs, then at the point when you started laddering up.

Remember, the Grid interview lets you eavesdrop on the language your subject has developed in order to navigate their way around the world. So if your subject felt that they’d had their trust broken rather too often, and perhaps found it hard to judge whether someone was trustworthy, you’d expect a lot of constructs that related to how trustworthy their people were. I’ve conducted interviews in which money and worldly success accounted for a good half of the subject’s constructs, and interviews where these didn’t figure at all; interviews in which religious belief and practice were important, and interviews where religion got no mention, and so on.

Again, before you think about analysing the results, you need to be certain that you’ve given the available constructs the best chance of coming to the surface (and I’ll say more about this in a later posting) and you should look for broad themes rather than fine distinctions.

I frequently use content analysis of constructs in projects where a group of people will be interviewed and their results pooled. Mostly I’ll be working on an organisation change project, where the first task is to determine the present culture of the organisation – in order to ask the question ‘how does it need to change?’ Mostly I’ll be asking managers to give me constructs about their colleagues and subordinates (whose identities are kept anonymous – I’m only interested in the constructs). I showed you a number of examples of this sort of study in the third posting, so I’ll not repeat myself here. What I do need to say is that when you’re doing a content analysis of the constructs, one of the most important skills to develop is that of recognising what’s been left out – observe that money hasn’t been mentioned in the constructs about your relationships ... observe that none of your managers mention innovation when thinking about their colleagues ... etc. Unfortunately there isn’t a Universal Directory of Constructs where you could look up the answer to the ‘what’s missing?’ question – though I will try to help you where I can – but let me say that if you’re aware of the need to develop this skill, and you practice before going ‘live’ on a new topic, then the skill usually develops at a good deliberate speed. (The only circumstances where it won’t are if the interviewer doesn’t listen; Grid interviewing is a great way to improve one’s listening skills).

3. Responses to Element Creation Questions:

Remember the example of using Grid to examine the reasons why doctors chose their specialties? There I used a set of element creation questions – asking for preferred and non-preferred specialties, etc. – and there were patterns in the answers to those questions, patterns which could themselves be analysed for useful information before we even got to construct elicitation. That’s one example of an application where the ‘beef’ begins to emerge in the very first stage of the interview.

The third practice session I gave you – on motivation, where you were asked to think about the high points and the down bits and the plateau times – was an example of what we call an unrehearsed element set. That is, you’re asking questions which most people may not formally have asked themselves before, and it’s unlikely that the responses will come tripping off the subject’s tongue. I don’t know whether you’ve ever been asked to draw an ‘emotional graph’ of your life-span, with its ups and downs – that’s a technique often used in counselling, and just looking at the highs and lows can be instructive even before you start asking for the reasons.

As a rough generalisation, unrehearsed element sets are more likely to be events or activities than things or people; usually, we’ve thought about the people in our lives and many of the ‘things’ that we have to choose between, so the ‘beef’ comes when you start to elicit constructs, or when you go on to laddering – or maybe even later. When you’ve had more practice in designing Grid sessions, I hope you’ll come to realise that many interview purposes can be achieved in two or three different ways, with different element sets: for example, you can do a very interesting ‘help me choose a car’ interview by creating element creation questions that single out various kinds of ‘critical incident’ while driving – now there’s an unrehearsed element set for almost anyone except a driving instructor, and if you feel like some more homework why not give that a try?

4. Using the constructs to create some new elements.

If you go back to the constructs you obtained in the career choice exercise, you could start to draw a profile of MY IDEAL CAREER or THE CAREER MOST LIKELY TO PROVE DISASTROUS, by deciding which pole of each construct you prefer and how strongly you feel about it. Then you’re part-way to preparing an aide-memoire to help you – or your interviewee – ask useful questions about potential job choices, or to understand why you might feel like a square peg in a round hole if you’re unhappy in the present job.

If you’d like to try something potentially more powerful, more rewarding, go back to your session on relationships and use the constructs to create some new elements. For example, if you’re in the habit of falling for people who turn out to be bad for you, try creating and rating the element I FALL FOR AND IT ENDS IN TEARS, or however you feel like phrasing it … one woman who did this had a huge Aha! moment and I realised that I had a habit of falling for the lonely hero whom only I could understand and accompany on his quest, but when the chips were down I’d be pushed aside … I require the Universe to bring me a thoughtful pipe-smoking historian who’ll absolutely adore me. Or there was the chap who realised that I keep falling for women who like being spoiled, treated with old-fashioned courtesy, made a fuss of … except that there are times when I want to be spoiled and given a treat, and it didn’t happen.

Experiment with various ‘created’ elements, getting a sense of how to create an element or elements that will be informative – for example:

WHAT S/HE SEEMED LIKE WHEN WE MET and WHAT S/HE SEEMED LIKE WHEN WE PARTED (where s/he represents a special relationship that went wrong);

MY IDEAL RELATIONSHIP;

THE RELATIONSHIP MY PARENTS WANT ME TO HAVE;

and so on. I’ll say more about creating elements part-way through in a later posting. I want to finish by raising a question, thus:

Who’s doing the analysis anyway?

So far, I’ve rather given the impression that the interviewer also does the analysis – look at what’s been generated in the interview and draw some conclusions from it. In fact, it’s much better if the interviewee is involved in the analysis also: better for the interviewee’s comfort, better for their trust in the process, but – crucially – because it’s their information they’ve generated and they’re interested in it, maybe more likely to see what’s important, to get involved. In fact, there’s a cardinal rule in good grid practice: Don’t construe other people’s construing, meaning ‘don’t you go drawing conclusions from the results – ask your interviewee.’ I’ll talk some more about feedback, ownership, and interviewee involvement in the next post – for the moment, may I just plant the seed that Grid isn’t something you do to people, it’s something you do with people.

Love,

Valerie.