5.3.2 Eyeball Analysis

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 

Regardless of what other analyses you may have planned, your next step is to

read the grid as a whole, and familiarise yourself with what’s there. If the grid

is one which you have elicited yourself, and you’re familiar with it already,

you need to step back from the detailed elicitation procedure and provide

yourself with an overview. This is a simple description of what the grid

presents, before your very eyes and with nothing up your sleeve. An eyeball

analysis was what I drew on in Section 2.2, when I described the grid which

appeared in that section as Figure 2.1.

You should always carry out an eyeball analysis before doing anything else,

especially if:

. The grid is the result of an interview by another interviewer.

. It is the outcome of a self-interview.

. The grid was elicited by a semi-automated software package. (See the

comments on software in the introduction to Chapter 6.)

For the moment, it’s best to assume that you are focusing attention on this

single grid and this particular topic, and that you’re not intending to make

inferences about how your interviewee thinks in general.

Use the following six-step procedure.

 (1) What is the interviewee thinking about? Note the topic of the grid; if

there is any information available on any qualifying phrases which were used

during elicitation, note that, too.

(2) How has the interviewee represented the topic? Note the elements. If there

is any information available about the way the elements were agreed, note it.

(3) How does s/he think? Constructs! How many constructs were obtained,

given the length of the interview? As suggested in the account of process

analysis above, the number of constructs you get is related to the

meaningfulness of the procedure; by and large, you’ll get more constructs

about a topic

. which interests the interviewee

. in which the interviewee has special expertise

. which the interviewee has to confront frequently in his or her work

. which the interviewee has thought about before.

The obvious question: what are the constructs; that is, what constructs does the

interviewee use in making sense of the topic? Read both poles of all the

constructs and note the particular distinctions being made. Bear in mind that

these are constructions from a given point of view, so remind yourself of the

qualifying statement.

(4) What does s/he think? Element Ratings! Note the scaling interval used (a

5-point scale, a 7-point scale, or whatever). Look at the ratings. Is there

anything obvious about the whole matrix of ratings? Probably not, but check.

Are they mainly 1s and 5s (or 7s) with no finer shading? Are there lots of

missing ratings because certain constructs didn’t apply to certain elements?

Lots of 3s (4s, in a 7-point scale), which could mean the same thing?

Column by column (element by element) find out what is being said about that

element. How has it been rated on all of the constructs? What picture of that

element is being conveyed? Note, particularly, whether that element receives a

lot of 1s and 5s (or 7s), and on which constructs. Repeat for all the elements.

(5) Look at the supplied elements and constructs and ratings. Familiarise

yourself with the ratings given to each of the elements on any constructs which

have been supplied. Similarly, familiarise yourself with the ratings given, on

all the constructs, to any self elements, or ideal self elements, which have been

used. What, in this context, is being said about the Self? How does the

interviewee characterise his or her ideal? How do ratings on the supplied

construct compare with the others?

(6) Draw your conclusions. Summarise the main points of what you have

gleaned. Where you find yourself making interpretations of your own, you

should do so in the light of the process analysis you have already conducted. If

you’re working collaboratively with your interviewee, the process issues

should be discussed, as well as the substantive information derived from the

grid. And, in a collaborative analysis, the interpretation is, by definition,

collaborative: you both propose your viewpoint on what the significance

might be, and what it all means, and you both assess that viewpoint, as

critically as you decide it deserves. You negotiate a common understanding,

according to the stance I have outlined in the previous section.

And that’s all there is to the basic, first-step analysis procedure. If, at the end of

this procedure, your reaction is ‘so what?’, there may be something wrong.

Most people have at least something interesting about their construing! A

reaction of this kind may occur:

. because you’re tired and jaded, having looked at a lot of grids at the same

time. No problem: take a break.

. Because you’re working with a sample of interviewees. You’re working with

several sets of grid results, and it’s the type of investigation in which the big

hits will emerge only after you have completed the analysis of the whole set.

Be patient!

. Because you don’t quite know how you’ll continue the analysis of the grid,

but you’re sure you’ll think of something soon. That is a problem.

In the latter two cases, look at the material in Section 7.1.2 on design. But

usually all will be well. Onward.

Go straight to Exercise 5.2!