Index Analyses of the Grid

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 

Another possible next step, which may, depending on your purpose, replace

describing the structure in the grid, is to summarise the way in which the

interviewee construes, by means of one or more of several indices. They won’t

be obvious to your interviewee, who is unlikely to have any sense of

ownership, but they can be helpful to you. They depend on the structure

within the ratings of the grid, and say something about the interviewee’s style

of thinking in this one grid. A comparison with other people is implied.

There are many such indices, each summarising some structural characteristic

of an individual grid. They can be broadly classified as follows:

. cognitive complexity: the number of constructs about a topic; the number of

constructs being differently used (functionally independent construction: see

Landfield & Cannell, 1988); the number of distinct construct clusters

(articulation). Burleson et al. (1997) provide a review of measures and an

interesting example.

. extremity scores of various kinds, such as ordination, integration, and

constriction, which examine the proportion of extreme as distinct from midpoint

ratings.

These measures are used in counselling and therapy, where they are likely to

be just one of many information sources which the interviewer has about the

interviewee. I don’t cover them herein. That’s because I’m wary of making

statements about an interviewee (‘he has a cognitively complex way of

thinking about this issue; she construes in a differentiated way’) on the basis of

just one grid elicited from that person, with no other information or personal

history in support. My guess is that you’ll be doing single grids with

individuals without necessarily getting to know them deeply, using other,

supporting, sources of information. Should this change, you’ll be able to find

out more in the references I listed above. For further particulars, look in

Fransella et al. (2004, Chapter 5) in the first instance.

Before any of this, though, a word about some basic assumptions.