|
Chapter Twenty - Additional Resources
Box 20.1 : Eliciting constructs and constructing a repertory grid
A person is asked to name a number of people who are significant to him. These might be, for example, mother, father, wife, friend, employer, priest. These constitute the elements in the repertory grid.
The subject is then asked to arrange the elements into groups of threes in such a manner that two are similar in some way but at the same time different from the third. The ways in which the elements may be alike or different are the constructs, generally expressed in bi-polar form (quiet -talkative; mean – generous; warm – cold). The way in which two of the elements are similar is called the similarity pole of the construct; and the way in which two of the elements are different from the third, the contrast pole of the construct.
A grid can now be constructed by asking the subject to place each element at either the similarity or the contrast pole of each construct. Let x = one pole of the construct, and blank = the other. The result can be set out as follows:
CONSTRUCTS
|
ELEMENTS |
|
A |
B |
C |
D |
E |
F |
1. quiet-talkative |
x |
x |
x |
|
|
x |
2. mean-generous |
x |
|
|
x |
x |
|
3. warm-cold |
|
x |
|
|
x |
|
It is now possible to derive different kinds of information from the grid. By studying each row, for example, we can get some idea of how a person defines each construct in terms of significant people in his life. From each column, we have a personality profile of each of the significant people in terms of the constructs selected by the subjects. More sophisticated treatments of grid data are discussed in examples presented in the text.
|
Source: Adapted from Kelly, 1969
Box 20.2 : Allotting elements to constructs: three methods
Example 1: Split-half form
ElementsConstructs
1 2 3 4 5 6 7 8 9 10
x x x x x 1 fast-slow
x x x x x 2 late-early
x x x x x 3 dangerous-safe
Since the subject is forced to allocate half of the elements to one pole, the chance expectancy of matchings occurring on 10 elements when two constructs are compared is 5. Deviation scores can be computed from chance level. Thus 5 matchings = 0; in constructs 1 and 2, matchings = – 3; in constructs 1 and 3, matchings = +1; and in constructs 2 and 3, matchings = – 1. The probability of particular matching scores being obtained can be had by reference to statistical tables.
Example 2: Rank-order form
ElementsConstructs
1 2 3 4 5 6 7 8 9 10
10 1 2 5 8 7 3 4 9 6 1 fast-slow
9 4 10 1 6 8 5 2 3 7 2 late-early
7 9 5 6 10 2 1 4 8 3 3 dangerous-safe
Spearman’s rho (r s) Relationship scores
Constructs 1 and 2 = .15 (0.15) 2 x 100 = +23
Constructs 1 and 3 = .24 (0.24) 2 x 100 = +58
Constructs 2 and 3 = -.16 (-0.16) 2 x 100 = -26
Example 3: Rating form
Elements Constructs
1 2 3 4 5 6 7 8 9 10
4 4 2 1 4 3 5 1 5 2 1 fast
1 1 3 5 1 3 2 2 5 5 2 late
5 1 3 2 2 1 4 5 1 2 3 dangerous
A 5-point rating scale is shown in which, in this example, single poles of the constructs are rated as follows:
Not at Very much
all like Average like
½ ___________ ½ ___________ ½ ___________ ½ ___________ ½
1 2 3 4 5
Bannister and Mair suggest several methods for calculating relationships between constructs from the rating form (pp. 63-5). For a detailed discussion of measures of construct relationships, see Fransella and Bannister (1977, pp. 60-72).
Source : Adapted from Bannister and Mair, 1968
Box 20.3 : Laddering
Constructs |
Elements
teachers |
|
A |
B |
C |
D |
E |
F |
G |
H |
masculine |
2 |
1 |
5 |
4 |
3 |
6 |
8 |
7 |
serious |
6 |
2 |
1 |
3 |
8 |
4 |
5 |
7 |
good teacher |
|
|
|
|
|
|
|
|
authoritarian |
|
|
|
|
|
|
|
|
sexy |
|
|
|
|
|
|
|
|
old |
|
|
|
|
|
|
|
|
gets on with others |
|
|
|
|
|
|
|
|
lonely |
|
|
|
|
|
|
|
|
like me in character |
|
|
|
|
|
|
|
|
like I hope to become |
|
|
|
|
|
|
|
|
A matrix of rankings for a repertory grid with teachers as elements
You may decide to stop when you have elicited seven or eight constructs from the teacher elements. But you could go on to ‘ladder’ two or three of them. This process of laddering is in effect asking yourself (or someone else) to abstract from one conceptual level to another. You could ladder from man-woman, but it might be easier to start off with serious-light-hearted. Ask yourself which you would prefer to be – serious or light-hearted. You might reply light-hearted. Now pose the question ‘why’. Why would you rather be a light-hearted person than a serious person? Perhaps the answer would be that light-hearted people get on better with others than do serious people. Ask yourself ‘why’ again. Why do you want to be the sort of person who gets on better with others? Perhaps it transpires that you think that people who do not get on well with others are lonely. In this way you elicit more constructs but ones that stand on the shoulders of those previously elicited. Whatever constructs you have obtained can be put into the grid.
|
Source: Adapted from Fransella, 1975
Box 20.4 : Elements
Construct |
Self
|
1 |
2 |
3 |
4 |
5 |
6 |
7 |
8 |
9 |
10 |
11 |
12 |
13 |
14 |
15 |
Construct |
KIND
|
X |
|
X |
|
|
|
X |
X |
|
|
X |
X |
X |
|
|
X |
CRUEL |
CONFIDENT
|
X |
X |
|
X |
X |
|
X |
|
|
X |
|
X |
|
X |
|
|
UNSURE |
__________
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
__________ |
__________ |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
__________ |
__________ |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
__________ |
|
Box 20.5 : Difference score for constructs
|
|
|
|
|
Construct |
|
|
|
|
1 |
2 |
|
15 |
|
|
|
-2 |
|
|
|
1 |
|
|
|
|
|
2 |
-2 |
|
|
|
Construct |
|
|
|
|
|
|
|
|
|
|
|
15 |
|
|
|
|
Box 20.6 : Grid matrix
|
