12.01.2015 Views

RESEARCH METHOD COHEN ok

RESEARCH METHOD COHEN ok

RESEARCH METHOD COHEN ok

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

PROCEDURES IN GRID ANALYSIS 441<br />

Box 20.5<br />

Difference score for constructs<br />

Construct<br />

1<br />

2<br />

−2<br />

1<br />

−2<br />

Construct<br />

2 15<br />

Chapter 20<br />

15<br />

ΣD<br />

Constructs Match Difference score<br />

1–2 6 6 − 8 =−2<br />

By matching Construct 1 against all remaining<br />

constructs (3 ...15), we get a score for each<br />

comparison. Beginning then with Construct 2,<br />

and comparing this with every other construct<br />

(3 ...15), and so on, every construct on the grid<br />

is matched with every other one and a difference<br />

score for each obtained. This is recorded in matrix<br />

form, with the reflected half of the table also<br />

filled in (see difference score for Constructs 1–2<br />

in Box 20.5). The sign of the difference score is<br />

retained. It indicates the direction of the linkage.<br />

A positive sign shows that the constructs are<br />

positively associated, a negative sign that they are<br />

negatively associated.<br />

Now add up (without noting sign) the sum of<br />

the difference scores for each column (construct)<br />

in the matrix. The construct with the largest<br />

difference score is the one which, statistically,<br />

accounts for the greatest amount of variance<br />

in the grid. Note this down. Now lo<strong>ok</strong> in the<br />

body of the matrix for that construct which has<br />

the largest non-significant association with the<br />

one which you have just noted (in the case of<br />

a 16-element grid as in Box 20.4, this will be<br />

a difference score of ±3 or less). This second<br />

construct can be regarded as a dimension which is<br />

orthogonal to the first, and together they may form<br />

the axes for mapping the person’s psychological<br />

space.<br />

If we imagine the construct with the highest<br />

difference score to be ‘kind–cruel’ and the<br />

highest non-significant associated construct to be<br />

‘confident–unsure’, then every other construct<br />

in the grid may be plotted with reference to<br />

these two axes. The coordinates for the map<br />

are provided by the difference scores relating<br />

to the matching of each construct with the<br />

two used to form the axes of the graph.<br />

In this way a pictorial representation of the<br />

individual’s ‘personal construct space’ can be<br />

obtained, and inferences made from the spatial<br />

relationships between plotted constructs (see<br />

Box 20.6).<br />

By rotating the original grid 90 degrees<br />

and carrying out the same matching procedure<br />

on the columns (figures), a similar map may<br />

be obtained for the people (figures) included<br />

in the grid. Grid matrices can be subjected<br />

to analyses of varying degrees of complexity.<br />

We have illustrated one of the simplest ways<br />

Box 20.6<br />

Grid matrix<br />

Confindent<br />

+8 Kind<br />

Unsure<br />

-8 0 +8<br />

-8 Cruel

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!