Cognitive Measures

Although the repertory grid has generated various ways that purportedly measure certain cognitive dimensions, it is worth clarifying some terms right at the beginning of this chapter. First, the term "cognitive" is used in its broadest sense, as Kelly (1969) rejects the notion that personal constructs are merely "cognitive." This is because he regards the "cognition-emotion" distinction as being of little use to psychological science, preferring to move towards a more holistic orientation which views behaviour, thought and emotion as the result of a process that involves the construction of meaning (see Mancini & Semerari, 1990).

Second, the postulation of cognitive measures responds more to the need to define trait-like variables such as "cognitive complexity," "identification," "self realisation," etc. than a reflection of the phenomenological spirit of the theory. The GRIDCOR programme has incorporated some of these indices with the intention of increasing the idiographic potential of the grid by allowing for intra-subject and inter-subject comparisons. This selection is based on their clinical value and on the support found in the psychological literature. However, and mainly because the repertory grid is a non-standardized test (due to the large variety of grids and grid components), there are no tables with which to make generic interpretations as there are with other tests having fixed content.

Third, we would join Fransella and Bannister (1977) in their call for caution regarding the proliferation of measures, different ways of calculating them, etc., and the tendency for some measures to lack clear psychological meaning. In this sense, we have tried to choose the most conceptually solid measures for inclusion in the GRIDCOR programme.

Measures of Self-Construing

Various aspects of the self can be evaluated using repgrid data. The GRIDCOR programme gives a simplified output of the original data which summarises, in a clinically relevant manner, some of the responses provided by the subject.

Self-definition and Self-Ideal discrepancy

An option of the GRIDCOR programme shows the constructs that are relevant to the self (Table 6). The self-definition column shows a list of those constructs that have extreme scores loading in the direction of the "SELF" element as well as their opposites. This enables us to see which are the most representative constructs of how the subject sees himself. Daniel, for example, sees himself as "anxious," "with no sex drive," "unmanly," "dreamer," "irresponsible," and "intolerant" (this implies that he does not see himself at all as "relaxed," "happy," "responsible," etc.). The column to the right shows the perceived discrepancies between the self and the ideal. This column includes those constructs in which the difference in scores given to the SELF and the IDEAL elements implies a change in the direction of the construct pole (i.e., a minimum of four points of difference in a seven interval scale). This column not only indicates that Daniel sees himself as "anxious" but also that he would like to be "relaxed." There are also other qualities that he would like to have, such as: "adjusted," "responsible," "understanding," and "happy." In the therapeutic context, these constructs describe the changes that the person would like to make and, therefore, can be considered important "target goals" for therapy.

Self-Ideal differentiation

The SELF-IDEAL correlation, which is simply the product-moment correlation of these elements, can give purely quantitative information on any discrepancy. Although this coefficient can be found in the correlations table between elements (already described in the previous chapter), GRIDCOR gives a summary of some important general indices (Table 7). For those who prefer distances, the distance between SELF and IDEAL can also be found at the distance matrix for elements.

Although Norris et al. (1976) relate it to "self-alienation," the SELF-IDEAL discrepancy can be considered a measure of self-esteem. The SELF-IDEAL correlation gives us a quantitative evaluation of how respondents value themselves in their own terms, as opposed to more traditional self-esteem scales which score the subject according to items previously selected by the investigator. At the moment, there are a growing number of studies that point out the importance of self-esteem for effective functioning and a subjective sense of well-being. Low self-esteem has been associated with a great deal of psychological disorders (see Robson, 1988). Daniel presents a very high negative correlation (-0.84) between SELF and the IDEAL which is indicative of a very low self-esteem and is coherent with his symptomatology and discomfort.

Self-Others Differentiation

In PCT, the construction of the self is intrinsically related to the construction of others:

"The way in which the self is construed must necessarily be the way we construe others. This is because we do not have a self concept as such, only a bipolar 'self-not self' construct" (Bannister & Agnew, 1977, p. 99).

For this reason, the processing of information about others also involves a degree of automatic self-comparison. Therefore, the differentiation between the SELF and OTHERS is of particular relevance as a central feature of interpersonal construing. Adams-Webber and colleagues have carried out an extensive series of experimental studies (see Adams-Webber, 1990 for an overview) in which a psychological constant was found in the way the self and others were construed: on average, samples of normal adults place elements on the same pole as the self in 62% of their ratings. This proportion, which coincides with the "golden section" (a proportion considered aesthetic by the Greeks and ideal by information theorists) has been replicated in several international samples. A developmental trend has also been observed in this ratio, showing that it settles at 62% when adolescence is reached.

The finding (replicated in experimental and clinical studies) that depressed subjects show a greater SELF-OTHERS differentiation than normal subjects (see Neimeyer, 1985, for a review) is of significant clinical interest. The reverse of this differentiation score has been labelled "identification" by some authors (Jones, 1961), but we consider it to be a measure of perceived social isolation. Norris et al. (1976) support a similar view when suggesting that this measure could be used as an indicator of how lonely the subject feels. Likewise, Harter, Neimeyer and Alexander (1988) have reported that sexually abused women display greater SELF-OTHERS differentiation than control subjects.

In the GRIDCOR programme, the differentiation between the self and others is calculated by creating an artificial OTHERS element which is the result of averaging the scores of all but the SELF and the IDEAL elements. Just as with the discrepancy between the SELF and the IDEAL, the differentiation between the self and others can be seen in the distances and correlations matrices (distance SELF-OTHERS). The GRIDCOR programme gives the SELF-OTHERS correlation right after the SELF-IDEAL correlation appearing in the summary of cognitive measures. In Daniel's case, we can see that this correlation is negative but moderate (r = -0.262), which indicates that a slight social isolation is perceived. This means that he construes himself as somewhat different from others.

Ideal-Others Differentiation

Although less frequently used in the psychological literature, we consider the correlation between the IDEAL and OTHERS to be a measure of perceived adequacy of others. For example, while a high negative correlation can indicate that the subject is dissatisfied with the people that surround him/her, a positive correlation suggests a perception along the lines of "everyone is great." In Daniel's case, others are also rather distant from what he considers to be the ideal (r = -0, 176, see general indices). Beck and his colleagues (1979) regard such negativity about the social world to be one of the hallmarks of depression.

From a clinical perspective it is worth stressing that the two indices that contain the IDEAL element (correlation SELF-IDEAL, correlation IDEAL-OTHERS) can suggest opposite but complementary lines of action. A substantial difference between the IDEAL and SELF/OTHERS could indicate the need to improve self-esteem as well as how others are perceived. However, it could also indicate an excessive and over-demanding IDEAL which would be very difficult to satisfy realistically. If such a demanding ideal appears, it is usually positively and highly correlated with a particular, idealised, element which may prompt useful therapeutic discussion.

Tentative self-construction profiles

In accordance with the available data and with our experience, a series of general self-construction profiles can be identified based on the three indices just explored (Table 8).

In the way of a hypothesis, we suggest that if subjects with this profile also present a positive IDEAL-OTHERS correlation, this would correspond to the depressed person who has a negative self-perception and also believes that he/she is the only one who is ill/bad. This would indicate a double tendency in information-processing terminology: information relative to the self is biased negatively, whereas information relative to others is biased positively. Under such conditions, the subject may show considerable dependency on others or be preoccupied with fantasies of being "rescued" by more poweful or competent others.

If, on the other hand, the IDEAL-OTHERS correlation is negative, the person is likely to be a depressed person who thinks that others are equally maladaptive. This suggests that he or she may relate his or her depression or despair to the attidudes held by others. If this is the case, clinical exploration may reveal feelings of anger and resentment towards others or, more specifically, towards a particular family member.

A note of caution is worth mentioning after having described these profiles. They should be taken as merely indicative and are at their greatest descriptive value when the correlations (positive and negative) are high enough. They would also require a more thorough investigation if they were to be considered something more than a working hypothesis.

Daniel's scores on these three measures come close to a depressive isolation profile, although this is obviously not his clinical diagnosis. If symptoms of depression are present in Daniel, they are likely to stem from his incapacity to deal with his panic attacks, which he perceives as uncontrollable.