The assessment of ego development in borderline personality disorders.

The purpose of the study was to assess levels of ego functioning in a cohort of patients who met the criteria for borderline personality disorder. The Sentence Completion Test (SCT) and the Descriptions of Significant Others (DSO) test were used to measure dimensions of ego maturity. The borderline patients' responses on the SCT were compared with the responses of an outpatient sample of psychiatric patients and with general population norms established for the test. The scores on the SCT for a selected subgroup of five subjects were compared with their scores on the DSO. Contrary to expectation, the responses on the SCT by those with borderline personality disorder did not differ from those for the psychiatric outpatient sample or the general population norms. However the within-group comparisons between the SCT and the DSO for the subgroup of five subjects showed comparable trends between tests. Discussion of the results focuses on the interpretations of the borderline patients' responses to the two tests. Emphasis is placed on the need to distinguish subsets of ego functions, which can be assessed separately using a variety of measurement strategies.

Object representations in descriptions of significant others: a methodological study.

The aim of this methodological study was to extend Blatt, Wein, Chevron, and Quinlan's (1979) coding system for assessing levels of mental representations of others. Blatt devised a system for scoring spontaneous descriptions of parental figures on a scale that spanned five levels of object differentiation. A single level score was assigned. In this study a procedure was developed to generate scores for each of five representational levels, for two object descriptions, for each subject. The new procedure was based on the assumption that in each individual, all levels of differentiation are maintained albeit in varying proportions. The new coding method was tested in a contrast study in which the responses of a cohort of borderline patients were compared with those of a cohort of nonpsychiatric subjects. As predicted, all subjects were assigned scores at all five representational levels; however, the borderlines had a higher proportion of content units in the lower, least differentiated levels of object representation, and the nonpsychiatric subjects had higher scores at the more differentiated, conceptually complex levels. Results of the psychometric properties of the coding method are discussed. Work in progress to refine the method is outlined.