Psychological tests and the borderline diagnosis.

Contrary to recent optimistic claims there is no solid evidence to support the hypotheses that (a) a Borderline's Rorschach is more disturbed than a schizophrenic's, or (b) a Borderline presents disturbance on unstructured tests but not on structured tests. The group research clearly refutes the first hypothesis. The supposedly supportive research for the second hypothesis is either irrelevant, extremely weak, or refutable. Future research needs to explicate the subject sample's symptom, defense, and demographic characteristics, to estimate the base rates of the syndrome and hit rates of the test pattern in order to calculate the conditional probability of being Borderline given the test pattern, and to use a multimethod, multidiagnostic model of validation.