THE VALIDITY OF THE CROSSOVER-EFFECT IN R EACTIONTIME-STUDIES OF SCHIZOPHRENIC-PATIENTS

Attentional dysfunctions have lang been considered to play an importan t role in the pathogeneses of schizophrenic disorders. One of the most often replicated findings is the ''crossover effect'' (COE) in the se quential reaction time performance of schizophrenic patients. In recen t years the validity of the COE has been criticized. It was argued tha t it does not indicate an attentional dysfunction intrinsic to schizop hrenia, but is merely a result of procedural aspects of the experiment al set-ups used. This paper describes the first rigorous experimental test of this hypothesis. It was hypothesized that if the frequencies w ith which preparatory intervals (PI) were preceded by longer vs, short er preceding preparatory intervals (PPI) are held constant in the irre gular series, the COE would vanish in contrast to the classical experi mental task with unbalanced PPI-PI sequences. Sixteen schizophrenic in patients, 16 non-schizophrenic psychiatric inpatients and 16 healthy c ontrols were tested. Results confirm the hypothesis in two ways. First , in the classical procedure, a COE was observed in all three groups, although to a smaller degree in healthy controls. Second, with balance d PPI-PI relationships the COE pattern could no longer be identified, i.e., reaction time performances under regular and balanced irregular trial series were almost parallel in each of the three groups. The que stion is discussed whether the widely recognized COE in the sequential reaction time performance of schizophrenic patients must be re-interp reted as an epiphenomenon of unbalanced PPI-PI sequences.