LSD response in Eysenckian trait types identified by polypredictive CFA

Earlier research has related the two major Eysenck dimensions, extraversion/introversion and neuroticism, to individual differences in drug responses along arousal–sedation and affective continua. The investigators note that looking at these two dimensions simultaneously can reveal trait combinations that predict differing pharmacological sensitivities; drugs acting on serotonergic systems, such as LSD-25, have been used experimentally to probe such personality-linked response patterns. A fortunate pre-existing dataset collected by Kohnen and Lienert (1987) provided observations of three binary LSD-elicited symptoms — blackouts, thinking disturbances and affective reactions — which the present paper re-analyses to test trait–response associations.

A further modification, termed the 'chessboard' CFA, collapsed the contingency structure to a square arrangement that maps four personality cross-classifications directly onto four response configurations, allowing a tetraprediction test that simultaneously assesses all four hypothesised personality–response correspondences. Where numerical formulas or intermediate table labels appear in the extraction, they are partially garbled; the description above reflects the analytic logic as reported. The extracted text indicates group cell counts were available and used in the CFA procedures (examples below).

Using the chessboard modification that collapses certain symptom distinctions, a one-to-one mapping was obtained between the four personality cross-classifications and four response configurations: hysterics → T+ A+ (schizoaffective-like), stable extraverts → T- A+ (affective changes), dysthymics → T+ A- (thinking disturbances), and stable introverts → T- A- (non-responding after excluding blackouts). The tetraprediction (four-fold) chessboard CFA was reported as statistically effective, although some numerical thresholds in the extracted text are garbled and not fully reconstructible from the extraction.

The discussion situates these results within ongoing debates about whether psychiatric syndromes lie on continuums with normal personality traits. The authors acknowledge limitations in mapping temperament types to disease categories, noting theoretical ambiguities—particularly the difficulty of linking stable introversion to an epileptic temperament—and the absence of a psychoticism dimension measure in the dataset. They argue that extraversion and neuroticism are better-established, more reliably measured dimensions, providing some justification for restricting prediction to these two axes. Methodologically, the paper emphasises that polypredictive CFA approaches (uniprediction, biprediction, and chessboard/tetraprediction) can increase the efficiency and detectability of type-specific associations when theory-driven hypotheses are specified a priori. The authors present their mappings and statistical results as tentative and exploratory rather than definitive, acknowledging theoretical and empirical uncertainties.