ArticleLinnoila M, Liljequist R, Olkoniemi J, Saario I.
Pharmakopsychiatr Neuropsychopharmakol. 1977 Jul;10(4):246-53.
For 2 weeks 40 volunteers received either 5 mg diazepam, t.i.d., or 10 mg chlordiazepoxide, t.i.d., and placebo. A choice reaction test, two coordination tests, and an attention test were administered to the subjects on the 14th day of each treatment. Thirty minutes before the tests, the subjects ingested either alcohol., 5g/kg or a placebo drink, incombination with the last capsule. After the test the subjects rated the quality of their treatment as placebo, tranquilizer, or stimulant. The psychological tests taken before the treatments were Eysenck's EPIC-NESI, Taylor's Manifest Anxiety Scale (MAS), and Cattell's 16 PF inventory. A multiple regression analysis was computed. Personality factor scores found to be associated with a strong effect of the benzodiazepines were 16 PF's A, C, L, N, and Q, and EPIC's E. The effect of alcohol was associated with a high score of 16 PFs B factor. Personality factors associated with "placebo reactors" and nonreactors were investigated, as well. Those subjects on placebo indicating their treatment to be active were classified as "placebo reactors". A discriminant analysis revealed that 16PF's O and I factors discriminated effectively "placebo reactors" from nonreactors, and EPIC's SE and 16 PF's L factor nonreactors from "reactors".