G. A. Lienert

Defining tranquilizers operationally by non additive effect in experimental stress situations

Clinically and theoretically, a tranquilizer should be more effective under stress than under non stress conditions. By transformation into a statistical non additivity equivalent, a 2×3 factorial design was derived with a drug factor (Cloxazolam as a tranquilizer) and a social stress factor. The results in healthy young subjects show a significant drug-x-stress interaction, indicating the non additivity of both treatments. Psychological interpretations, according to Yerkes-Dodson law, and therapeutic and experimental conclusions are added.

LSD RESPONSE IN EYSENCKIAN TRAIT TYPES IDENTIFIED BY POLYPREDICTIVE CFA

Abstract The four personality type combinations derived from high and low extraversion ( E+ E− ) and high and low neuroticism ( N+ N− ) have been related to response patterns composed of three symptoms (affective disturbances, thinking disturbances, and blackouts) scored as present (+) or absent (−) after a single oral dose of the hallucinogenic drug LSD-25. Hypotheses for expected response patterns for each personality group were derived from a data set obtained by Kohnen and Lienert (1987). Significance of associations was tested by two strategies of polyprediction configural frequency analysis (CFA): multiple uniprediction and biprediction CFA. Both strategies yielded a significant hyperpresentation of all three symptoms present in E+N+ (hysterics), merely thinking disorders in dysthymics (E−N+), merely affective symptoms in E+N− (stable extraverts), and merely blackouts in N−E− (stable introverts). Authors tried to relate these symptoms to Kretschmers temperament types and could afterwards show by a chessboard modification of prediction CFA, that by applying two combined hypotheses for two personality types each, the significance of the predicted associations could be increased.

Point symmetry adjustment of phi-coefficients in the factor analysis of psychometric test items

Abstract Existing computer packages for data analysis are not generally sensitive to the underlying assumptions of factor analysis and issues concerning the appropriateness of data. One serious problem in the factor analysis of test items is that of response skew. This paper discusses some strategies for mitigating this problem and emphasizes the use of alternatives to the Pearson-Bravais inter-item correlation coefficient. In particular, it is shown that the Holley-Guilford G -index of association may be interpreted logically as a coefficient of correlation between two binary items, dichotomised at the median of their underlying continua, following a point symmetry adjustment. The G -index is a point symmetry adjusted phi-correlation coefficient. It is argued that principal components analysis of the inter-item matrix of G -indices is less distorted than that of phi-coefficients where parametric assumptions are in doubt.

Personality classification agreement identified by configural frequency analysis

Abstract Agreement tables are defined as squares rxr contingency tables with identical row and column categorial variables. It is proposed to test for categorial agreement cellwise by configural frequency analysis, rather than globally by Cohens K coefficient. The configural agreement tests are illustrated by an example from experimental personality research: it is shown that Eysencks and Cattells personality classifications are in good cell wise categorial agreement.

Longisectional Interaction Structure Analysis (LISA) in Psychopharmacology and Developmental Psychopathology

Longisectional interaction structure analysis (LISA) is a method for evaluating multivariate observations in a sample of individuals (patients) at two or more than two subsequent times (stages). It combines cross-sectional configural frequency analysis (CFA) for defining interactions between variables at a given stage with longitudinal interaction structure analysis (ISA) in relating variables observed at two subsequent stages, nonparametrically. The interactions are identified locally as types rather than globally as contingencies, where types are defined as (cross-sectional or longitudinal) patterns occurring in more individuals than expected under H0 of no (cross-sectional or longitudinal) interaction. LISA is applied to data sets from a clinical follow-up study and from a longitudinal study within developmental psychology. It is shown to be a useful technique for the interpretation of such data.

Configural Frequency Analysis as a Statistical Tool for Developmental Research.

Configural frequency analysis (CFA) is suggested as a technique for longitudinal research in developmental psychology. Stability and change in answers to multiple choice items and in Yes-No item-patterns obtained with measurements repeated two or three times are identified by CFA and illustrated by developmental analysis of an item from Gorhams Proverb Test.

Item Homogeneity Defined by Multivariate Symmetry

The concept of item homogeneity, implying that items have equal difficulties and equal intercorrela tions, is defined by multivariate axial symmetry. A nonparametric test for J-item homogeneity is pro posed and illustrated by a numerical example.

Modified Phi Correlation Coefficients for the Multivariate Analysis of Ordinally Scaled Variables

The Pearson-Bravais correlation coefficient is the most widely used intervariable association index with multivariate analysis of underlying structure. However, it is a suboptimal coefficient when assumptions of multivariate normality are violated because of its tendency to become attenuated. A nonparametric class of intervariable association indexes is presented that follows the logic of point symmetry adjustment of the Pearsonian phi. This approach is extended to multipoint scales. These indexes may be shown to have useful Euclidean properties and to be attractive alternatives to the Pearson correlation coefficient for multivariate analyses.

Examining the relationship between the eysenck personality inventory and the sixteen personality factors questionnaire using predictive configural frequency analysis

Abstract A nonparametric analogue of canonical correlation analysis, called P-CFA, was used to determine how the Eysenck Personality Inventory is related to Cattells Sixteen Personality Factors Questionnaire . The P-CFA methods used showed that Eysencks personality dimensions of Extraversion and Neuroticism can be predicted quite well from suitably chosen primary scales in the Cattell questionnaire. The results were consistent with those obtained in another study using canonical analysis. Beyond that, P-CFA methods were used to predict Eysenckian trait-type categories from Cattells primary scales. The results were weak and not predictable from canonical analysis, but showed potential analytic capabilities of P-CFA that are not available in canonical analysis.

Übereinstimmungsmaße für subjektive Merkmalsbeurteilungen

Kapitel 5 zeigte, wie Zusammenhangsmase ermittelt und uberpruft werden konnen. Die dort ubliche Frage lautete, ob zwischen 2 Merkmalen X und Y, die an einer zufalligen Auswahl von Individuen erhoben wurden, ein Zusammenhang besteht. Fur ordinalskalierte Merkmale haben wir zur Beantwortung dieser Frage Spearmans rho und Kendalls tau kennengelernt.