Graphoanalysis, a neglected psychodiagnostic tool

The hypothesis that Graphoanalysis (GA) is a valid predictor of scores on the Guilford-Zimmerman Temperament Survey was tested by correlating Graphoanalyses of writing samples from 30 Ss, done separately by two Graphoanalysts (Gs), with results of the Guilford-Zimmerman Temperament Survey ( GZ), taken by the same 30 Ss. The hypothesis that GA is a reliable measure was tested by calculating Pearson's r and a z-score for the two sets of GA results. A specific group of GA traits was assigned to each GZ trait and correlated with it; each GA trait in the group was also separately correlated with its assigned GZ trait. Correlations of groups of GA traits with GZ traits produced significant correlations (p<0.05) associated with the GZ traits of Sociability, Objectivity, and Friendliness, and their respective GA correlates, out of 18 correlations calculated. Correlations of GZ traits with individual GA traits produced 8 significant correlations, out of 102 calculated. Since these numbers of significant correlations might have occurred by chance alone, the results are not held to support the hypothesis that GA scores will predict GZ scores. Fifty-one traits were evaluated by each G. The degree of agreement between them was tested by Pearson's r; the absolute agreement, by the Z-score for binomial distributions for an n of 30 or more. r was significant (p<0.05) for 16 of the 51 traits. The z-score showed significant results for 37 of the 51 (p<0.05). These results tend to support the hypothesis that GA is a reliable experimental tool. Difficulties in scoring and data analysis in both measures were discussed.