Publication Date


Advisor(s) - Committee Chair

Sam McFarland, Dorsey Grice, Richard Miller

Degree Program

Department of Psychology

Degree Type

Master of Arts


Since the Bender's growth in popularity as a tool for assessing organic brain damage, other factors have been shown to influence the accuracy of design reproductions. Mental deficiency, cultural background, early age and :ex are among those factors. Attempts to demonstrate the effect: of personality traits have been few, but supportive. Research into the influence of personality on Bender performance has largely been group comparison designs. The degree of overlap found among groups has reduced the degree of confidence with which the results can be applied to individual performance:. The present study used a correlational approach to determine how validly State and/or Trait Anxiety can be predicted from Bender records.

Results indicated significantly more positive correlations than chance between 22 of the Bender deviations and deviations scored for severity and both the State and Trait Anxiety measures. This supported the hypothesis that Bender deviations are affected by the presence of anxiety. A factor analysis of the Bender records did not produce any robust factors, probably because of the high sample homogeneity with respect to their Bender performance. Alpha coefficients were increased from those obtained on the Deviation Total and Deviation Severity scales by selecting two subscales from each, based on the item-total coefficients of the deviations. Correlations consequently increased with State and Trait Anxiety on three of the four subscales. Lastly, a step-wise multiple regression procedure was applied to both the Deviation Total and Deviation Severity scale to obtain multiple Rs with :State and Trait Anxiety. Adding deviations continued to increase the multiple R of the lists through eight to thirteen steps. Corrections for the small sample size estimated the possible shrinkage of the multiple Rs.

Discussion focused on finding no correlations between are, sex and Bender performance, as expected. It also stressed some of the procedural flaws and data assumptions, most noticeable of which was the need for a more reliable Bender scoring blank for further research. Finally, it was concluded that the results supported the hypothesis that Bender deviations are influenced by anxiety. However, the results further pointed out that Bender deviations, as presently defined, are affected by other factors as well. It was therefore cautioned that sender reproductions should be considered only a rough predictor of the presence and extent of anxiety in an individual.


Personality and Social Contexts | Psychology | Social and Behavioral Sciences