Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)


The understanding of children's abilities to express affect in drawings of the human figure has clinical and developmental significance, yet is an area that until recently has received relatively minimal research attention. The present study investigated the abilities of child psychologists, graduate psychology students, and non-mental health raters to correctly identify the attempts of 72 kindergarten, second, fourth, sixth, eighth, and tenth-grade students to depict six common affective expressions in human figure drawings. The affective expressions were: happiness, sadness, anger, disgust, surprise, and fear. Thirty-six adult raters (18 female; 18 male) reviewed 216 drawings, identified the perceived affective expression, and rated their confidence in their selections. The effects of child sex, adult sex, affective expression, and rater group assignment on rater accuracy were analyzed. The results clearly indicated significant differences in recognition rates for the six affective expressions with Sadness and Happiness correctly recognized most frequently ((TURNEQ)80%), followed by Anger and Disgust ((TURNEQ)60%) and Surprise and Fear ((TURNEQ)35%); the overall recognition rate was 58%. Accuracy of recognition was considerably lower (i.e., 41%, overall) when rater commission errors were considered. Significant increases in adult recognition rates directly coincided with increases in child age; however, these differences were not consistent across affective conditions. Recognition rates were not affected by the sex of the adult rater. Graduate psychology students obtained the highest mean accuracy rate (61%), followed by child psychologists (58%) and non-mental health raters (55%): the significant differences between rater groups were attributed to individual rather than actual group differences. Raters apparently recognized their limitations on this task as indicated by significantly lower confidence ratings for incorrectly identified than correctly identified drawings. Higher accuracy ratings were associated with lower confidence ratings and vice versa, suggesting that accuracy and perception of accuracy were inversely related. Caution was encouraged when making clinical inferences from the data because of the seminal nature of the study. It was concluded that the adult recognition rates were too low to be clinically reliable, despite being well-above chance expectations. It was suggested that drawings may have a place in clinical child assessments if important precautions are considered. Recommendations for further research were given.