Psychological assessment of the child patient using Sandplay

This study investigates the use of Sandplay as a projective assessment technique with a child inpatient population. The Sandplay measure is compared to two other standard assessment techniques, the Rorschach ink blot test, and an objective measure, the Achenbach Child Behavior Checklist (CBCL). It was hypothesized that the Sandplay, a less structured, more non-verbal tool, would correlate positively with the Rorschach and CBCL measures. The Sandplay and Rorschach tasks purport to measure aspects of an individual's object relations and the CBCL provides an index of psychopathological symptomology. Subjects (n = 41) were latency age children admitted into the inpatient psychiatric unit at Elizabeth General Medical Center and ranged from 5 through 12 years of age. Half the children received the Rorschach first, half the Sandplay task first. Primary nursing staff filled out the CBCLs within a week of the child's admission. Rorschachs were administered and scored using the Urist Mutuality of Autonomy scale procedure which focuses on a child's range and average mode of object relations, scoring for all implied or explicit movement responses. Sandplay was administered and scored based on prior work by Caproni, Segal, and Maza (1989, 1990, and 1991, respectively) in which judges of a final photo of the child's production rate the Sandplay along object relational dimensions.; For the main hypotheses, few correlations were found until the children were separated by gender. While more statistically significant correlations were found when data of boys and girls were separated, these findings must be viewed with caution because sample size is so small (boys' n = 30; girls' n = 11). Boys were found to perform significantly worse than girls on all Sandplay variables than did girls although no statistical tests assessing the magnitude of these differences were used. This finding suggests possible differences in the way latency age boys and girls process and express emotional disturbance on a non-verbal task. It would be worthwhile to investigate latency age gender differences in manifestations of psychopathology on these tasks with a larger sample size. Surprisingly, the more standard Rorschach and CBCL tasks were not found to correlate, suggesting these measures may not measure comparative constructs.