Abstract

ROSA, Helena Rinaldi  and  ALVES, Irai Cristina Boccato. Goodenough-Harris reliability in children. Aval. psicol. [online]. 2008, vol.7, n.2, pp. 171-179. ISSN 1677-0471.

This study investigated the reliability by the half-split method and retest method of the Goodenough-Harris Test. The sample was composed by 1540 children, from 5 to 11 years old, of both sexes, from São Paulo city. The task of drawing a man was administered individually. The reliability by split-half method was obtained for each age and for the total sample. Pearson correlations, corrected by spearman-Brown formula, ranged from 0,746 to 0,895 and were of 0,923 to total sample. The retest sample was constituted by 67 children from 6 to 9 years old, 38 boys and 29 girls, from four public schools. The retest was applied after 8 to 15 days. Correlations were statistically significant to both age ranges (0,822 and 0,703) and for total sample (0,808). The reliability obtained by two methods can be judged satisfactory according to literature criteria.

Keywords : Human Figure Drawing; Psychological tests; Reliability; Intelligence; Psychometry.

        · abstract in Portuguese     · text in Portuguese     · Portuguese ( pdf )