A Study of the Harris Revision of the Goodenough Draw‐a‐man Test

Measuring intelligence with the Goodenough-Harris Drawing Test.

“…In these studies, all 18 comparisons reported across ages 4 to 17 reflected significantly superior performance of girls over boys. Factor analysis suggests that boys have a greater concern for proportion and that girls are more attentive to details (Sinha, 1970), a finding that supports Harris's notion that girls' superior performance on the Woman scale can be accounted for by the inclusion of points for clothing and facial features that girls are more likely to attend to.…”

“…Examination of scorer's disagreements item by item has revealed systematic errors that reflect ambiguity in the scoring criteria for some items (Phillips et al, 1973;Sinha, 1970). Phillips et al listed 12 items they found to be vaguely defined, along with their clarification of the scoring criteria for them.…”

“…Serwer, Shapiro, and Shapiro (1972) reported a similar split-half reliability of .87. Sinha (1970) obtained interitem reliability coefficients between .84 and .90 by the Kuder-Richardson Formula 20.…”

The Draw‐a‐man Test: A Study of Scoring Methods, Validity and Norms With English Children at Five and Eleven Years

“…Yule et al (1967) obtained correlations between three judges of 0-86-0-91 for a sample of 131 English children at 10 yr. Auricchio (1966) reports the value of 0-87. Sinha (1970), investigating a large sample of London schoolchildren aged 7-11 years, obtained an overall scorer-reliability of 0-81 between two judges, but for separate age groups correlations ranged from 0-65 to 0-88. The correlation coefficients from 0 89 to 0-91 in this study accord very well with other reports.…”

“…Yule et al (1967), using 131 children from the Isle of Wight, average age 10-1 yr, correlated this measure with four subtests of the WISG, to obtain coefficients from 0-31 to 0-41. Sinha (1970) reports a wide range of correlations between the Harris Scale and a number of group measures of verbal and non-verbal ability. The correlations with Stanford-Binet in Tables 2 and 3 are similar to those reported from America (Vane, 1967); and from South Africa, Strumpfer and Mienie (1968) reported a correlation of 0-54 with an individual intelligence test among children of average age 11-4 yr. Pringle and Pickup (1963), who gave the Goodenough test to 37 English school children each year from 7-to 10-years of age, obtained correlations with Stanford-Binet in the range 0-11-0-37.…”