Diagnostic Implications of Wisc Digit Span Scores

Abstract

Some meaningful (significantly different from zero) correlations between Digit Span and other WISC variables were obtained for a sample of 324 children in 7 diagnostic groups. There is no evidence of reliab!e sex or diagnostic differences in these correlations.As part of a larger study by Schoonover and Hertel (1970), later reanalyzed by Winne and Schoonover (1972), correlations between Digit Span scaled scores and other WISC variables were obtained for each of nine clinical groups and the total sample, to determine whether the pattern of correlacions was characteristically different for each clinical group. The sample consisted of 231 males and 93 females; N for the individual samples ranged from 7 to 50. A complete description of the sample is given in the report by Schoonover and Hertel ( 1970). Table 1 gives the maximum N and obtained correlations for the two sexes and the combined groups.? Correlations that are significantly different from zero as ascertained from tables prepared by Sockloff and Edney (1972) are indicated by an asterisk. All correlations for the total male, total female, and combined sample are significant but the correlations developed for the individual diagnostic groups are generally not significantly different from zero. The single exception is the male chronic brain-syndrome sample; this group yielded significant correlacions becween Digit Span and all other WISC variables except Picture Completion.Each of the 130 pairs of correlations for males and females was tested by Fisher's z-transformation to determine whether there were significant sex differences (Walker & Lev, 1953, pp. 253-254). Three critical ratios were significant at the 1% level and six at the 5% level. In most instances, significant differences were found when the correlation for one sex was non-zero while the corresponding correlation for the other sex was not significantly different from zero.Finally, h e data from the combined samples were tested with the z-transformation to determine if there were differences among the clinical groups. Of 'All correlations and statistical tests were performed on the WANG 700-C Programming Calculator using programs developed by the second author. Requests for reprints should be sent to the first auchor. Tabled significances of differences for all comparisons may be found in Document NAPS-02512. Order from Microfiche Publications, 440 Park Ave. S., New York, N. Y. 10016. Remit $1.50 for microfiche or $5.00 for photocopy. 'One male psychoneurotic was not given the Coding test so the N for this correlation is 42. Vocabulary was not given to 33 males or 11 females so correlations are generally based on fewer Ss than those for ocher subtests.