Support for abbreviation of the Wide Range Achievement Test-Revised spelling subtest in neuropsychological...

The Wide Range Achievement Test-Revised (Level 2) provides valuable data in neuropsychological assessments, but its inclusion is costly in time. Although the Reading and Arithmetic subtests are inappropriate for abbreviation, the Spelling subtest is unnecessarily lengthy for most clinical needs. In a retrospective study of the reliability of data obtained from split-half and split-third tabulations with mixed psychiatric and forensic subjects, strong correlations were obtained between the split-halves and between split-half and full administration scores. A Spearman-Brown correction of the relationship between the split-halves yielded a correlation of .97, higher than that reported for any WAIS-R subtest (Wechsler, 1981). Standard scores generated by split-half tabulations were within 10 points of those generated from the complete raw data in 94.6% of cases. These findings suggest that the Spelling subtest could be abbreviated in clinical settings with minimal loss of reliability.