RETEST RELIABILITY OF THE WOODCOCK LANGUAGE PROFICIENCY BATTERY

The Woodcock Language Proficiency Battery: English Form was administer ed to 152 of the 478 students beginning medical studies at an Australi an university in 1990 and 1991. Two yeats after admission 147 of the s tudents were available for retest. Test-retest reliability coefficient s ranged from .70 to .85 for the 8 subtests, .83 to .91 for the three cluster scores, and the coefficient for the Broad Language Cluster sco re was .95. The sample had deliberately over-represented students low in English language proficiency but the reliability coefficients were only slightly reduced when this bias was corrected by forming a repres entative sample of 72 students. The mean incremental gain on the Broad Language Cluster of 1.8 was statistically significant and corresponde d closely to the 2.0 predicted from the test norms for a 2-year develo pmental period. These results support previous reports of the excellen t psychometric properties of this instrument.