SCAN test-retest reliability for first- and third-grade children.

The SCAN is a popular screening test that was developed to provide a rapidly administered, standardized method for determining the potential of central auditory processing disorder (CAPD) in children between the ages of 3 and 11 years. It can be administered in 20 minutes with a portable stereo cassette player and contains three subtests: filtered words (FW), auditory figure ground (AFG), and competing words (CW). Published SCAN test-retest reliability data (Keith, 1986) used a 6-month retest interval and indicated that SCAN scores may be unreliable. No additional reliability data are available, and studies indicate that SCAN has been used by both researchers and clinicians despite reliability concerns. This investigation examined the stability of SCAN outcomes for 25 first-grade and 22 third-grade children (ages 6 to 9 years) using a 6- to 7-week retest interval. Time of day and examiner were held constant, and participants were normal-hearing, were Caucasian, and spoke English as their primary language. ANOVA outcomes indicated that both raw and standard scores improved significantly from Test 1 to Test 2 for two of the three SCAN subtests (FW and CW) and for the composite (COMP) score. Additionally, COMP-percentile-rank and age-equivalent outcomes demonstrated significant improvement from test to retest for both grades. The AFG subtest was the only SCAN measure for which a significant test-retest difference did not emerge. The highest test-retest correlation values (r) were moderately strong (0.70 < or = r < or = 0.78) and occurred for the CW and COMP scores. Implications of correlations and factor analyses are discussed. It is suggested that examiners base recommendations for additional testing, follow up, and remediation on the COMP score only. Further, it appears that second administration of the SCAN can provide a better estimate of an individual child's best performance, but lack of second-score norms confounds simple interpretation of such scores.