In this study, the authors examined the validity of a holistically scored retell within a confirmatory factor analysis framework by comparing the fit of a three-factor model of reading with the data from a diverse sample of seventh and eighth graders. The final model demonstrated adequate fit, χ2(32) = 97.316; comparative fit index = .96; Tucker–Lewis index = .94; and root mean square error of approximation = .08. Retell’s chi-square difference, Δχ2(1) = 16.652, p < .001, and factor loading (.250, p < .001) were higher for the comprehension construct. Similarly, retell’s correlation to comprehension measures (r = .155–.257, p < .01) was stronger than its relationship to measures of fluency (r = .158–.183, p < .01) or word identification (r = .132, p < .05). However, retell had a large residual variance (.938) and low interrater reliability (κ = .37), suggesting that improvements to the instrument are needed. Despite overall latent differences, retell did not demonstrate differential item functioning.
Reed, D. K., Vaughn, S., & Petscher, Y. (2012). The validity of a holistically-scored retell protocol for determining the reading comprehension of middle school students. Learning Disability Quarterly, 35, 76-89. doi: 10.1177/0731948711432509