Academic journal article School Psychology Review

Examining Readability Estimates' Predictions of Students' Oral Reading Rate: Spache, Lexile, and Forcast

Academic journal article School Psychology Review

Examining Readability Estimates' Predictions of Students' Oral Reading Rate: Spache, Lexile, and Forcast

Article excerpt

Curriculum-based measurement in reading (CBM-R) is an assessment that involves students reading passages aloud while an examiner records the number of words read correctly per minute (WRCM). Extensive evidence supports the reliability and validity of CBM-R for estimating students' relative levels of reading achievement (e.g., Marston, 1989; Wayman, Wallace, Wiley, Ticha, & Espin, 2007). CBM-R is also intended to allow educators to model individual student growth through the administration of alternate passages across time. Observed WRCM are plotted in time series fashion to model student growth. A critical feature of measures used to model growth is that the construct measured and difficulty level of alternate probes remain constant across time (Francis et al., 1994). The use of probes of varying difficulty has the potential to result in estimates of growth that are not true reflections of actual growth in the behavior, skill, or construct being modeled (Ardoin & Christ, 2009).

Readability estimates, which utilize counts of variables, such as number of syllables per 100 words, to estimate passage difficulty have been employed to select and revise passages within CBM-R passage sets. For instance, in the development of Dynamic Indicators of Basic Early Literacy Skills (DIBELS) passage sets, passages were first evaluated using readability estimates, revised to fall within a predetermined range, and then once again evaluated using readability estimates (Good & Kaminski, 2002). Similarly, the equivalence of AIMSweb passages was assessed primarily using readability estimates. An additional step, however, was employed in the development of AIMSweb passage sets: approximately 20 students per grade read each passage to evaluate its appropriateness (Howe & Shinn, 2002). Unfortunately, although there is a large and dated body of research evaluating the utility of readability estimates (Fry, 1989), readability estimates have not been evaluated for the purpose of selecting or modifying passages within CBM-R passage sets.

Results of the few studies that have examined readability estimates in the context of CBM-R are not promising. Compton, Appleton, and Hosp (2004) assessed the relationship between readability estimates, students' oral reading rate, and students' reading accuracy. Among other variables (e.g., decodability, percentage of high-frequency words in the text), the Spache (1953) and Flesch-Kincaid (Flesch, 1948) readability estimates were calculated for 15 passages administered to 248 students. Results suggested that the readability estimates did not predict WRCM or accuracy. However, both the relationship between accuracy and percentage of high frequency words, as well as the relationship between WRCM and percentage of high frequency words and decodability, were significant. In a similar study, Ardoin, Suldo, Witt, Aldrich, and McDonald (2005) examined eight readability estimates used in the CBM-R literature to predict students' WRCM. Results revealed only a modest relationship between WRCM and passage difficulty as determined by the readability estimates. Unfortunately, the Spache (1953), DaleChall (1948), and Flesh-Kincaid (Flesch, 1948) estimates, identified as the most commonly used readability estimates in CBM-R research by Ardoin et al., were found to be the poorest predictors of WRCM. Of the eight readability estimates, the Forcast (Sticht, 1973) estimate was found to be the best predictor of WRCM. These results supported those of Powell-Smith and Bradley-Klug (2001), who found the relationship between students WRCM and both the Dale-Chall (1948) and Fry (1989) estimates to be insignificant.

Lexile scores, which were used as one of many estimates for selecting the CBM-R passage that make up the AIMSweb passage sets (Howe & Shinn, 2002), have been examined in only one study related to students' oral reading rate (Walpole, Hayes, & Robnolt, 2006). A Lexile score is a quantitative measure of readability that is determined by word frequency (Stenner, 1997). …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.