Masters Degree Thesis, December 2003

The current study examines the logic behind the Wordchains test using data from 738 students in grades 7 to 12 from the U.S. The logic behind Wordchains is that a speeded test made of uniformly easy items can help measure the development of automaticity and differences among fluent and non-fluent readers. This study finds support for the claim that Wordchains is made of the uniformly easy items. The result of frequency analysis suggests that most of Wordchains items are frequent enough to be familiar to all participants. The results of an item difficult analysis were that 96 % of the Wordchains items had p-values within that indicated they were uniformly easy. However, findings did not establish that the Wordchains test could separate fluent from non-fluent readers. There was a significant linear trend between Wordchains scores and grade levels. However, this linear tend was not significant in higher grades 10-12. Furthermore, comparison of Wordchains scores and teachers' ratings yielded only moderate correlations, indicating that the Wordchains scores shared insufficient variance with teacher ratings of students' fluency. These findings indicate that further development of the Wordchains is needed as well as further studies of concurrent validity.