ACTION RESEARCH PROPOSAL

Action Research Proposal

Action Research Proposal

Background Information

Individual differences in word decoding accuracy and efficiency are important indicators of overall reading proficiency for beginning readers (e.g., National Reading Panel, 2000) and for adults functioning at low levels of literacy (Sabatini, 2002 and Wagner & Venezky, 1999). However, findings regarding the persistence of these relations at high levels of reading skill have been mixed. Therefore, the present study will be designed to test the limits of approaches to individual differences in text selection that emphasize word-level processes (e.g., Perfetti, 1985) within a sample of mature, successful students—those attending a research university.

Despite the popularity of the college sophomore as a research participant, large-scale studies of college students' literacy typically have included measures of text selection only, not component reading skills (e.g., Bray et al., 2004 and Reder, 2000; but see Cunningham et al., 1990 and Perfetti & Hart, 1999 for exceptions). Those investigators who have examined postsecondary students' component reading skills have studied relatively small samples, often contrasting groups selected for particular ability levels (e.g., Bell & Perfetti, 1994, Bruck, 1998, Haenggi & Perfetti, 1994, Jackson & Doellinger, 2002, Martino & Hoffman, 2002 and Sabatini, 2002). For example, Bruck (1998) studied adults who had been identified in childhood as having a specific reading disability; Wilson and Lesaux (2001) compared 28 college students with and 33 without a recently confirmed diagnosis of reading disability; and Bell and Perfetti (1994) studied only 30 university students chosen to represent three skill profiles. Furthermore, none of these studies examined the extent to which inefficient decoding might have interfered with students' success in their college coursework.

The contribution of word decoding accuracy and reading speed to college students' reading selection and academic success cannot easily be predicted from the broader literature on reading acquisition. This literature can be used to both to support the hypothesis that word-level processes are important at all levels of proficiency and to support the alternative hypothesis that the relevance of these processes as sources of individual differences becomes minimal as readers become highly skilled.

Many researchers studying acquisition of English reading skill (e.g., National Reading Panel, 2000, Perfetti, 1999 and Share & Stanovich, 1995) would agree with a description of reading selection success that emphasizes the importance of efficient word recognition, made possible by mastery of the alphabetic principle. Abundant data support this bottom-up view. High levels of performance on tests of nonword and word reading accuracy and speed have been found to be associated with good text selection in readers at all levels of development. Correlations between text reading fluency and text selection are often so high that fluency measures have been proposed as stand-ins for selection tests (Fuchs, Fuchs, Hosp, & Jenkins, 2001). Share and Stanovich (1995) concluded from a review of the literature that, “even among adults, word recognition accounts for a substantial amount of variance in reading ability” (p. 3).

On the other hand, associations between component skills and text selection do weaken as reading becomes more skilled ...