Beyond basic comprehension: Reading to learn and reading to integrate for native and nonnative speakers.

摘要

This study created measures to distinguish the constructs of reading to learn (RL) and reading to integrate (RI) from the more traditional reading construct, reading for basic comprehension (BC). The new measures were taken by 251 participants: 105 undergraduate native speakers of English (NSU), 106 undergraduate nonnative speakers (NNSU), and 40 graduate nonnative speakers (NNSG). The research subproblems included determination of the influence of overall basic reading comprehension level (score on the reading comprehension section of the Test of English as a Foreign Language, TOEFL-RC, or the Nelson-Denny Reading Comprehension Test, ND), native language background (NS vs. NNS), medium of presentation (print vs. computer), level of education (graduate vs. undergraduate), and computer familiarity (score on familiarity questionnaire) on RL and RI measures; and the relationship among measures of BC, RL, and RI. Results revealed that native language background and level of education had a significant effect on performance on both experimental measures while other independent variables did not. All reading measures showed some correlation, yet RL and RI had lower correlations with BC measures, suggesting a distinction between BC and the new measures. Results of discriminant analyses further supported the distinction between BC and the new measures and, for some group contrasts, suggested a hierarchy of difficulty. Discriminant analyses also identified a lower BC threshold for NNSs, below which they were unable to perform the more complex reading tasks. Additionally, this study used interviews as another source of evidence to distinguish these three reading constructs. Additional subproblems explored through interviews with a representative subgroup of 49 participants were readers' perceptions of task variability and difficulty; awareness of text structure; and awareness of cognitive and metacognitive strategies used in completion of reading measures. Also, while some participants could articulate the structure of the problem/solution texts in interviews, there was no conclusive evidence to support a link between recognition of text structure and ability to complete the more complex, experimental tasks. Results from the qualitative interview sub-study revealed that participants perceived a hierarchy of difficulty among the testing tasks that matched the theoretical hierarchy; BC least difficult, followed by RL, then RI.