The role of background knowledge in reading comprehension of subject-specific texts

This thesis investigates the impact of background knowledge in L2 reading comprehension of subject-specific texts, in particular its interaction with grammar knowledge. It explores how different levels of discipline-related background knowledge, grammar knowledge, and self-reported familiarity affect individual differences in L2 reading comprehension in terms of its outcomes and process. A number of studies made assumptions about readers' knowledge based on their study disciplines or reports by readers themselves, and thus this study also explores the difference between two operationalizations: tested background knowledge and self-reported familiarity. A mixed-methods approach was used by combining two studies: a testing study and a think-aloud study. Altogether 404 students of the School of Economics and Business, University of Ljubljana, Slovenia took part in the study; 22 in the piloting study and 382 in the main study, out of which 358 were engaged in the testing study and 24 in the think-aloud study. The quantitative and qualitative datasets were obtained from five research instruments: a grammar test, a test of discipline-related background knowledge, a reading comprehension test based on three finance texts, a post-reading questionnaire, and think-aloud verbal protocols. The results of multiple regression revealed that tested background knowledge was a significant medium strength predictor of reading comprehension, slightly stronger than grammar knowledge. In contrast, self-reported familiarity was not found to impact reading comprehension and was not its predictor. This evidence casts doubt over self-reporting as an operationalization of knowledge in L2 reading. Apart from having a facilitative effect on L2 reading comprehension, background knowledge was found to have compensatory and additive roles when interacting with grammar knowledge. The findings showed that readers with higher discipline-related background knowledge could use it to make up for lower grammar knowledge and vice versa, thus suggesting the compensation effect between the two variables. In addition, the results revealed that readers were able to use their background knowledge regardless of their level of grammar knowledge, albeit slightly less at higher levels of grammar knowledge. This finding suggests that the threshold hypothesis could not be supported. Finally, the group of students with both high background knowledge and high grammar knowledge outperformed other groups in reading comprehension, which suggests that the two variables affect reading comprehension in an additive way. The qualitative data from verbal protocols in the think-aloud study and readers' scores from the testing study were obtained to compare the processing patterns and strategies used by readers with high and low background knowledge. Although both groups were found to use the same types of strategies, they differed in the frequency of their use. While the high background knowledge group used more correct paraphrases, elaboration, inferences, and evaluating, the low knowledge group adopted a more local-level approach by paying more attention to individual words, phrases, and sentences and reporting on various comprehension problems and inability to see the bigger picture. The results suggest differences between the groups with different levels of background knowledge with regard to semantic and pragmatic processing at the local and global level. Analysis of the verbal protocol and post-questionnaire data revealed that specialist vocabulary was the main source of difficulty in L2 reading comprehension of subject-specific texts.