The effect of reader knowledge and textual features on second-language reading outcomes in an Indonesian EFL context

Abstract

Vocabulary and sentence-level grammar skills are the bedrock on which fluent L2 reading skills rest. The importance of vocabulary knowledge in particular is self-evidentmtexts consist of wordsmbut the actual contribution that individual differences in vocabulary knowledge make to L2 reading outcomes seems to vary. Greater vocabulary knowledge, as reflected in the basic dimension of size, correlates strongly with better reading outcomes (Laufer, 1992; Qian, 1999, 2002), but there are also studies that show little, or no, link between the two (Yamashita, 2013). The inconsistency is due to differences in target population, task, setting and other study-specific factors. It may also reflect the difficulty of the text relative to the readersr proficiency level. This study examines the relationship between L2 reading outcomes and reader based vocabulary and grammar knowledge as it is moderated (Hayes, 2013) by the text based features of lexical difficulty and syntactic complexity. By examining the potential interaction between reader knowledge and text demands, the study provides a more refined picture of the factors affecting reading outcomes, with a specific focus on vocabulary size, alone and in combination with sentence-level grammatical knowledge. The latter is also a potential moderating factor on the link between vocabulary knowledge and reading outcomes. n nParticipants were EFL students (n = 71) in the second year of their English major at an Indonesian university, with a range of English proficiency levels. The participantsr vocabulary knowledge was measured using the Timed Yes/No vocabulary test (Harrington, 2006), a measure of vocabulary size and processing skill. Grammar knowledge was established using a test of sentence-level grammatical knowledge. Text based syntactic complexity was established using the TextEvaluator readability formula (ETS, 2013), and lexical difficulty was defined by the lexical frequency of occurrence as assessed using VocabProfile software (Cobb, 2010). Each participant completed four reading texts, each reflecting a unique combination of two levels (high vs. low) of syntactic complexity and lexical frequency. n nThe first research question established the predictive power of individual differences in the participantsr vocabulary size and grammatical knowledge for reading outcomes: (RQ1) Do individual differences in linguistic knowledge (vocabulary size and grammatical knowledge) predict L2 reading outcomes? What is the relative contribution of each? Then, the role that differences in text based lexical and syntactic complexity play in reading outcomes was characterised: (RQ2) Are textual features (lexical frequency and syntactic complexity) stable predictors of reading outcomes? What is the relative contribution of each? Lastly, the relationship between reader knowledge and text features in predicting L2 reading outcomes was analysed, with a specific focus on potential moderating effects of text features: (RQ3) Do reader knowledge and text-feature factors interact to predict reading outcomes? nn nFindings for RQ1 showed vocabulary knowledge was a stronger predictor than grammatical knowledge. This replicates previous studies showing the importance of vocabulary knowledge in explaining variation in L2 reading (Brisbois, 1995; Zhang, 2012; Yamashita, 1999) and is at odds with other studies (Shiotsu a Weir, 2007; Shiotsu, 2010). Findings for RQ2 showed that lexical difficulty, as indexed by lexical frequency, accounted for more variation in reader outcomes than syntactic complexity. This finding supports previous findings (Brown, 2013; Eslami, 2014; Twessi, 1998). For RQ3 there was a mean interaction effect between lexical frequency and reader vocabulary knowledge such that the relationship between vocabulary and L2 reading for high lexical frequency texts was stronger than for low lexical frequency texts, but the effect was not statistically significant. There was no effect for differences in text syntactic complexity. The results also showed that text lexical frequency significantly moderated the relationship between reader grammatical knowledge and L2 reading, such that the relationship between grammatical knowledge and L2 reading for high lexical frequency texts was significantly greater than for low lexical frequency texts. Given the absence of significant interaction between vocabulary knowledge and lexical frequency, the sample was split into high and low vocabulary size groups (30% each) to see if more distinct proficiency differences would yield significant differences for lexical frequency effects. The subgroup analysis did not yield different results, though it did show that the grammatical knowledgenlexical frequency interaction evident in the complete sample was only evident in the low vocabulary size group. n nThe findings are consistent with Perfettirs (1985) verbal efficiency theory and Kintschrs (1998) construction-integration theory and their account of lower-level processing in L2 reading comprehension. The thesis demonstrates the importance of looking at learner knowledge in relation to text demands in understanding the relationship between learner knowledge and L2 reading outcomes. Pedagogical implications relate to a better understanding of how learner level and text difficulty affect performance. For the Indonesian EFL participants examined here, it also indicates that vocabulary development should be a central focus in the development of English L2 reading skills.