Hybrid Modeling of Intelligence and Linguistic Factors as Predictors of L2 Writing Quality: A SEM Approach

Abstract

The aim of this paper was to investigate the role of linguistic and intelligence factors in L2 writing. The sample included 347 Iranian learners of English. Six tests were administered to measure the participants' grammar knowledge, depth of vocabulary knowledge, breadth of vocabulary, verbal intelligence, narrative intelligence, and writing ability. Two SEM models were compared to each other. Model 1 only included grammar knowledge, depth and breadth of vocabulary knowledge as the linguistic factors of writing. Model 2, proposed by the researchers, included verbal and narrative intelligences as well. The models were then linked to the data to see which one fits better. The results of structural equation modeling show that Model 2 has better fit indices producing better parameter estimates. In the end, the applications and implications of the findings for L2 writing pedagogy and assessment are discussed.