| Within instructed second language acquisition (SLA), Processing Instruction (PI) has enjoyed a rigorous research agenda for more than 15 years. Research comparing PI (with or without explicit information about a target linguistic structure) with either traditional, production-based instruction (e.g., Cadierno, 1995; VanPatten & Cadierno, 1993) or meaning-based output instruction (e.g., Farley, 2000; Morgan-Short & Bowden, 2006) has shown that PI activities are as good as, if not more effective than, other types of instruction. The effectiveness of PI has been attributed to the way in which PI seeks to alter learners' non-optimal input processing strategies via structured input (SI) activities for the creation of richer intake data (Sanz & Morgan-Short, 2004; VanPatten, 2004; Wong, 2004). However, to date no studies have compared SI with other types of input-based treatments to determine whether the effectiveness of SI is due to (a) altering learners' processing strategies, or (b) simply providing learners with meaning-bearing input. Therefore, this dissertation compares the effects of various input-based treatments (input flood, input flood + text enhancement, focused input, and structured input), along with a control group, on the interpretation and production of Spanish 3rd person accusative clitics. Participants included 290 adult learners enrolled in an intermediate Spanish course. Learners completed a pretest, a computer-based treatment, an immediate posttest, delayed posttest (3 weeks after treatment) and an extended delayed posttest (6 weeks after treatment). Assessment tasks measured both interpretation and production of accusative clitics. Results for the interpretation task revealed that although all treatment groups (except the control group) showed significant gains over time, only the SI group significantly outperformed the control group at the second delayed posttest. On the production measures, all input-based groups (except the control group) showed improvement; however, no significant differences emerged among the four input-based treatments. The findings for the production measures are not so straight forward; however all groups significantly improved or approached significance. The theoretical, methodological and pedagogical implications of these findings are discussed, along with limitations to the study and avenues for future research. |
