Dissertation Defense: Gabriele DeRobles
Candidate Name: Gabriela DeRobles
Advisor: Ronald P. Leow, Ph.D.
Title: The Effects of Type of Written Corrective Feedback and Level of Proficiency on Processing and Accuracy in Heritage Language Learners of Spanish
In the field of Instructed Second Language Acquisition (ISLA), written corrective feedback (WCF) has been shown to play a facilitative role in improving written accuracy (e.g., Ashwell, 2000; Ferris & Roberts, 2001; Ferris, 2006; Sheen, 2007; Bitchener, 2008; Liu & Brown, 2015). However, studies that examine the effects of different types of feedback (which vary in explicitness) on L2 learning have yielded inconclusive findings (e.g., Chandler, 2003; Van Beuningen, De Jong, Kuiken, 2008; Benson, Dekeyser, 2018). Traditionally, the literature in this area of research has focused solely on L2 foreign language learners or English as a second language learners with little attention to heritage language learners (HLLs). While there is growing interest in HLLs, and particularly in how they can be supported to develop their academic writing skills in the HL, an area of particular vulnerability for HLLs, whether type of feedback provided exerts influence on their written accuracy over time, remains understudied, and therefore, not understood. In addition, this dissertation not only explores the efficacy of WCF on the improved accuracy of HLLs on errors relating to mood, aspect, and orthography, but also the extent to which the learner’s proficiency may moderate the effects of different types of feedback. Furthermore, this study investigates the role of depth of processing (DOP) on subsequent language development via the use of think aloud (TA) protocols.
The purpose of this study is three-fold: 1) compare the accuracy scores of HLLs at high and very high proficiency levels resulting from indirect (i.e., error code corrections) or direct (i.e., full reformulation of the error) feedback on a composition, 2) gather concurrent data on how HLLs at high and very high proficiency levels process types of WCF, and 3) investigate the relationship between DOP and types of feedback, and the effects this relationship may have on learning outcomes. Participants were randomly assigned to one of two experimental conditions (direct WCF+ TA or indirect WCF+TA) or the control (no feedback) condition. All participants completed an untimed semi-guided writing task (Draft 1). They revised their compositions using their respective feedback, and a week later, they wrote a second composition on a similar topic (Draft 2).
The results showed that at both proficiency levels (High and Very High), both direct and indirect WCF groups significantly outperformed the control group. However, for learners with lower proficiency, providing direct feedback on subjunctive and orthographic errors led to higher accuracy. The qualitative data revealed that HLLs in both proficiency groups used explicit metalinguistic knowledge as well as their implicit non-metalinguistic knowledge of their HL (i.e., intuition) when revising orthographic and grammatical errors. Regarding the relationship between the type of WCF and DOP, both feedback types (i.e., direct and indirect) elicited similar levels of DOP (i.e., low, medium, and high). However, differences were observed when proficiency was taken into consideration. Finally, there was a positive relationship between deeper levels of processing and improved accuracy from Draft 1 to Draft 2, particularly for the indirect WCF group.
Wednesday, June 5, 2019 at 2:00pm to 4:00pm
Edward B. Bunn, S.J. Intercultural Center, 302P
37th and O St., N.W., Washington