The Secret Is in the Processing: A Study of Levels of Explicit Computerized Feedback in Heritage and L2 Learners of Spanish

Abstract

The field of Instructed Second Language Acquisition (ISLA) has expressed interest in pursuing a research agenda that expands the current heritage language (HL) strand of research to investigate how this heterogeneous population re-learns their family language, and how this experience differs from that of second language (L2) learners. This dissertation examines an unexplored aspect of this strand - Depth of Processing (DOP) - and its potential role in the development of the Spanish pluperfect subjunctive in contrary to fact conditional sentences in the past, in both HL and L2 learner participant populations, and how it is facilitated by computerized feedback in more and less explicit conditions.

The current study focused on four main issues: 1) it investigated how Spanish HL participants processed input in their heritage language, and how it was similar or different than L2 participants; 2) it examined how the Speaker profile and/or 3) Type of feedback (i.e. +Explicit feedback, -Explicit feedback, control) had an effect on subsequent performance on assessments; and 4) it addressed the potential correlation between levels of processing on performances following the experiment. Levels of processing were measured by means of think-aloud protocols collected during the experiment. Performance was examined by way of two written assessment tasks - a controlled production and a semi-spontaneous picture description task.

Results showed that HL participants primarily employed non-metalinguistic strategies (e.g. grammatical judgment, prior knowledge) to process the input, whereas L2 participants relied more on metalinguistic strategies (e.g. mentions of previous lessons and explicit grammar). Additionally, although the HL participants significantly outperformed the L2 participants on both tasks at the immediate posttests, the significant difference was not maintained on the delayed posttest. Regarding the levels of explicitness of feedback, both the +EF and -EF groups significantly outperformed the control group on both tasks. At the same time, no significant difference was found between the +EF and –EF on both the immediate and delayed posttests. Finally, with respect to DOP, the +EF facilitated more instances of higher levels of processing than the –EF condition, and DOP was positively and significantly correlated with better performances at the delayed posttest.