Exploring alternate vehicles for DDL integration into teaching practicesa case for pedagogic natural language processing resources.

Resumo

Objectives. Academics have praised the benefits of Data-Driven Learning (DDL) across the literature for the last decades, yet the area seems to struggle with how to contribute to the development of the epistemological underpinnings of language theory. Scholars argue that, in order to prevent DDL as a field of study to become stagnant, DDL must expand its reach beyond tertiary education contexts and ‘normalise’ its usage across the wider L2 education community. Such efforts, however, appear hindered by the difficulties associated with managing corpora, especially for the untrained L2 teacher, who tends to show reluctance towards adopting corpus methodologies in the classroom. Thus, this thesis advocates for expanding DDL’s methodological reach to include digital, language processing tools available on the Internet as vehicles to implement DDL as a viable approach to L2 teaching in the form of pedagogical natural language processing resources (P-NLPRs). Methodology. The research carried out for this thesis comprises three separate studies developed within the framework of the ERASMUS+ project ‘Transforming European Learner Language into Learning Opportunities’ (TELL-OP): Study 1 presents a preliminary exploration on the uptake of P-NLPRs among European L2 educators. It analyses the results of a survey distributed to language teachers in Spain and the UK (n=230), aiming to ascertain the role of a high familiarity with the tools with regard to the potential increase in their frequency of use. Furthermore, the study also explores potential factors that may influence how familiar an individual is with P-NLPRs, such as previous background training in using mobile devices for educational purposes, or the explicit support from institutions towards using MDs in-class. Study 2 explores the implementation of a mobile app designed with built-in open source DDL-oriented tools for L2 learners from Spain (n=55), Belgium (n=70), and the UK (n=3). A mixed-methods approach was adopted to collect both quantitative data via questionnaires and qualitative data through focus groups. The results provided valuable feedback on how to tackle the potential implementation of DDL practices through mobile devices. Study 3 focuses on exploring the perspectives of L2 educators on the usage of P-NLPRs as vehicles to implement DDL. The research gathered data from the participants of a training module designed within the framework of the TELL-OP project, with the intent of providing basic training on the usage of P-NLPRs for teaching purposes. The approach to data gathering was mixed-methods, as quantitative data was collected through questionnaires prior and after the course, while individual interviews were conducted five years later to perform qualitative analyses of the data. Conclusions. The results suggest P-NLPRs can serve as vehicles for the implementation of DDL practices, as using these resources promotes the main elements DDL is based on: autonomy, individualization, induction, and authenticity. This has overarching implications in how DDL research can be approached in the future; methodologically speaking, P-NLPRs boost the potential approaches to implementing DDL by providing a greater variety of tools with multiple pedagogical purposes, contributing thus to the ‘normalisation’ of DDL in the L2 classroom. Epistemologically speaking, by including P-NLPRs as part of the DDL toolset, researchers have more resources at their disposal to study how DDL approaches may impact learning processes, thus cementing the contribution of DDL to areas such as Applied Linguistics or Second Language Acquisition.