Task design in computer-based L2 listening

Abstract

The proposed multi-phase study introduced here seeks to put forward a theoretically-and-empirically-based framework for the design of tasks for computer-based L2 listening environments. After a thorough review of the literature, in the present proposal, we argue that the construction of a framework for the design of computer-based listening tasks should not only consider the personal goals of the listeners, but also it should be context-appropriate. Accordingly, we will develop our proposal in three phases. In phase one, listeners goals will be elicited and used to establish context-appropriate criteria for the design of computer-based L2 listening tasks. The participant in this phase of the study will be a total of 48 students from different public/semi-private high-schools located in the second, fifth and eighth region in Chile. The systematic analysis of the focus groups data will foreground a preliminary framework for the design of computer-based L2 listening tasks by integrating Ellis (2009) criteria for the design of tasks for language learning, González-Lloret (2014) criteria for the design of technology-based TBLT, listening pedagogy and the emerging criteria for task design in computer-based L2 listening environments. In phase two, we will validate the framework with worldwide experts to do with L2 listening comprehension, computer-based listening and three novice listening researchers from Chile. The inclusion of Chilean researchers will help foreground the proposal for the Chilean context. Upon framework construction, in phase three we will describe and design sample computer-based listening tasks for learners of English at low proficiency levels.

We also design a task acceptance instrument that will be used to measure levels of acceptance of the proposed tasks. Two reasons motivate the current study, one at the contextual level and the other at the theoretical level. Contextually, in Chile as in most English as a Foreign Language Context (EFL) listening is problematic. After high-school students, primarily from public and semi-private schools, take 4 or up to 7 years of instruction in the foreign language, they are unable to comprehend spoken English. This situation has become evident from 2010 onwards as the results of the SIMCE-Inglés have been consistently low, 49 points on average. With such low results, the national goal of making of Chile a bilingual country by 2030 becomes a simply farfetched ideal and tangentially calls for theory-informed and context-bound measures to address this issue.

At the theoretical level, a thorough review of the literature shows that existing frameworks for task design in SLA (i.e Ellis, 2009) and in technology-mediated TBLT environments (González-Lloret & Ortega, 2014) fail to cater for students’ needs and wants. They have also been constructed primarily using a top-down approach where experts rely on theories, but where the voice of the language learner is not necessarily taken into account. Also, this piece of research would address recent calls to computer-based listening researchers to derive pedagogical insights and/or design guidelines drawn from empirical research (see Cross, 2017; Hubbard, 2017).

The expected contributions of the proposed study are: (1) a validated theoretically-and-empirically based framework for the design of computer-based L2 listening tasks; (2) a sample of computer-based L2 listening tasks and (3) a task acceptance instrument to assess the quality of computer-based L2 listening tasks.

A Preliminary Framework for the Development of an Interactive Virtual Tutor for Second Language Listening Comprehension

Abstract:

Nowadays, second language listeners seek not only to understand the target language speaker but also to interact with the innumerable aural materials available on the Internet and to become active members of web 2.0 communities such as YouTube, audio blogs and podcasts. This shift in listeners’ goals has led to an investigation on how to improve computer-based L2 listening materials. One focal area of research concerns the use and design of help options. Thus, L2 listeners interacting with computer-based materials can replay, forward and rewind aural texts, read along from transcripts and translations, consult definitions in online dictionaries, understand the contexts of the texts in the cultural notes and assess text comprehension with the feedback. Despite the increasing opportunities for learner-computer interaction, L2 listeners tend to neglect or use help options ineffectively.

CALL scholars agree that training is paramount to address this non-use/ineffective use of help options. The way this training should take, however, is still unclear. Some researchers suggest training should be done in the language classroom (Hubbard, 2009; Garrett, 2009; O´Bryan, 2008) and others through Interactive Virtual Tutors (IVT) (Heift, 2006; Cárdenas-Claros & Gruba, 2009; Cárdenas-Claros, 2011). This proposal addresses the second suggestion.

An IVT, for the purpose of this study, is a parser that by analyzing aural input highlights potential obstacles, at the linguistic level, to understanding and suggests the most adequate input enhancement (help option) to the student.

Listening scholars acknowledge that multiple factors influence comprehension of aural materials, but the IVT for L2 listening comprehension will only focus on linguistic features of texts given the elusive nature of other factors (i.e. schemata, memory, attention, etc.). The complete deployment of an IVT for L2 listening comprehension is a multi-disciplinary endeavour that requires multi-staged research, but the primary data for its conceptualization relies on analyses of language.

This three-year qualitative study seeks to build up a theoretical-and-empirically based framework to guide the future development and implementation of an IVT for L2 listening comprehension materials. Accordingly, this qualitative study is structured into three phases: Phase 1 seeks to identify and describe obstacles, at the linguistic level, that negatively affects the comprehension of L2 listening materials. The participants will be twenty adult learners of English from Chile, who will individually interact with five listening tasks. Upon interaction, learners will be guided to reflect on the difficulties, at the linguistic level, they experienced to understand the aural texts. Verbal recalls will be stimulated through recorded interaction of the participants with the listening tasks. After cyclical analysis of data, obstacles will be identified and documented. The second phase seeks to construct and validate a taxonomy of obstacles, at the linguistic level, that affect L2 listening comprehension. The validation component of this phase will be conducted with 7 scholars with expertise in L2 listening comprehension. They will individually be asked to evaluate the taxonomy taking into account four main criteria: obstacle description, relevance of the obstacle, impact, and order in the hierarchy. Kappa de Fleiss calculations will be performed to establish inter-judges reliability and the experts’ suggestions will be incorporated to refine the taxonomy. The last phase seeks to design and validate a theoretically-and-empirically-based conceptualization that establishes the relationship between obstacles for L2 listening comprehension and potential help options that the IVT could offer to assist listeners’ comprehension of L2 texts. Similar to Phase 2, scholars with expertise in CALL, HCI & ITS will be asked to evaluate the theoretical proposal. The main criteria for evaluation will be theoretical soundness, practicality, usability, design transferability and potential for interaction.