Learner attitudes towards data-driven learning: investigating the effect of teaching contexts

Concordance-based Data-Driven Learning (DDL) aims to help second language learners infer language usage rules from language usage regularities. A number of DDL pedagogical treatments have focussed on phraseological units such as collocations, widely recognised as a central component of second language learning. This study evaluates DDL effects from an emic perspective, reflecting the learners’ perceived usefulness of the approach, as opposed to etic perspectives, representing changes in language competence as a result of the approach. It compares a group of Chinese learners and a group of Belgian learners of Italian as a Second Language/ Foreign Language (SL/FL). The findings indicate that the Belgian students seem to have gained familiarity with the approach faster than the Chinese, though the latter seems to perceive greater long-term benefits of the approach, and are more favourable to future mobile phone applications. The study aims to shed light on possible learner-related differences in DDL treatments and on the insightfulness of emic data in assessing DDL effects.


Introduction
When investigating the effects of DDL, i.e. the direct and immediate use of corpus data by language learners (Leech, 1997;Meunier, 2010), one important perspective is the emic perspective, which aims to reflect the learners' personal attitudes towards the approach.
Previous research has highlighted the learners' positive attitudes towards the perceived relevance and authenticity of corpus data, and the empowering effect 1. University for Foreigners of Perugia, Perugia, Italy; luciana.forti@unistrapg.it; https://orcid.org/0000-0001-5520-7795 How to cite this article: Forti, L. (2019). Learner attitudes towards data-driven learning: investigating the effect of teaching contexts. In F. Meunier, J. Van  of the inductive learning that is often associated with DDL (Chambers, 2007). Nevertheless, little attention seems to have been devoted so far to the potential effects that different teaching contexts may have when adopting DDL materials and strategies.
This paper examines learner attitudes towards DDL for learning collocations, comparing a larger sample of Chinese learners and a smaller sample of Belgian learners of Italian as an SL/FL.

Participant samples and DDL treatment
All participants in this study were university students of Italian as an SL/FL. The larger sample of Chinese learners consisted of 123 participants, which were distributed in eight classes with a range of 13-17 per class. Their level, as assessed by the university placement tests, was pre-intermediate, and they were part of an eight-week long intervention, with one one-hour lesson a week. The smaller sample of Belgian learners consisted of 22 participants, divided into an elementary class of ten, and an intermediate class of 12, with an exposure of one one-hour lesson each. Various activities were developed for both samples on the basis of concordance lines extracted from the native Perugia Corpus (PEC) (Spina, 2014), and focussed on verb-noun collocations (see supplementary materials for a sample activity).

Questionnaire construction and administration
The questionnaire items analysed in this study are part of a larger questionnaire made of eight Likert scale items and four open questions. The results for the Chinese participant sample were partly published in Forti (2017). This paper reports on results for the four Likert scale items dealing specifically with features of the DDL treatment.
The Likert scale items were developed according to the recommendations contained in Dörnyei and Taguchi (2010), with items being formulated either positively or negatively in order to avoid single-sided choices. The values of the scale were operationalised as follows: 1, totally disagree; 2, disagree; 3, partially disagree; 4, partially agree; 5, agree; and 6, totally agree.
The questionnaire was distributed in class at the end of the pedagogical intervention, in pen-and-paper modality.

Results and discussion
Fifty questionnaires were collected from the Chinese students, while ten and 12 questionnaires were collected from the Belgian elementary and intermediate students respectively. For questionnaire items two and four, one Belgian intermediate student selected three values, so these two responses were removed.     In sum, Belgian students have less trouble with the groups of sentences, as they seem to feel not as confused as the Chinese. This might be due to French belonging to a language family that is genealogically closer to Italian, especially in terms of reading system. However, the Belgian students do not seem to perceive DDL as useful as the Chinese do, which might also be due to the vicinity with the language being learned, and perhaps with the very limited exposure they had to the approach.
On the other hand, Chinese students seem to be more favourable to the idea of a concordance-based app in comparison to both groups of Belgian students. This might be due to a possible increased familiarity that the Chinese learners have with mobile phones, in comparison to the Belgian students.

Conclusions
In this paper, we briefly outlined some results related to a DDL pedagogical intervention which took place in two different teaching contexts. The results are based on emic data, aiming to elicit the learners' perceptions in relation to the usefulness of the approach.
The findings indicate that the Belgian learners were generally more comfortable with the proposed concordance work, although the Chinese perceived its usefulness more distinctly. Furthermore, the Chinese students seem more favourable to the idea of an app with a concordance-based version of DDL, in comparison to the Belgian students.
The study would definitely benefit from more fine-grained and sophisticated statistical analyses, as well as from an integration of other open-ended questions which were also contained in the administered questionnaire. The hope is, however, that this study can shed light not only on the possible differences in assessing the effects of a specific DDL treatment in different populations of students, but also on the importance of emic data in gaining insight into how DDL is perceived by those who it is meant to help. Disclaimer: Research-publishing.net does not take any responsibility for the content of the pages written by the authors of this book. The authors have recognised that the work described was not published before, or that it was not under consideration for publication elsewhere. While the information in this book is believed to be true and accurate on the date of its going to press, neither the editorial team nor the publisher can accept any legal responsibility for any errors or omissions. The publisher makes no warranty, expressed or implied, with respect to the material contained herein. While Researchpublishing.net is committed to publishing works of integrity, the words are the authors' alone.

5.
Trademark notice: product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe.
Copyrighted material: every effort has been made by the editorial team to trace copyright holders and to obtain their permission for the use of copyrighted material in this book. In the event of errors or omissions, please notify the publisher of any corrections that will need to be incorporated in future editions of this book.