The BEST Dataset of Language Proficiency.

Angela De Bruin,Jon Andoni Duñabeitia,Manuel Carreiras

doi:10.3389/fpsyg.2017.00522

Abstract

The BEST dataset presents data from 650 adult participants from the Basque Country who completed a series of Basque, English, and Spanish language proficiency tests (hence the name BEST). The BEST dataset is the result of a collaborative project developed at the Basque Center on Cognition, Brain and Language (BCBL) with the aim of providing researchers from the Basque Country with a series of norms that can be used to characterize the test samples. The files contain background information about the participants (age, gender, maximum education level, and handedness). It furthermore contains the values per language for age of acquisition, self-rated exposure, self-rated proficiency, the scores resulting from an interview, the number of correctly named items in a picture-naming test, and the LexTALE scores.

Highlights

Researchers investigating processes stemming from or related to multilingualism often face the challenge of correctly characterizing their multilingual samples in terms of language use, proficiency, dominance, and exposure
The current study provides a compelling dataset of norms obtained from a large number of Spanish–Basque–English multilinguals from the Basque Country that will facilitate participant selection and classification as a function of their background and language skills
The current study focuses on multilinguals living in the Spanish autonomous community of the Basque Country, an area containing a wealth of bilingual speakers of Spanish and Basque, the two co-official languages, who know English as a foreign language to a certain extent

Summary

INTRODUCTION

Researchers investigating processes stemming from or related to multilingualism often face the challenge of correctly characterizing their multilingual samples in terms of language use, proficiency, dominance, and exposure This is typically done by using a variety of objective (e.g., normed tests) and/or subjective (e.g., self-reports) measures which tend to vary largely across laboratories and studies, since to date the existence of a comprehensive set of measures and norms is limited to certain language combinations (see Gollan et al, 2012). The BEST dataset is the result of a collaborative project developed at the Basque Center on Cognition, Brain and Language (BCBL) with the aim of providing researchers from the Basque Country with a series of norms that can be used to better characterize the test samples We present both the range of scores per task, the quantile distribution of these scores, as well as a cluster analysis grouping participants according to the different measurements. We present detailed information about these tests, which can be accessed together with the normative data via https://figshare.com/s/2b377367585a7e5353fb and http://hdl.handle.net/10810/20563

PARTICIPANTS

CONCLUSION

Findings

ETHICS STATEMENT