Language assessment for the workplace

2014

The present document is the result of a literature-based study of recent developments in the field of language requirements and language testing for immigration and integration purposes. The study was carried out as part of the regular research programme 2011-2014 at the Swiss national Research Centre on Multilingualism (RCM)1. The introduction of language requirements and the use of formal language tests as prerequisites for obtaining entry visas, residence permits and/or citizenship are fairly recent but rapidly evolving practices in many Western countries, particularly in Europe. These requirements and assessments have been made the object of academic debate and research, and the subject of several scholarly publications. This report seeks to present a structured overview of several aspects of these requirements and assessments; the basis for our study is provided by a corpus of academic publications. The summary of literature put forward in this paper is divided into two main ch...

Describes work in teaching and testing foreign languages for personnel working in the Teleservices Industry in Ireland

In many multilingual and multicultural countries it is an acceptable and standard practice to translate or adapt tests into more than one language. On the other hand, tests are often available in only one language, but are used across more than one language or cultural group. In South Africa (SA), as is the case elsewhere in the world, it is a common practice to follow this approach, especially when tests are used for admission into tertiary education, or in employment testing. This study aims to assess the fairness of this practice. After conducting a comprehensive survey of the literature to motivate the focus of the empirical study, a test of reading comprehension, available only in English, but which is used across language groups to decide about admission to university, was evaluated, as a case study, for bias and valid use across diverse language groups. A convenience sample (N = 563) consisting of three language groups, Afrikaans (n = 181), English (n = 187), and an African-language group (n = 195), was selected for the study. The results of the study indicated high levels of bias. It was found that these problems remained even after the removal of problematic items. Implications for fair testing practices in a multilingual SA will be discussed.

Melbourne Papers in Language Testing (MPLT), 2010

PEOPLE: International Journal of Social Sciences, 2017

Foreign Language Annals, 1984