Language Testing

Although much more limited in scope and intensity than second language (L2) immersion, MLPS (Modern Languages at Primary School; known elsewhere as FLES, Foreign Languages at Elementary School) is being extensively implemented across the European Union as a reflection of national and European Commission (EC) policies. Despite an increase of research interest in the area, issues of MLPS assessment have not been widely addressed. There are good reasons for developing such assessments:

• gauging the return on a massive public investment;

• feedback to parents and the public;

• informing national policy development;

• self-evaluation at the level of institutions, departments, teachers and learners.

There are, however, major problems at present in doing so across schools:

• variability of context;

• embeddedness of children’s language in a flow of events;

• their relative lack of cultural knowledge;

• unfamiliarity of teachers with concepts of L2 testing;

• lack of consensus concerning what MLPS proficiency might mean.

Examples are given from Scotland of how some of these problems were addressed in the pilot phase of the national MLPS initiative of the early to mid-1990s, leading to the development of more comprehensive and standardized assessments in the national Assessment of Achievement Programme (AAP) intended for the late 1990s and early years of the twenty-first century.

The National Assessment Programme in Education (PPON) measures the level of performance on school subjects at the end of Dutch primary (elementary) education. Two assessments, in 1991 and 1996, of English as a foreign language have been conducted. This article reports on the outcomes of both assessments with regard to listening, reading and word knowledge. There is a slight decrease in pupils’ performances when the outcomes of the 1996 assessment are compared to the 1991 assessment. Large differences in performance exist between pupils from higher and lower social levels. Boys outperform girls. Time emerges as a rather stable and significant variable in foreign language teaching and learning. The assessments have provided important insights into the possibilities and limitations of foreign language teaching and learning.

Alternative assessment (as opposed to formal testing) is gaining a great deal of attention in current educational discussion. This article attempts to address some of the epistemic and practical issues facing alternative assessment, with particular reference to teacher assessment of spoken English as an additional language/second language (EAL/ESL) in the early years of primary (elementary) education in England and Wales. We first examine the claims made by advocates of alternative assessment in terms of validity and educational relevance. It is argued that such claims are founded on an uneasy articulation of different principles underpinning psychometric measurement and pedagogy. Next we look at some of the reasons why psychometric approaches may not provide an adequate response to pedagogic and policy developments. Then some of the theoretical and practical problems involved in teacher assessment of speaking, focusing on learners with EAL in primary education, are discussed. We focus on the importance of clarity about the epistemological bases of different types of assessment. Additionally, the article highlights the need to be alert to the ways in which political and ideological concerns - together with the influence of professional (teaching) culture - are influential in shaping the properties of assessment systems.

The recent introduction of literacy benchmarks and assessment regimes in schools which are designed for students with an English-speaking background but which are also used to assess the second language and literacy progress of ESL¹ learners has repercussions for teaching and learning, on professional understandings and on ESL learner self-efficacy. Insights into how to map and monitor ESL language development within the mainstream context have been gained through the development of a number of ESL standards for schools. This article presents principles behind the construction of ESL standards for schools, drawing on examples of ESL standards developed in Australia, England and Wales, and the United States, and examines how differences in purpose in these standards - planning, professional understanding, and reporting - influence how ESL standards might best be constructed. It raises issues of consequential validity which arise in the construction and use of valid ESL standards. Arguments supporting the need for ESL standards in the current climate of English-speaking background-based assessment for all are put forward through a close analysis and understanding of the construction and impact of different types of ESL standards.

This article explores the nature of formative assessment in a primary (elementary) language learning context. The research is situated in nine inner-city schools where an Early Years Intervention Project is being implemented to address problems of low levels of achievement in English, with specific reference to the language support of learners for whom English is an Additional Language.¹ School-based assessment data are presented and analysed in relation to the construct of formative assessment. It is argued that the distinctions between formative and summative assessment are not as straightforward as sometimes portrayed and that the interplay between reliability and validity for purposes of class-based assessment is highly complex.