Engelsk institutt

Facts about the EVA corpora of pupil language

The EVA corpora of pupil language were compiled by the EVA Project in cooperation with the HIT Centre, University of Bergen, http://www.hit.uib.no, who were responsible for putting the pupil data into electronically searchable form. The material consists of three parts: two corpora of spoken language (one from Norwegian and one from UK pupils) and a corpus of written language. These corpora can be accessed via query forms, using TACTweb searching, at the following HIT Centre websites:

http://kh.hd.uib.no/eva/eva.htm (Norwegian pupils: spoken)

http://kh.hd.uib.no/eva/eva-uk.htm (UK pupils: spoken)

http://kh.hd.uib.no/eva/eva-w.htm (Norwegian pupils: written)

There is a user name and a password needed in order to access the corpora. These are intended to ensure that the corpus is only used in the interests of education and research. They are obtainable from angela.hasselgren@eng.uib.no

The EVA speaking and writing corpora (Norwegian pupils) were compiled from material produced by 14-15 year old pupils taking the speaking and writing tests developed as part of the EVA Project in spring 1994. The UK pupils' speaking corpus was compiled from recordings of 14-15 year old pupils in the North of England carrying out the tasks of the EVA speaking test.

The corpora are of the following sizes:

EVA Speaking Corpus, Norwegian pupils: 62 pupils,34,544 words

EVA Speaking Corpus, UK pupils: 26 pupils,17,629 words

EVA Writing Corpus, Norwegian pupils: 93 pupils,50,170 words

The material is tagged for the following variables: task, personal identity (ID), school, tape, gender and test grade. (The school and ID are kept anonymous through the use of codes). Words or expressions (or word forms) may be searched through the query form. It is possible to restrict the search to a subgroup, eg all pupils over a certain grade. The way a word is used can be displayed in context or in terms of the way its use is distributed according to variables, eg its relative use by boys and girls. While it is possible to see large parts of the context in which an item occurs, it is not readily possible to obtain whole transcripts.

The corpora are of value both to teachers and researchers of school English. The spoken UK corpus provides a rich source of examples of words and expressions in native speaker use, which can be contrasted with Norwegian pupils' use. The study of the Norwegian corpora can highlight weaknesses and characteristics of pupil English.

Facts about the EVA project

The University of Bergen, English Department was given the task in 1993, by the Norwegian Ministry of Education, of developing a system for evaluating the communicative English language ability of pupils in the Norwegian school system. The work has being carried out in the framework of the EVA Project. Material is developed in response to the need for:

means of diagnostically assessing pupils' language ability over a range of skills;

concrete criteria for assessment of this ability, applicable at all levels of formality;

systematic documentation of ability;

a way of enhancing classroom assessment processes, including self-assessment.

The material for both 6th and 9th grades has been published by Nasjonalt Læremiddelsenter, and is available for use in schools. This material contains not only tests, but material for ongoing classroom assessment, for use by teachers and pupils. It also contains booklets for teachers with advice on using the material, and on following up assessment with measures for building pupils' skills

A significant aim of the project has been to enhance research into the development of English in Norwegian school children. Doctoral and hovedfag research has been carried out using the data which have emerged from the project.

For more information, please contact:

Dr Angela Hasselgren