Introduction | Research papers Research projects | Statistics | Language Trends

New Research into residence abroad

Dawn Ebbrell reports on this research seminar held at CILT on 11th June 2001, organised by the Residence Abroad Project with the support of CILT and the Subject Centre for Languages, Linguistics and Area Studies.
(This report was first published in the CILT Higher bulletin, issue 3.)

With speakers from the UK and Germany, the day focused on a wide range of issues surrounding residence abroad including intercultural experiences, methods and tools for measuring language acquisition and attitudes looking at both quantitative and qualitative methodologies.

Ulrich Raatz (Duisburg), Rüdiger Grotjahn (Bochum) and Verena Wockenfuss (Duisburg) of Project TESTATT (Bochum, Duisburg, Portsmouth) presented the results of the use of scales and tests for measuring important factors in student development during the period of residence abroad. Factors covered were anxiety, introversion and extraversion, attitudes, motivation and C-tests (in German, French, English and Spanish). Tests have been based to varying extents on tools used in earlier studies e.g. the European Language Proficiency Survey 1993-1995). All scales adopted a 4 point Likert scale; C-tests consisted of sets of 6 texts each with 20 gaps.

332 responses from students at the universities of Duisburg, Bochum and Portsmouth were collated and analysed using SPSS to establish the reliability of the instruments and their concurrent validity.

Compared to students at Portsmouth and Bochum, German students in Duisburg were found to have a lower level of language anxiety. Across groups high anxiety scores correlated with lower levels of extraversion, lower C-test scores, lower motivation and less positive attitudes towards the target language (L2). A factor analysis of the anxiety scales showed 2 factors: emotional reaction and self-confidence. Results also showed that high levels of extroversion were associated with low levels of anxiety and high levels of motivation. Within extroversion three clear factors emerged sociability, spontaneity, and shyness. On the attitude scales, items were taken from the European Language Proficiency Survey (ELPS). Positive attitudes towards the native language (L1) tended to be linked to positive attitudes towards the L2. Positive L2 attitudes related to lower anxiety levels and higher motivation and C-test scores. Factor analyses of the L1 and L2 attitude scale revealed 3 factors for both scales: pro-social behaviour, egocentric behaviour and competence. With respect to motivation, German learners were found to have a higher level of motivation for language learning than their English counterparts. There were no significant differences within the groups. Overall, the TESTATT project has developed valid, reliable and practical research instruments which have already produced interesting data.

Richard Towell (University of Salford) reported on a longitudinal study examining the progress of 12 undergraduates over a 4-year period. The student group was assumed to be homogenous having all followed the British secondary system at the same time. It was assumed that fluency depends on computational linguistic knowledge, associative linguistic knowledge and the ability to use these two types of knowledge. Acquisition and development depend on acquiring the knowledge and turning the knowledge into production systems. It was expected that exposure provided during residence abroad would enable the learners to develop their fluency as they proceduralise the knowledge they possess but that the L2 would not be spoken as fluently as the L1 even after residence abroad since their L2 knowledge is proceduralised to a lesser extent. A series of tasks were developed to measure fluency as described below. Data was collected prior to and immediately following the period of residence abroad and as tasks in English at the end of the final year of study. Research measures were:

• global fluency: speaking rate in syllables per minute measured on a cartoon re-telling task

• working memory related fluency: % of potential recall of forms measured on a shadowing task where students were required to listen and speak at the same time thus making use of working memory without being able to make use of conscious control.

• storage related fluency: % of potential recall of forms on a mini-story retelling task. The investigator read out the story and the student had to retell it immediately. As the stories were long, learners needed to use their own proceduralised language in retelling it.

Students who scored lowest prior to residence abroad increased their scores on the tests but did not reach the level of those who were initially more fluent. From a methodological point of view, it was felt that experiments need to be devised which will separate out linguistic knowledge and proceduralisation as separate influential factors and that reaction time experiments may reveal more than the mini-story retelling task.

Anne Ife's study (Anglia Polytechnic University) looked at the impact of the period of residence abroad on the vocabulary development of learners of Spanish. Previous research has shown that more advanced learners do not progress as much as lower level learners. The research tool used was the Association Vocabulary Test (AV3T) developed by Gemma Vives Boix and Paul Meara at Swansea. The test is designed to measure the interconnectedness of words in the learner's lexicon. The more connections present then the closer the lexicon is assumed to be to that of a native speaker lexicon. Learners responded to 3 sub tests each containing 40 items. Each item consists of a group of 3 words and learners must decide which of the 3 is not associated with the other two; the scoring method alters the score by estimating the number of guesses made. The experimental group included two successive groups of British students going to study abroad in Spain for between 5 and 10 months. The groups were divided between advanced and intermediate level. All students took the association test and a translation test (translating words from the test) both before and after their period of residence abroad. It was expected that intermediate students would progress more on the translation test and advanced students would progress more on the association test. Results indicated that all of the students made progress on the translation test if they went abroad for two semesters rather than one, advanced students made more progress on the association test. The gap between the intermediate and advanced level students did not narrow following the period of residence abroad. The longer students are away the more progress is made although this effect is more noticeable for the intermediate groups.

Paul Meara (Swansea) addressed the issues of whether reported loss of vocabulary knowledge by students is in fact real, what can be done about it and whether teaching practice is to blame. There is an inherent difficulty in using control groups to investigate residence abroad since the control group would have to remain at home. Meara suggested a way of dealing with this by making the problem something that occurs within a group rather than between groups by using matrix models to make predictions about how vocabulary knowledge will develop and testing that prediction. In the case of vocabulary development there is no value in basing predictions on small numbers of items. A system has been developed at Swansea to set up a 15 minute test of 400 words from a set containing items between 4000 and 6000 in a frequency list. Students make decisions about the level to which they feel they know each word from 4 choices ranging from Don't know to Sure I know. The test generates transitional probabilities for items moving between states. Testing to date would suggest that that the matrix model is accurate to within 10%. Pilot tests suggest a decline in vocabulary knowledge from the year abroad, proficiency fell during the final-year study despite language teaching

Robert Crawshaw (Lancaster) reported on qualitative research which has been done on the nature of intercultural interaction during the period of residence abroad. Data was collected from 250 students using diaries, focus groups, and interviews. A taxonomy was developed to analyse the data with the aid of NUD.IST (Non Numerical Data. Indexing Searching and Theorising), a software package designed to aid the analysis of qualitative data. The use of the software to locate and label text units illustrating concepts when applied to a large base of data allows anecdotal evidence to become more grounded. For example, when examining statements about stereotypes it was found that the majority of generalisations were hedged; students only tended to make generalisations if prompted and generalisations were often qualified by statements. Through examining data from diaries it was possible to trace students personal development in adjusting to their new environment over a period of time.

Jim Coleman (University of Portsmouth) reported on the results of a recent project which built on the results of the 1993-95 European Language Proficiency Survey (ELPS). In this latest survey (1997-2000) items on proficiency, self-assessed confidence, reasons for study and attitudes were re-used from the ELPS. The project involved over 2325 respondents from 18 universities, 80 students in a longitudinal study and used interviews and focus groups as well as questionnaires. The new survey examined 7 ages of residence abroad: pre A-level, end of year 1 and 2 at university, during residence abroad, the start and end of the final year of study and language graduates in employment.

Some of the findings of the study are outlined as follows. There was a strong correlation between the views of staff and students on the purpose of the residence abroad. Improved proficiency in the target language, insight into society, institutions, way of life, increased personal maturity and independence and increased employment skills or personal transferable skills were all rated highly by both groups. There was a strong correspondence between perceived importance and the level to which specific objectives were felt to have been achieved except in the area of linguistic development and employability gains. Scores on C-tests showed great improvements between pre- and post- residence abroad with gains having diminished by the end of year 4. The level of embarrassment over errors becomes much less acute following residence abroad. Anxiety levels seem to rise from school to university, falling sharply during residence abroad and rising on return to university study. Reasons given for study are much the same as in the previous survey but with a slight increase in integrative reasons, similarly there is little change in the reinforcement of negative stereotypes during residence abroad. Particularly encouraging are attitudes towards employability. Over 75% of the language graduates surveyed said residence abroad helped them get both their first job and subsequent jobs with over 90% saying that skills gained through residence abroad were of direct benefit in their job. 99% said residence abroad was a worthwhile investment in time and money. Large proportions also rated residence abroad as the most important part of their degree (54.5 % of students at the University of East Anglia and 70.4% of students at Portsmouth).