Change in Language of Instruction in Russian Medium Schools: Multilevel Analysis of Attitudes and Language Proficiency

Abstract

This study aims to explain the positions regarding change in language of instruction in Russian-medium schools. The empirical basis for this article is a representative survey conducted in 2009 among pupils (n = 1026) and teachers (n = 548) in Estonia. The results indicate that in the case of teachers’ language proficiency, individual resources (citizenship, Estonian-language media) dominate over regional (ethnic composition) ones. Regarding pupils’ language proficiency, apart from individual factors (gender, Estonian-language media), school-level activities are essential (language immersion, teachers’ in-service training). While pupils’ positive attitudes towards education change are characterized by individual and school cultural resources, teachers’ attitudes are more autonomous and express general societal affiliation. The article suggests that the “cultural translation” ability of the relevant individuals and in the public turns out to have critical relevance in the context of implementing integration policy and in the communication of changes in education.

Keywords:

• educational change
• language proficiency
• language attitudes
• language of instruction
• social geography
• multilevel modelling
• language policy
• Estonian Russian-speaking population

Acknowledgements

The data analyzed in this article were collected within research project SHTHT07261 financed by Estonian Ministry of Education and Research. This research was supported by the European Union through the European Regional Development Fund (Centre of Excellence in Cultural Theory), by Estonian Research Council grants ETF9308, ETF8347, and IUT20-38. The authors are thankful to the anonymous reviewers for their profound comments, which have helped in completion of the article.

Notes

1 In the present paper the Russian-medium and Estonian-medium schools signify the main language of instruction used in the school. However, in Estonia since the academic year 2011–2012 there has officially only been Estonian-language education at the upper-secondary level (with some exceptions), although 40% of the subjects are still taught in Russian in formerly Russian-medium schools.

2 The main documents regulating language policy in Estonia are the following: Estonian language development plan 2011–2017, Estonian language strategy 2011–2017, Estonian Integration Strategy 2008–2013. Based on these strategic documents, the issues of minority education are reflected in specific area regulations like Basic school development plan 2007–2014, Basic and secondary school act, Pre-school education act, State curricula for basic and secondary school.

3 There has been increasing public discussion about the transition to Estonian-language instruction before parliamentary elections (e.g., 2011, before the final deadline of transition). In addition, changes in government structure have brought the Russian-medium schools to the fore as a topic: e.g., Mailis Reps, of the Central Party, taking into consideration the interests of the Russian-speaking electorate since becoming Minister of Education and Research in 2005. She criticized the absence of state-led programs related to Estonian-language instruction at Russian-medium schools. At this time the initial deadline of 2007 for transition to Estonian instruction was replaced by a more realistic 2011–2012 (with the exception of some schools which were granted extensions). On the other hand, since the main responsibility of transition lies with the school owners, discussions and solutions at the local government level are quite different regionally (e.g., resistance to the change in language of instruction and public discussion about private Russian-medium schools as a solution in Tallinn and north-east Estonia in 2011 and 2012).

4 In the present paper the terms context and social space are used as synonymous in a similar way to studies in the field of education changes (Hargreaves 2005; Hernandez and Goodson 2004), indicating characteristics on various levels (individual, school, region).

5 The number of pupils learning in language immersion classes has grown year by year, although the number of pupils learning in Russian-medium schools has increased, at the beginning of 2000s constituting 24%, but by 2010 constituting 20% of the total number of pupils; the number of pupils learning in language immersion classes has grown from 1.5% in 2005 to 3.4% in 2011.

6 In Estonia every teacher should participate in training for at least 160 hours over a five-year period. The themes and topics should be in accordance with the developmental plan of every school, but also with general priorities. In the field of bilingual education the in-service training is planned and organized at the state level, and is fully free for schools and teachers. At the same time, the choice of which in-service trainings to participate in is up to schools and teachers.

7 There have been no major legislative or strategic changes in the field of Russian-speaking education since 2009, after this study was conducted. At the same time, another study conducted in 2011 indicates that there has been a significant change in the attitudes and readiness to begin with the more intensive Estonian-language learning and teaching in kindergartens; also among Russian-speakers the support for Estonian-language or bilingual kindergartens has grown (Masso, Kirss, and Kriger 2012).

8 Here, self-estimated level of language proficiency rather than objectively measured language proficiency is used as one of the main dependent variables. Although the measurement scales may be understood differently by various social groups, previous studies indicate the self-estimated and actual language proficiencies correlate significantly since Estonian Russian-speakers have various everyday opportunities to prove their level of language proficiency, such as with official language tests for teachers, and pupil school examinations (r = .4; see Masso 2009a).

9 The general model used in this analysis could be summarized with the following formula: where language proficiency y_i(jk) of the ith child from the (jk)th school/area combination is modelled by the overall mean α together with a random departure u_j due to school j, a random departure u_k due to area k, and an individual-level random departure e_i(jk). In the present study, higher-order interactions are not used in the analysis as suggested in the case of complex multilevel models (Hox 2010). Although in the present analysis we have set up a three-level MLwiN structure, conceptually we have only a two-level model, but with region and school crossed at level two (for details of cross-classified structure, see Rasbash and Goldstein 1994). Cross-classification enables us to take into account influences from two different “contexts” – place of residence and school environment – and in this way to ascertain the importance of the two classifications in analyzing individual positions regarding change in language of instruction.

10 On the one hand, for estimating the accuracy of the model, the single and multilevel models are compared in the analysis. On the other hand, the chi-square likelihood ratio test is used for assessing the differences between models. Also the indicator of standard error is used to estimate the variance over sample size (the smaller the standard error the more reliable the results).

11 Indicators, such as the total number of pupils at the school, which were included in the teachers’ analysis in order to make a comparison with pupils are neither theoretically nor statistically significant when explaining teachers’ Estonian language proficiency.

Additional information

Notes on contributors

Anu Masso

Anu Masso is senior research fellow at the Institute of Journalism, Communication and Information Studies, University of Tartu. Her main research interest is personal social space (e.g., perceived cultural distances, geo-cultural mobility, and its relations with linguistic capital) in Estonia. Her second interest is research methods (quantitative and qualitative) in social sciences.

Maie Soll

Maie Soll is a doctoral student at the Institute of Journalism, Communication and Information Studies, University of Tartu. Her main research interest is language education of the minorities and the development of ethnic identity of pupils. Her second interest is pupils’ informational sources in the context of educational innovations.