Scottish Government

Evaluation Report: Early Partial Immersion in French at Walker Road Primary School, Aberdeen

1. INTRODUCTION

The present report covers the first two years of implementation of the 'early partial immersion in French' (EPIF) project of Aberdeen City Council, with support from the Scottish Executive Education Department, and located at Walker Road Primary School, Aberdeen.

Year One and Year Two

The section on Evidence from Year One contained within the present document focuses exclusively on the Primary One class with which the project began in October 2000. The main data were collected in June 2001, when the pupils were approaching the end of their summer term.

The section on Evidence from Year Two focuses on this same class when it was in Primary 2 and also on the Primary One class which had entered primary school in August 2001. The main data were collected in late May 2002, almost one year after the data-collection for the Year One report.

Varieties of 'immersion education'

'Immersion education' is widespread across the world and exists in several different varieties. Common to all of them are the following characteristics:

• The 'immersion' language is not the students' first language. It may be a 'second' language, e.g. a language spoken by other groups in the same country, or it may be a foreign language.
• The students not only learn the immersion language but also learn other important subject matter through the medium of the language.
• The teacher is a highly fluent speaker of the language, and very often a native speaker.
• The teacher uses the language extensively for a wide range of teaching and interactional purposes, thereby providing the students with substantial exposure to it and giving a strong impetus to the development of students' comprehension skills.
• The teacher does not force the students to speak the language, and initially allows them to speak in their first language if they so wish. When the students initially express themselves in the immersion language they tend to do so through songs, poems, games, phrases. Gradually, of their own volition, they begin to create their own spontaneous utterances in the immersion language.

The various varieties of immersion education differ from each other mainly in relation to the starting point (early - delayed - late) and the extent (total or partial). Thus there can be 'early total', 'early partial', 'delayed total', 'delayed partial', 'late total' or 'late partial'. These are summarised in Figure 1 below:

FIGURE 1: VARIETIES OF IMMERSION

The 'early' varieties tend to begin at pre-school level or at the start of primary education; the 'delayed' varieties tend to begin at some point between the ages of eight and fourteen; and the 'late' varieties tend to begin after that, including with adults. It is wrong therefore to assume that immersion education necessarily has to involve young children, although in the Aberdeen project it obviously does.

The research background to immersion education

A very large body of research has been conducted internationally on immersion education. The international research, e.g. Cummins (1999), Johnson & Swain (1997), indicates clearly that immersion learners make substantially better progress in the language than do learners of the language when taught as a school subject. In Canada for example all of the different varieties of French immersion consistently yield higher levels of proficiency in French than is yielded by 'Core French' (taught as a subject), with 'early total' immersion the strongest model.

The present report does not seek to present these international research findings. However, readers who are interested in the international research background to immersion education can easily find the present writer's international research review on the Scottish CILT website (see Bibliography), and they are entitled to download it free of charge. This was commissioned by the Scottish Executive Education Department.

Meaning of 'partial immersion' in the Aberdeen model

The Aberdeen model fits into the 'early partial immersion' box as in Figure 1 above. This implies that from Primary One onwards pupils are not only taught French as a language but that they also receive some of their more general primary school education through the medium of French, particularly the Expressive Arts. Views as to exactly how much time per day is given to partial immersion in French vary somewhat but at present it would be up to a maximum of one hour of French per day, supported by a trained and qualified primary school teacher who is a native speaker of French, working in collaboration with the pupils' classteacher, support staff and the headteacher.

However, there can be considerable variation even within the one model of 'early partial immersion'. The Aberdeen model differs for example from the variety of 'early partial immersion' that tends to operate in Canada and the USA. There, 'partial' means roughly 50% of the curriculum, and substantive subject-matter is introduced from the start. In one well-known partial immersion experiment in the USA (Thomas, Collier & Abbott, 1993), for example, the children were introduced to Mathematics from the first year of elementary (primary) school in a number of schools through partial immersion in French, Spanish or Japanese. By the end of the second year of elementary school they were achieving through their modern language the same state and national attainment norms for Mathematics as were achieved by matched groups of children doing their Mathematics through English as first language.

In the Aberdeen model by contrast, 'partial' means less than 50% and in Primary 1 is mainly directed to Expressive Arts. A distinctive feature of the Aberdeen model however is that it plans for the percentage of French to increase, with areas additional to Expressive Arts being handed over in whole or in part to French as the children progress through their primary schooling. ( Appendix 1 charts how this plan is intended to unfold). In comparison with models of 'early partial immersion' in certain other countries, then, the Aberdeen model may seem somewhat 'canny', but in comparison with what happens elsewhere in Scotland it is strikingly bold and imaginative.

The Aberdeen approach of less than 50% to begin with, but building up over time, allows space in the early years for pupils to establish a foundation of initial literacy in English with the aim of achieving a good level of literacy in both English and French by the end of their primary education. In the Aberdeen plan, Reading and Writing in French are due to be introduced as from Primary 3, though the present report shows that a start on this has been made even in Primary 2.

Differences between the Aberdeen model of early partial immersion and the general model of MLPS (modern languages at primary school) in Scotland.

The 'partial immersion' model adopted at Walker Road Primary School differs substantially from the much more widespread model of MLPS (modern languages at primary school) adopted throughout Scotland. Key differences between the 'early partial immersion' model and MLPS are set out in Figure 2 below:

FIGURE 2: 'EARLY PARTIAL IMMERSION' (ABERDEEN) AND MLPS IN SCOTLAND

Figure 2 strongly suggests two things:

• First, it is reasonable to expect a far higher level of proficiency in French from 'early partial immersion' students, as compared with MLPS students, by the end of their primary education, in view of the much larger number of hours of contact with French.
• Second, precisely because so much of the children's time is given over to French, it is important to monitor not only their proficiency in French but also the extent to which they are learning their general primary school subject matter through the medium of French. The research on Gaelic-medium primary education in Scotland (Johnstone et al, 1999) showed clearly that children receiving Gaelic-medium primary education, whose home language was English, did not lose out in their learning of science, mathematics and English at primary school but in fact made some gains. In many instances, however, the Gaelic-medium education was of the 'early total immersion' variety, and so it will be of interest to establish the extent to which 'early partial immersion' in French yields similarly impressive outcomes.

Conducting the evaluation

Thus far, a light-touch,'dipstick' approach

The evaluation is being conducted by Scottish CILT and was commissioned by the Scottish Executive Education Department (SEED). When SEED approached Scottish CILT with the commission, it was their view that the evaluation need not start until possibly Year 3 of the project, thereby allowing teachers, pupils and parents an opportunity to settle down in the project before anything was subjected to an evaluation process. The Director of Scottish CILT (the present writer) considered however that if possible the evaluation should follow the project from the start. His letter of 5 June 2001 to the Director of Education (Aberdeen City Council Education Department) states:

'Expert evaluation researchers are agreed it is best all round if the evaluation process can accompany the innovation from the start. Mutual understandings can then build up and the process can become 'normalised' as useful insights are regularly fed back not only to the funding body but also to the participants themselves. If an evaluation comes in at a later stage, it tends to be perceived as something different, unknown and possibly threatening and it can be difficult for the evaluators to understand why things are as they are, since they were not involved from the start.'

Permission was willingly given for the evaluation to collect data from Year One of the project, but in order to be as unobtrusive as possible the evaluator proposed that in the first two years or so it should be of the 'light-touch' variety. For these initial two years a 'dipstick' approach was therefore adopted, whereby one evaluation visit would be made towards the end of each year.

The evaluation is under the responsibility of Professor Richard Johnstone, Director of Scottish CILT, who made the evaluation visits and who is the author of the present report. Professor Johnstone was also the director of the SEED-funded evaluation of Gaelic-medium primary education in Scotland. He was also commissioned by SEED to undertake a review of the international research on immersion education (Johnstone, 2000), and he has been commissioned by the Council of Europe to prepare a chapter on early language learning for their forthcoming publication on policy-advice for educational decision-makers across Europe.

General aims of the evaluation

The general aims of the evaluation in its 'light-touch' mode thus far have been:

• To gauge the perceptions of immersion teachers, headteacher, other school staff, parents and representatives of Aberdeen City Council Education Authority as to how the project is progressing.

For example, is it meeting (or even surpassing) their plans, hopes and expectations, or are any problems beginning to surface?

• To build up an initial picture of what happens in the Aberdeen project, affecting pupils in P1 and P2.

For example, what sorts of special provision are made, in order to take account of the fact that it is based on early partial immersion? What sorts of processes seem to be engaged? Do the methods of teaching seem to be consistent with the international research findings on immersion methodology? What does this initial picture suggest about pupils' participation, learning, attitudes and language development?

• To identify in a preliminary fashion any differences between pupils.

For example, are any differences beginning to emerge between P2 and P1 classes, or between boys and girls, or between 'high-aptitude' and 'lower-aptitude' pupils? Were there any pupils with particular needs or disabilities, and if so, then to what extent do they seem to be gaining benefit from their participation?

Data-collection and data-analysis processes

On each visit, data were collected by two main approaches:

• observation of lessons;
• interviews with key stakeholders, especially the immersion teachers, the headteacher, parents and representatives of Aberdeen City Council Education Authority.

A third set of data consisted of the project documentation made available by Aberdeen City Council and the school. This gave the evaluator a good sense of the project's overall aims and plans.

With regard to the observation of lessons, the evaluator considered it would have been premature and over-invasive if these were to be audio-or video-recorded for purposes of subsequent analysis. Instead a frame of reference was constructed in advance, consisting of a number of 'guiding questions' which would guide the evaluator in the detailed hand-written notes he took as the lessons proceeded. These 'guiding questions' were:

• What main topics are being addressed? Do these draw, for example, mainly or exclusively on the Expressive Arts curriculum, as intended in the project plan?
• What are the main characteristics of the immersion teacher's language? Does the immersion teacher talk in different ways when teaching P2 and P1 classes?
• What are the main characteristics of the pupils' language? Do P2 pupils, for example, show signs of progress in their language development as compared with P1 pupils?
• What are the main characteristics of the interaction that takes place in class, e.g. between the immersion teacher and the class, or between the immersion teacher and particular groups or individuals, or between pupils themselves?
• Are there any observable and consistent differences between boys' and girls' participation?
• Are there any observable signs, however fleeting, of any pupils experiencing real difficulties or disaffection?

With regard to the conduct of the interviews with the key stakeholders, these were semi-structured. They allowed the evaluator to feed in a small number of 'guiding questions' appropriate to each particular respondent or group of respondents, but not in a fixed formulation or pre-determined sequence, and they allowed the respondents considerable latitude in expressing and developing their own thoughts. As the interviews proceeded, the evaluator took detailed notes but felt it would have been over-intrusive at this exploratory stage to request that the interviews should be audio-recorded for further analysis.

The detailed notes which the evaluator took as the lessons and the interviews proceeded were subsequently analysed carefully in relation to the 'guiding questions' indicated above, and this analysis is incorporated into the two 'evidence' sections of the present report: Evidence from Year One and Evidence from Year Two.

Thus far, data have not been collected:

• from the pupils themselves, e.g. by interview or questionnaire in order to gain their views;
• on the performance of pupils in tests, e.g. in order to measure their levels of proficiency in French or to ascertain how they were faring in relation to the national 5-14 levels.

The reasons for not collecting data from or on the pupils themselves (apart from observing them in their class lessons) is that the 'dipstick' approach does not provide sufficient time for this to happen, and that in any case the systematic collection of pupil-data at this stage would have run the risk of being intrusive at such an early stage in the project, particularly in view of the young age of the pupils. However, as the evaluation moves into Year 3, it will be desirable to collect a wider range of more systematic data on the pupils, and this aspect is discussed in Section 5 of the present report.