John Bald/language and literacy

Assessment in language learning - the key skill, and how to focus on it.

As we learn a new language, we adjust and extend our thinking from the structures we have learned while learning our first language. These involve, in the initial stages, big changes in pronunciation and in our understanding of the use of letters to represent sounds in words. The primary toolkits on this site show how these can be introduced and explained in the first few lessons - and in particular in the very first lesson, where colours illustrate most of the main changes, in a context that is self-evident.

To communicate using the new language, however, we need to know how to put our words together to make a sentence, and to understand sentences that we hear or read. This can be achieved by building up sentences based on personal experience and preferences, and gradually extending these to encompass as much of the scope of the language as a learner wishes, or is able, to know. This sentence construction uses grammar for communication, rather than as an end in itself.

The techniques in the toolkits, if possible supported by the teacher's own resources using Clicker,  enable each child to learn to compose and understand sentences, without the illusions created by copying. The approach is very easy to use, and builds assessment into the teaching, so that there is no extra work for the teacher. Any system of assessment that does not focus on sentence construction is giving a false picture of a child's ability to understand and use the language.  I'm happy to discuss this further.

Assessment without Levels Commission Draft Report. Ofsted

A government commission on assessment following the abolition of levels has completed its work, and a draft of its report has been leaked to The Guardian. A link is half-way down this article. It is well worth reading.

The commission was attended by an Ofsted representative, and includes a statement of "Myths" about Ofsted's expectations of assessment. Unfortunately, these myths are close to reality as far as too many teachers and school managements are concerned, and the problem stems from an approach to inspection that is based on unreliable data. Teachers are being required to write down far too much, and this must stop.

 The commission's members were:

• John McIntosh, CBE
• Shahed Ahmed, Elmhurst Primary School
• Daisy Christodoulou, Ark
• Professor Robert Coe, Durham University and the Centre for Evaluation and Monitoring
• Sam Freedman, Teach First

• Mark Neild, Sir Isaac Newton Sixth Form
• Natalie Packer, National Association for Special Educational Needs
• Dame Alison Peacock, The Wroxham Teaching School

Assessment without NC levels: Dr Rachel Hawkes

Dr Rachel Hawkes has produced a framework of ten steps to language learning that teachers may find invaluable in meeting the demands of senior management for evidence of progress. The ten steps have important features, such as writing words and phrases from memory rather than copying, that are consistent with the approach to languages set out on this site, and can easily be adapted to meet individual interests and requirements, as, like most of the materials on Rachel's site, they are free to download. This may well have save a lot of teachers a great deal of work and stress.

The one element that I would personally change is in the degree of accuracy required in writing in the initial stages. The approach on this site is based on the idea that children learning a new language need to adjust their thinking so that they can use new structures, and part of this is understanding the different relations of letters to sounds. Having them practise single words and then sentences, using tracing on the sleeve and mini boards, so that children only write in a book what they know they will get right, builds confidence more effectively than simply accepting errors. Children also make fewer errors when they fully understand what they are trying to do, and why. A little practice in the early stages saves a great deal of frustration later on.

"The intellect is developed quite naturally, without being forced." A compliment unpacked.

This, from a parent in Falkirk, is as fine a compliment as a teacher can receive, and gives an insight into the principles behind successful teaching. It is easy to explain, and, once the principles are understood, easy to put into practice. Here's how:

Once we understand that learning is based on the formation of networks in the brain, we can use discussion and analysis of children's written work to see what they do and don't know and understand, and what they may be able to learn next. We use this knowledge to design materials and teaching - it does not always need to be written down.

Around what children already know, we can choose from many routes. Some will involve practice to strengthen existing netwoks - part of this is developing memory - and give children quicker access to what they know. This is important in mathematics, where children need to apply their knowledge in solving problems, and in English, where they need to be able to spell quickly and accurately in order to concentrate  on what they want to say.

Some will introduce new material, and this gives important choices. Some will simply add new material, such as facts or data. Some will open up new networks. In languages, learning vocabulary on a thematic basis is an example of the former. Learning patterns in the construction of words, and developing an understanding of words shared with the learner's first language is an example of the latter. It includes facts and data, but also leads to new networks. Just learning, say, the words for the contents of a classroom or a pencil case, has no application beyond itself. Seeing how a word may have moved from one language to another, with some alterations, or how a verb pattern may keep recurring when we want to say different things, builds networks which extend themselves - as we use more verbs - and also reinforce themselves, as the same patterns keep recurring.

Working with individual children, we can, unless we are obliged to follow some government "strategy", make a judgement on what the child needs to learn next, and exploit their own interests as source material. We are constrained in this only by our own knowledge, which we need to keep extending. With classes, we have the difficulty of hitting more targets at once, but this can be made manageable by grouping children according to their learning needs, and tracking progress by building it into a scheme of work.

All of the teaching here, including the German, is based on these ideas.

 

Assessing Languages without Levels.

At the London Osiris conference in July, I promised to write a framework for assessing progress in languages without using the levels set out in the former national curriculum. Here is a working draft, on which correspondence is invited.

Three reasons to change.

1. The NC levels are botched. Originally, all subjects had ten levels.  For most, these covered Y1-9, but for languages the 10 levels were squashed into Y7-9 - a third of the time. This led to substantial overlap between the levels, and led schools to set artificial targets. Some suppose children start at whatever level they have in English, whether or not they know a word of their new language.

2. The NC levels do not reflect the patterns of progress children really make in when learning languages. This is a fatal weakness. Schools that have used levels successfully have rewritten them, including their own sub-levels. As these are based on observation of children actually learning, they are more accurate than the original NC, and if schools have produced accurate levels for themselves, there is no reason to change.  

3. Levels bear no relation to the European Common Framework of Reference for Languages, agreed by the UK government and used for assessment in all EU organisations. Failure to use the Common Framework puts UK pupils at a serious disadvantage when applying for EU jobs - if they don't know what the requirements are, they have little or no chance of meeting them. The Framework is an essential tool for assessing languages, and is downloadable free, here. 

A Way Forward.

1. The Framework has three broad levels:

Basic User A1 (Breakthrough) - A2 (Waystage).  British pupils assessed as part of the Early Language Learning in Europe Project  have been assessed as working at level A2, though this was exceptional. Pupils starting from scratch, or starting a new language, might consider Level A2 as a suitable target by the end of Year 8.

Level A2+. A bridge between Basic and Independent User, and a demanding goal for higher-attaining pupils at the end of Y9.

Independent User. B1 (threshold)  B2 (vantage).  Independence in language needs to be built up - if pupils do not know any of the new language, they are not in a position to operate independently. In the UK system, independence is a reasonable goal for the GCSE years for higher-attaining pupils, with A2+ a reasonable goal for all who continue with a language to 16.

In effect, this gives three broad levels (A1 to A2+) for Years 7-9, with the possibility of movement towards independence for the highest-attaining pupils. The move from A1 to A2 corresponds very roughly to the earlier NC goal of two levels of progress per KS per subject.  The global requirements of each level are set out in Table 1 at the above link, p. 24. They are consistent with the requirement of the new NC to show progress across all areas of the programme of study.

The third level, Proficient User, is beyond the scope of this posting.

2 How to proceed.

1. The Framework has a list of topics and requirements that correspond to most schemes of work for Y7-9, and offers the option of subdividing its levels to allow for more precise or slower patterns of progress. 

2. For written work, the main point in my Osiris course presentation, also set out here  is that pupils should never copy when they write, but should hold words and sentences in their heads and write from memory. The steps towards this are set out in the presentation.  If this approach is used consistently, then the writing pupils do in their books is a record of what they can do, and not what they can copy. The approach can be supplemented by a test of what they can write in five minutes, in a room containing no language displays or prompts. The scripts can be retained as a record.

2. Reading can be assessed using a bank of scripts based on the topics in the scheme of work. These can be adjusted to cover progress within levels by adding new sentences or vocabulary to the basic script. Pupils must not have any prompts when carrying out the assessments, which should be as short as possible.

3. Assessing spoken language is more difficult, as it is potentially time-consuming. The use of disembodied listening tasks at full speed, common in most textbooks and recorded material, is disastrous for beginners and leads to distress and drop out. Recording yourself, or a native speaker, at a speed children can understand, and without any distractions, allows listening to be assessed using a bank of simple questions. This needs very careful presentation and practice.

4. Pupils can keep a record of what they are able to say on a personal file in a computer. Once again, a time limit, say of 3 to 5 minutes, with no prompts, provides a reliable record. As the Toolkits approach teaches the spoken and written language together, and introduces positive and negative sentences at a very early stage, it leads to close correspondence between what they can say and what they can write. Each new file contains a record of progress.

5. Progress within levels can be assessed in terms of criteria to suit the department. These might include range of vocabulary recognised and learned (Vocab Express is recommended by two former presidents of the Association for Language Learning), the use of positive and negative sentences (which avoids parroting) and accurate pronunciation.

 

Footnote The key feature of all of the assessment above is that pupils should not have any prompt while they undertake it. All of the tasks are very short - 5 mins at most - and should be structured so as not to cause pupils stress. In particular, spoken language should not, at KS3, go at a faster rate than pupils can understand, or present them with vocabulary that they do not know. The idea that they will not understand everything they meet in the new language, and need to cope with this, is valid, but this is not a reason to overwhelm pupils in the early stages of learning, when they have few, if any, resources in the new language to deal with it.