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.