14.6
Conclusion

The key lesson to be learnt about all assessment, not just language assessment, is that testing should be closely linked to the learning outcomes of our courses. This means being very precise about what it is we want to measure and selecting a suitable assessment form (eg exam versus portfolio or project) and an appropriate test type (grammar, reading, mixed-skill, etc). None of the tasks described in this module is without its drawbacks; there is, in fact, no such thing as the perfect language test, and the best we can do is to ensure that our chosen instrument of assessment meets as closely as possible the five key criteria of validity, reliability, transparency, practicality and washback.

Just as importantly, we know from, among other things, research into learning styles (see Module 12) that not all learners respond well to the same mode of learning or testing. We are therefore most likely to be fair to students and to allow them every opportunity to show their ability, if we ensure our assessment procedures feature a range of methods and tasks. This means combining communicative tests (eg free writing) with those that focus on sub-skills (eg grammar exercises) and balancing subjectively marked tasks (eg oral interviews) with more objective ones (eg completion of a negotiation task).

Activity 30
The following is a self-assessment exercise on this module. Referring back, as necessary, to the relevant section in the right-hand column, write a short paragraph about each of the following statements:

Section
The only tests university language tutors need to be concerned with are achievement tests. 14.1.1
Sound testing systems feature indirect testing of underlying elements of language, such as pronunciation, grammar or vocabulary. 14.1.2
Language testing should always deal with one skill at a time. 14.1.2
Integrative skills testing is not sufficiently reliable for use on university language courses. 14.1.3
Of the three methods of marking student work, the analytic and objective approaches have most to recommend them.

14.2.1, 14.2.2,
14.2.3

There is really only one valid approach to the objective marking of FL essays. 14.2.3
Advanced-level university students do not need much help with writing FL essays, other than vocabulary. By this stage, they should know how to tackle discursive titles. 14.2.4
If used for assessment purposes, student oral presentations should always be an individual task. 14.3.2
There are basically three types of oral exam question. 14.3.3
The most effective listening and reading comprehension tasks are open-ended questions. Multiple-choice or true / false questions have no place in university language testing. 14.4.1, 14.4.2
We should be aiming to banish English completely from our testing of foreign languages. 14.4.4
Cloze tests are both a reliable and valid form of language testing. 14.5.1
C-tests are both a simple and economic form of assessing overall language competence. 14.5.1
Assessment of translation work needs to focus on linguistic accuracy. 14.5.2
The danger of plagiarism makes the use of projects and portfolios in language work too problematic. 14.5.4, 14.5.5

 


previous button
next button

contents button