How to Test Speaking Skills in Japan

48 Part 2: How can we mark & administer speaking tests? Notice how this rating scale strives to capture a continuum of overall English ability by weaving together various strands of com municative competence, grammatical awareness, and vocabulary complexity Keep in mind these are only three possible variables to look at—given the complexity of speech, there are certainly other ways to define it It is also important to make sure that the top of the scale is not way beyond the ability of most learners (McNamara, 2000) Underhill’s scale does a good job of adding language that softens things at Level 5 ( ...may be at a loss for words... makes errors of idiom or collocation... ), but take care: avoid descriptors such as talks like a native speaker , as they may reinforce the idea that English fluency is an unattainable goal Finding the right wording to define each level will take some trial and error on your part The more you know your students and what they are capable of, the better you’ll be at matching the descriptors to what they can do While Underhill’s scale is about evaluating overall ability, you may find that you don’t always need such a general assessment For example, perhaps you just want to see if your students have picked up the latest material you’ve taught them In that case, you’ll need to work out how to describe each level in terms that are applicable to your immediate needs It’s a good idea to write the descriptors in language your stu dents can understand You may therefore want to have an alternate, simpler version of your rating scale, or even a bi-lingual one if you are working with lower-level students Knowing what each level means will help learners strive upwards because they’ll see where they can go In ad dition, the discriptors do need not be extensive or comprehensive in order to be useful They can be written to meet more targeted goals For example, if you are especially interested in promoting quick re-