Assessing Speaking – An Alternative Method

I believe one of the most challenging skills to assess and grade is a learner’s speaking skill. I have worked at institutes that have had devoted meetings to this topic and it can be been controversial. Indeed, I have been in meetings that have lasted several hours as rubrics have been proposed, amended, and protested against. I recall one particular meeting, where we the teachers, became engrossed on the wording of one particular item of a speaking rubric. However, after the meeting ended, we were no closer on reaching a consensus on how the rubric should be designed.

One of the main issues assessing speaking is that it is so subjective. What one teacher may consider a fluent and accurate speaker, may contrast with another teacher. After all, what is fluency and what is accuracy? In linguistic terms fluency is considered the time between utterances and accuracy is related to grammatical and lexical accuracy. However, these can be both thrown out of the window in real life situations. Depending on the context, as a native speaker, I might feel it perfectly natural to take my time and pause between utterances. In cases of expressing my feelings, I do this quite naturally. I come from the North of England, and in terms of accuracy, if I am talking with friends from my hometown, I will certainly not adhere to grammar structures that are taught in standard textbooks.

There are many other factors to take into consideration. One example might be the personality of the speaker. Some people have a more natural ability to communicate with others regardless of nationality or mother tongue. Others might be more introverted and not particularly enjoy talking with someone unfamiliar. How about nerves? How can we factor the white coat syndrome into a rubric on speaking? What about context? Is it fair that a certain learner might be more comfortable with the assessor, or their chosen partner in a speaking examination? How can all this be taken into consideration when assessing an individual? As you can probably guess, I am edging towards the idea that standard testing is not always feasible or applicable for speaking.

So, how can you possibly assess speaking? My own perspective is to take a longitudinal view on speaking assessment. I teach at the tertiary level at a university in Japan. I am assigned to teach 15 weeks of communication classes in a semester. In my early days, I used assign a final examination in which students would be required to talk for a certain amount of time and I would assess them using the rubric I had designed. Of course, the students were made aware of the criteria, but it just didn’t work. I had cases where students who had consistently performed well throughout the semester froze during the test. I also had cases of students who had rarely contributed to in class performed exceptionally. Following this system clearly did not reflect the effort, performance, or speaking ability of all of the individual members of the class.

What is the alternative then? At the moment, I am experimenting with a system of constant evaluation. In my communication classes, at the start of the semester, I explain to the class that there is no formal examination. However, there are certain criteria that I would like them to meet throughout the course. These include active participation in class discussions, a willingness to communicate with me and others, and for them to be both willing and able to formulate and ask questions. I am still in the early stages of developing criteria for these requirements which means my grading is still subjective. However, as I experiment with this form of assessment, the goals and criteria have become more concrete. Even at this early stage, I believe it is a fairer form of assessment than the one I described above.

I think if I am to adopt a constant evaluation for speaking skills, I need to make the learners fully aware of the expectations I have for the course. I also need to find a consistent and fair way of evaluating them whilst being the teacher. It is by no means an ideal replacement for a speaking test, but one in which is surely warrants  further research and discussion.

Neil Millington is the founder of If you are looking for free materials for your students, please check out the lessons on his site.