CAN WE ACTUALLY DESIGN THE PERFECT TEST?
By Evelin Suij-Ojeda
Assessing students’ performance is a complex process that involves measuring their language abilities so as to understand the progress they have made. One of the purposes of assessing learners relies on the need to provide useful feedback and enhance the language learning experience, taking into consideration students’ needs. It sounds easy and quite technical: nothing further from the truth!
Bachman (2010) defines assessment as the process of collecting information about something that we’re interested in, according to procedures that are systematic and substantively grounded. He also explains that in a language assessment, what we are concerned about is making an interpretation about some aspect of the test taker’s language ability. If that is so, then we need to design the “perfect” test so as to be able to assess in the most appropriate way. But, is it possible to achieve such goal? That has been the concern of many English language teachers, including myself.
But, what is a perfect test? What are its features? What should teachers/test designers should pay attention to? Brown (2010, p.40) describes it as being practical, valid, reliable, authentic while providing a positive washback. If that is so, our work as test designers is far more difficult and it takes into consideration not only the planning and designing of the assessment instrument, but also until what extent such tests or tasks impact language teaching and learning in our context. It sounds quite fantastic and interesting to me and I am completely sure that practice makes perfect and, since teaching is a craft (Kiely 2011) it is all about trying our best to fulfil this delicate and such important matter when learning any subject.
On one hand, an important issue that we need to pay close attention to is construct validity which refers to any underlying ability supported by theories. It is said a test has construct validity when it measures the ability it is supposed to measure in accordance with a theory of language behaviour and learning (Heaton, 1983). According to my teaching experience, the construct that should be present in all tests is: communication in real situation while making sure to avoid designing objective items easy to mark which definitely underpin validity.
On the other hand. reliability is quite important as well and, in my understanding, extremely care needs to be put on such matter in order to achieve it, especially when having writing skills tests where rater reliability is difficult to accomplish (Brown 2010). Reliability violation can take place when there are not clear scoring guidelines, especially when designing and applying subjective tests where there is not right or wrong answer.
Finally, what not to forget? Feedback! We teachers need to highlight the importance providing our students with effective and efficient feedback! It is of paramount significance that teachers understand that assessment is more than just assigning grades. but should be addressed as learning opportunities to enhance language performance. I agree with Brown (2010) when pointing out that feedback is a crucial factor, especially if we really want our students to succeed in learning a language. Not giving proper feedback can compromise most of the principles of assessment and that is why it is advisable to introduce more communicative assessment tasks which will allow teachers to provide enriching feedback, while increasing construct validity and authenticity.
Can we actually design the perfect test? I would say yes, we can!