Exams - a measure of what?
Exam season is upon us. Anxious students are cramming an impossible amount of content to revise, and anxious parents either worry about their children's anxiety or that they aren't anxious enough to do the revision they should be doing. Teachers are anxious about whether they managed to cram in enough revision lessons and school heads are anxious that their results will be better than last year. It's the climax of the school year and exams are the ends for which almost all educational means are justified. "Why do I need to learn this?" ask the students. "Because it will be in the exam.", replies the teacher. And whilst almost everyone agrees that exams are unpleasant, we rarely question the fact that they are the best way achieve their aims. Much has been written about the problematic effects of high stakes examinations, particularly by John White, on their deleterious effect on student wellbeing and their tendency to destroy curiosity by making students only value knowledge that leads to exam success. It is generally accepted that the benefits must outweigh the costs. After all, educational reform after educational reform keeps GCSEs and A levels exactly the same in any meaningful sense. The latest reforms, with the harder-to-achieve top grades, extra content and extra emphasis on memorised facts actually make all problems associated with high-stakes examinations worse. So what is so great about exams?
High-stakes exams are used primarily to create a quantifiable measure of student attainment in order for colleges, universities and employers to compare and select candidates. They are also used to judge the success of individual teachers and schools. Less obviously but perhaps just as impactful is that they are used socially and personally to judge a person's intelligence and worth. Judgements based on exam results have far-reaching effects, so it is important that exams are fair. In principle they are. Everybody learns the same information and is given the same opportunity to demonstrate their ability to use the skills and knowledge learned. In reality however, there is a high correlation between socioeconomic disadvantage and poor results, which means that the ability to learn skills and knowledge is affected by other factors. Many of these factors will be temporary, like instability at home, lack of role models, or health problems but exams rarely take these factors into account. If they are the main measure by which students, teachers and schools are measured, and they frequently are, then they are not fair and only serve to increase inequality.
Many subjects have coursework and this might mitigate some of these effects. Longer term projects mean that short term problems are less likely to have a significant impact on the final mark, and are less anxiety provoking, but they also carry the burden of more marking for teachers and, due to the difficulty of marking, the results are trusted less. For me, the real advantage of coursework is that it demonstrates a mode of thinking that is different to that of exams. By having two types of measurement, you are measuring more about that student than you are with one or the other alone. And yet coursework and exams test many of the same skills, like ability to express oneself clearly in writing, or to state specific facts. Competence in a subject can be demonstrated in many ways though. The data that we can collect about a student can be far more nuanced than exam results and coursework. We live in the age of big data, and this should open up untold opportunities for developing equitable measures of student, teacher and school attainment.
One of the problems with measuring in such a narrow way is that only certain skills are valued which means that in the classroom they become the only skills we value. This is the wrong way round. We should measure the things we value to begin with and there are many! If I was hiring, I would like evidence of good collaborative skills, presentation skills, measures of determination, creativity and many other qualities besides. Assessment of some description is taking place all the time in a school environment; teachers can assess students by the comments they give in class, by marking their work, or by observing their interactions with others. Assessments can be made along different modes of knowing; they can measure whether a student is practically skilful, knowledgable, or good at analysis. All of this data is richer than a blunt exam grade. And one of the most pernicious things about results is that they feel like a full stop, and an end to learning. When you have finished the course, there is no more to learn. People learn a lot more outside of school over their lifetime than they do in school. Any assessments we make on people should incorporate the idea of growth.
A Levels and GCSEs are neither a fair measure nor a measure of all we value. It is unclear why we even need GCSEs now that education is compulsory until 18. The reason they survive is that they provide an easily quantifiable, easily adjudicated, and easily standardised way to assess. In short, they are the easiest way to do it. What we need though are richer results that are quantifiable but contain qualitative information, that measure a range of skills, in a way that encourages development of learning rather than shutting it down. How we do that is beyond me, but if Cambridge Analytica can know someone intimately based on a few facebook likes, I would bet that universities and employers can know their students better than AAB or BCD, if only we had tools that could capture all the data that students produce in their school careers.