Accountability is an extremely powerful force in program design and delivery. It is a powerful “steering system” in that small moves on the steering wheel produce large changes in direction. Programs tend to organize themselves to maximize returns on the accountability measures, so it is important to be extremely careful when putting accountability into placei. All too often there have been examples of programs ending up “teaching to the test” when accountability structures only look at test results. As much as possible, accountability should adopt a light touch, focusing on the key measures reflecting program quality and nothing else. In many cases, the quality of accountability data can be improved by removing requirements rather than by adding them. Small amounts of more focused data are more useful than large amounts of irrelevant data.
Assessment is another form of evaluation, though this time instead of looking at program level indicators the questions are focused on individual learners. Assessment tries to capture what people actually learn in a given situation. For many people, the first example that springs to mind is formal exams. However, there are quite a number of other ways to record what people know. For example, people could be asked to write essays, or perform a simulated (or real-life) task. Probably the most important factor in high quality forms of assessment is not the means that is chosen, but that it is collected and collated in a systematic way. Assessment must be carefully tailored to the expected learning and the context in which it will be used.
There are a number of different ways to use assessment. At the time of intake, an initial assessment could be used as a diagnostic tool, to help the learner and educator get some sense of where to start. Formative assessments can help to check that learning is on track, or to find out if some instructional approaches are more effective than others for that learner. Summative assessment tries to capture the entire span of learning. Assessments can also be paired, as in pre-assessment and post-assessment, to provide evidence of learning over time. In this discussion the main concern is the form of assessment used to record learner progress in each of the three case study jurisdictions rather than diagnostic and other formative measures.
There has been a great deal of interest recently in standardized assessment, where every learner takes the same test, but it is important to be careful about this type of assessment. It might be useful in a classroom of schoolchildren where all have started at the same point and received the same instruction for a given period, but it tells us a great deal less about the knowledge an adult possesses. We know that adult education is strongly influenced by people’s previous schooling, their life experiences, and their current circumstances, so there will inevitably be differences in what they learn over a given period of time and standardized tests do not always capture this well. For example, two new Canadians may well score at the same level on a standardized test of French language ability, but one is a trained physician and the other has very little formal education. In this case, the score has quite different implications for the instructional program provided to each individual.