Fit-for-purpose assessment is valid, reliable, transparent and authentic - and manageable!
Why do we need these characteristics for assessment and what do they actually mean in practice?
Validity is about making sure that we're using assessment to measure exactly what we set out to measure - students' evidence of achievement of the intended learning outcomes. We need therefore to make sure that we know exactly which intended learning outcomes each element of assessment is addressing. But sometimes validity can be compromised by the form of assessment we choose - for example traditional exams sometimes end up measuring how well students can write about what they know, rather than how well they've got their heads round the subject.
Also associated with validity is assessing at the correct level. We need to make sure that the right level descriptor is appropriately reflected in the learning outcomes and associated marking and assessment criteria. The two Teesside Rough Guides entitled Learning Outcomes and Assessment provide help with how to write clear and effective learning outcomes and using assessment and marking criteria.
Reliability is about making sure that we're being fair and consistent, and that each mark or grade is accurate and realistic. In practice, this means that we've got to make a well-honed marking scheme for each element of assessed work (whether it is an exam question, an essay, a report or many other possibilities) so that we can be sure that we're being equally fair to all of our students. When there's a really good marking scheme, different assessors will agree on the marks to be awarded for particular exam answers or assignments. Also, there won't be any variation in the standard of assessment on the journey from the first piece of work you mark down to the last piece in the pile.
Transparency means we have to make sure that our students know how assessment works. They need to know what we're looking for in an excellent answer or piece of work. They need to know what they must do to reach a pass mark. They need to know what would not get them a pass. In other words, we need to help our students to see that what is being assessed is their evidence of achievement of the intended learning outcomes, and that these outcomes are useful to them as goalposts for their studying.
Authenticity has two sides. We need to be able to be sure that what we are marking is indeed the work of the students concerned - in other words that they haven't copied it or downloaded chunks from the web. At least in traditional exam situations, we're fairly sure about whose work it is. But plagiarism is largely a problem of our own making - we need to design out plagiarism in coursework assessment, by making what we assess more clearly students' individual efforts (for example critical incident accounts, reflective logs, making sure essay questions can't be answered by one book or article and so on). The other side of authenticity is about how 'real life' our assessment is in practice. For example, we can't expect to measure drama performance skills effectively by asking students to sit in an exam room and write about drama performance skills! This means exams and essays may not be the best modes of assessment to develop skills and knowledge.
Manageability also has two sides - assessment needs to be manageable for us and for our students. In the UK, it can be argued that there's too much assessment and, because of all of the pressure this causes, that it doesn't work very well. We need to be streamlining assessment so that it is of high quality and we're assessing (making judgements on important things) and not just marking (merely ticking off routine things, for example spelling, punctuation and grammar). When students themselves are overloaded with assessment, they are often driven to surface-learning mode, learning things rapidly just for the exam or assignment, then forgetting them just as quickly.
