This blog post was first published as a guest post on The Association of School and College Leaders’ (ASCL) website. It is the first in a series presented by Evidence Based Education which will explore the four pillars of assessment – purpose, validity, reliability and value.
The importance of purpose
We live in an information age, at a time when the quantity of information in our lives often outweighs the quality of it.
Assessments used to generate information on which decisions are made about student learning need to provide high-quality information fit for the purposes intended. There are dozens of reasons why you might assess pupils, and the ideal type of assessment is different depending on the purpose. However, it is not uncommon to lose sight of the function of assessment.
Assessment can also become all things to all people, where information collected for one purpose is also used as a measure of something entirely different. For example:
An end-of-year assessment designed to measure attainment in mathematics, but which tests only certain components of mathematics (multiplication, division and place values, for instance), cannot be used to draw conclusions about maths attainment in general (the intended purpose), only the components it includes. Matters are complicated when questions in the assessment contain overly complex wording. The assessment now requires sufficient reading skill to access the maths at the heart of the assessment, thereby disadvantaging weaker readers.
In addition, an assessment that is seemingly plausible on the surface may not be quite right. For example, if we are trying to capture how much has been learnt from point A to point B, even if we have a good baseline and follow-up measure, if sufficient time isn’t allowed for learning to become embedded, are we really measuring progress?
To capture progress of pupils’ learning, let’s say about the Roman Empire and its impact on Britain, we need fit-for-purpose baseline and follow-up assessments – how else could we detect improvement? But when should that follow-up assessment be given? And what should it test? We’re interested in understanding more about what pupils have retained for the long-term and can apply in different contexts, but does a single follow-up assessment at the end of a unit of teaching enable you to make a secure claim about progress?
Sometimes, what we think is an assessment of learning is more likely an assessment of short-term memory. Also, no assessment is 100% accurate (there is always ‘noise’ (error) around the ‘signal’ (the true meaning) of an assessment), so sometimes what we detect as progress is actually little more than inaccuracy.
Three steps for robust assessment
Everything about a great assessment begins with a clearly-stated purpose. As such, purpose is the most important of the four pillars of great assessment. We must know what we want to measure and why, in order to select the right tool to achieve our purpose. No matter what we want to measure – from self-efficacy to algebraic fluency to word decoding ability – we should make sure the tool we use to measure is actually capable of measuring the thing we want to measure: it should be valid for the defined purpose.
These three steps are always at the heart of any robust and purposeful assessment:
- The construct: What is the specific knowledge, skill or understanding (drawn from the curriculum) that we intend to assess?
- The end use: What do we want to do – the interpretation, the decision or action – with the information generated by the assessment process?
- The best tool: What and when is the most appropriate, effective and efficient way to assess in this instance?
Without clear answers to these questions, it is impossible to use assessment effectively. If we don’t have a specific goal for an assessment, there is no way of knowing if it is any good at providing the information we need from it. And without good information, guiding students along their learning journey is difficult.
What sorts of assessments do you use in schools? Whether they are external standardised assessments, home-grown tests, or past papers, is everyone clear about what their intended purpose is and how the information from them will be used? Are they fit for their intended purpose, or have they been warped over time – bent out of shape to fit a need in school? Or perhaps they’re done because … well… “we’ve always done them”!
Make sure the most is made of your time on assessment with appropriate, dependable measures to make appropriate, dependable claims and judgements.
“Understanding Purpose” is one unit of learning from the Assessment Lead Programme, offered by Assessment Academy. The programme is designed to offer a grounding to school teachers (primary and secondary) in assessment theory, design and analysis, along with practical tools, resources and support to help improve the quality and efficiency of assessment in your school.
Validity of assessment is an important issue we are conditioned to accept validity based on pure content as opposed to developing the ability of being able to manipulate information derived from teaching and applying where appropriate.
Hi Adrian, thanks for your comment. Our view – one I think we share with you – is that assessment is an effective tool to use for the improvement of student outcomes; there are, however, multiple purposes to which it can be put (school accountability measurements, progress tracking…), thus making the definition of the purpose to be a critical act.
So, no, the purpose of assessment is not assessment itself; rather, it is something that must be defined by those who want to create it, use it, and use its information.
You have written the entirety of this post on the purpose of assessment with measurement in mind. So is that the purpose of assessment?
I thought it was learning.
Sure by the end years of schooling measurement might become a high level need (even that is coming under question recently) but up until that point surely the point is how the assessment helps in the development of a child’s learning journey? Efficacy and engagement is more important than reliability and noise-signal ratios.