I've just read back through this before posting and realised it's a bit of a rant - didn't intend it to be, but it's important - I hope you agree. Please let me know your thoughts, make comments, share any ideas and suggestions you have.
Please note for the purposes of this post when I'm talking about assessments I really mean some kind of test or exam. I'm not talking about other types of assessment for learning that are used more dynamically as part of a lesson (e.g. targeted questioning, mini-whiteboard feedback, etc). The vast majority of schools use some kind of summative tests under controlled or exam type conditions at various points during the year, and these are what this post is referring to...
What are these assessments really for?
Unless we're talking about a FINAL assessment that results directly in the exam certificate, diploma or whatever's appropriate to the end of the course then ALL assessment should be used FOR LEARNING.
To clarify this lets just consider what other possible uses there could be for doing an assessment...
- Assessment for reporting (to parents, leadership, whoever)
- Assessment for categorisation/sorting/ranking of students
- Assessment for teacher performance management
- Assessment for occupying the student's time
Alright the last one is a bit tongue in cheek, but lets face it, if there is no formative or learning outcome then none of these look like a good reason to take time away from lessons to complete an assessment (where learning progress may be measured, but is not made).
What's really important is that the assessment allows you to learn something about the student that you didn't previously know - otherwise why spend the time doing it?
How can well intended assessments turn bad?
In my view the following list are the 6 key ways that assessments may end up being bad. Unfortunately I can lay my hands on too many that fall foul of one or more of these - we're working to try and improve them, but it's a big job...
1) Questions set at a level that is inappropriate to the students
Either too hard or too easy. What does an assessment really tell you if an entire class scores upwards of 90%, or another class scores less than 10%. Does this kind of result help you or the students to identify how to improve in ways that couldn't have been identified without the test?
2) Questions do not have an appropriate gradient of difficulty
Too big a change in difficulty will hide misconceptions in the gap between the levels. For more complex questions there should be a way for students to display partial knowledge that can be built on later through effective feedback.
3) Too many questions testing the same skill
Does the second, third, fourth question on a given topic give you any more information than the first about what the student knows? e.g. Do you need more than one question that uses Pythagoras's theorem?
4) Test is too long/short
Too long can overly penalise slow workers in terms of time required to complete it. Students can succumb to boredom and give up, meaning that the results and therefore feedback will be inaccurate.
Too short can have all of the features of (1) and (2) - vitally does it give you or the student any more information that you didn't know before the test?
5) Questions are too similar to a method or example given in class.
Does the question test true knowledge or simply the ability to reproduce something they've seen? (though sometimes assessing the ability to simply follow a method is important - this isn't always bad, but needs to be done deliberately rather than accidentally)
6) Viewing assessments as fixed part of a scheme of work/programme of study
Don't get me wrong - as a department leader I am keen that assessments are done as part of a planned, coordinated activity so that they can be used for departmental management as well as for learning. However we shouldn't get caught thinking that we have to do the same assessment at the same time every year because that's what we did before. All SoW or PoS should be live documents, being developed year on year to maximise learning. If improved learning means scrapping an old assessment or inventing/finding a new & better one then tradition shouldn't get in the way.
But the biggest problem of all... (and this can apply even if an assessment is "good" in all other aspects)
Only giving a grade, level, score or percentage as feedback
Basically this doesn't contribute to learning, and gives the student nothing constructive. It rewards the high achievers, it scolds the low achievers and doesn't tell them how to improve. As a result the assessment a complete waste of your and the student's time, which could have been better much spent actively learning something.
As such there must be a formative aspect to all assessments and a feedback process planned that shapes both student activity and teacher planning for the future.
Students need the opportunity to reflect following assessments so that they can learn from them. This reflection needs to be diagnostic and give guidance on topics/aspects that need to be improved, so that if they were to attempt a similar test again in a few weeks they would be able to demonstrate progress (I'm not suggesting that this similar test needs to actually happen).
This opportunity for reflection should be given adequate time during lessons, and be structured and guided, especially for weaker students. I've talked about methods we've used for doing this in an earlier post here. We're constantly developing this, and I see it as a key tool for future successes.
Another similar approach from @justmaths is here and this also includes some really good next steps sheets that can be used to rectify gaps in understanding (something that I'm planning on adopting in the near future).
So are there Good or Bad assessments?
I'll stick my neck out a little here and suggest that there isn't really any such thing as a good or a bad assessment. Even an apparently dry and dull page full of repetitive tasks could be educationally useful for a certain student or groups of students if targeted appropriately and reviewed/responded to in an effective way.
What really makes an assessment good or bad is how it's implemented and how the information it gives is used to improve learning. Both of these are down to skilled teaching and having the right routines and feedback structures in place to support it.
Basically if the assessment is "bad" then don't blame the test paper - it's the teacher that has set it that needs a rethink. In some cases a re-write of the test paper may well be needed for that student or group, however in others it is simply about how the test paper is used. Similarly an apparently "good" assessment could be bad and therefore worthless if given to the wrong group of students, or if it's not used constructively.
The key point I'm trying to make:
Lets make better assessments, but more importantly lets make better USE OF assessments.