EduTECH Day One: Eric Mazur - Assessment: The silent killer of learning
Eric Mazur is a physicist and educator from Harvard University
Mazur kicked off with an anecdote about meeting a woman on a plane who developed flash cards...for phones. It was clear from the outset that this man doesn't value memorisation and rote learning - I think we are going to get on.
He was quick to point out that memorisation is only about passing the test. After a few days only 35% of that information.
Today he is arguing something I feel passionate about - our assessment practices are outmoded and outdated. Really only designed for ranking students.
What is the purpose of assessment?
He gave us 90 seconds to list purposes of assessment...I got:
Needs for future learning
Others might value:
Personally really only value assessment FOR learning.
Will post when I have better wifi!
He identifies a key issue. It is often inauthentic problem solving. Couldn't agree more! This is my problem with external examinations! Why in this day and age would we use artificial testing of understanding when we can and so already measure understanding in authentic contexts when the students are actually with the practice.
At this point Mazur puts up a traditional written Maths problem students might be presented with in an exam and steps us through how this takes students right to the bottom of the blooms taxonomy - remembering. Why would we focus on this when computers do this so much better than us. Why aren't we focusing on authentic problems? But of course this involves taking risks and students are risk adverse. Assessment also kills creativity.
He then shares a photo of a typical exam room and asks what we see, then states the word - isolation! When do we do anything in isolation in the real world? Why are we testing students in a way they will ever experience in the professional lives??
Traditionally exams have been a high stakes assessment. Something for which students cram, which results in little more that short term memorisation.
Assessment produces a conflict - are we the coach or judge?
Teachers hide behind a thin veil of objectivity.
And then there is grade inflation and cheating. Cheating is not a problem with our students, it is a problem with our assessment.
How can we improve assessments?
Let's mimic real life - let's have open book exams.
Moving assessments online and give access to Google. Any assessment that have answers that can be googled is not an effective assessment. We need to go beyond the ungoogleable question and move to the ungoogleable assessment or exam.
Website worth visiting - Team Based Learning
At this point he shows a video of students taking a collaborative exam. 50% of the assessment is individual, 50% done as a team.
The videos shown as this point what I feel like we are seeing is not just "assessment for learning" but "assessment as learning"
Another suggestion he makes is importance of assessment feedback, this aligns with what we probably think of as assessment for learning.
Next point - let's focus on skills not content (take note AGS).
Mazur shares the work of Grant Wiggins and - Learning by Design. Think about the outcomes.
How might we resolve the coach/judge conflict. Make the assor a different person than the teacher/coach. We also need more peer and self-assessment.
Mazur suggests using calibrated peer review.
Check out: http://cpr.molsci.ucla.edu/
We must rethink assessment!
If we don't we continue to prepare students for the past, not the future.
Couldn't agree more: I published this article last year.