The debate and vote in the Legislative Assembly on Monday brought to the floor of the Assembly a discussion on a very significant issue for education in Alberta, as we constantly strive to understand how students learn and provide the best possible opportunity for all students to be successful. (March 16 — text of my comments published here; the full debate is in Hansard; or you can watch the video.)
Motion 503 focused on the PAT 3 tests. It requested the government to eliminate achievement tests at the grade three level and focus on diagnostic testing instead.
Several very good arguments are put forward in the debate — including that it is too early to test children; that the tests cause stress, anxiety and fear of failure; that it interferes with teaching as teachers feel they need to prepare children; that it drives teaching to the test; that diagnostic assessment provides a better learning tool; that assessment for learning focusing on the needs of the student is more effective; that the results are used inappropriately to rank schools and teachers, etc. I won't elaborate on those arguments here — the various speeches do that very well. But assuming even some of those arguments are true — why not just get rid of the tests and move on?
In the spirit of the motion, which was successful, we will have to examine our grade three PATs and make exactly that determination. However, as I said in my remarks, it is not prudent to throw out something which has been in effect for many years and which forms one part of a much larger Accountability Pillar, unless and until there has been a thorough examination of the subject and you know that what you are moving to will be more effective.
This is a much bigger issue than simply grade three PATs. Most of the arguments used can be translated to PAT6 and PAT9, even departmental exams. Indeed, the ATA has been actively promoting "Real Learning First". They have brought in education gurus such as Alfie Kohn and David Berliner as experts to influence the discussion. Why not scrap the whole bunch of them (the tests I mean)?
After all, teachers are professionals. They are in the best position, working with the student all year, utilizing a variety of methodologies to ensure that learning styles are understood, dealing with the students individual learning needs and using a variety of assessment tools and methodologies to ensure that concepts have been learned and understood. Let's just accept that — take the teacher's report on student achievement and use that to provide not only the learning assessment for the child and report to parent but also as assessment of learning to report back to the system and the public on the success of the system. Indeed, Alfie Kohn goes further. His website cites him as his country's "most outspoken critic of education's fixation on grades [and] test scores." He advertises a DVD "no grades and no homework = better learning". In a recent discussion I had with him after we both had presented at a teachers' convention, he postulated that assessment is destructive to learning (my paraphrase).
Berliner is not as dogmatic, at least in my brief experience with him. I heard him speak at the ATA Summer Conference and had the opportunity to meet with him after. His thesis is that "High Stakes Testing Corrupts America's Schools" (a portion of the title of a book he coauthored). When I heard him speak what I received was that when standardized tests are used for the wrong purposes they can seriously affect behaviour of educators and "corrupt" both the result and the system.
Are Kohn and Berliner and others right? Does our standardized testing actually harm Alberta students, destroy or seriously impair learning opportunities? Are we destroying creativity and inspiration, the very attributes we purport to aspire to develop? Are we engaged in "high stakes testing" and corrupting the education system? These are indeed very serious issues the understanding of which must go far beyond a one-hour debate in the legislature or indeed provocative presentations at conventions. These issues deserve serious analysis and discssion both in our education system and in our community. I thank my colleague, Genia Leskiw, PC - Bonnyville-Cold Lake, for raising this debate in the Legislature and making it more than a discussion among teachers at convention or an ATA issue, but bringing out for the community to engage, understand and help resolve.
Arguably, Alberta is recognized as having one of the best education systems in the world. We are mentioned comparatively with Finland, South Korea, and Singapore on the top of the world. People came from more than 25 jurisdictions over the recent past to look at our system and find out what makes it successful. The editor of the Des Moines (Iowa) Register recently won an award for her articles (written after visiting) on Finland and Alberta. Attending a Ministerial World Forum in January was very affirming because in discussion with or hearing presentations from many of the 65 Ministers of Education from around the world I discovered that most of them aspired to what we already have. At least by those measures we have much to be proud of, our system is doing well.
But are we doing the best we can for our students? Could we do better? Even if we do well today (or yesterday really) are we prepared for tomorrow? I don't know. What I do know is that by all accounts what makes our system what it is today can be attributed to many factors, but those most often looked at involve the quality of our teachers and teaching, our robust and provincially mandated curriculum supplemented by locally developed resources, and our accountability framework and assessment practices.
Standardized tests, forming a portion of the 16 different measures in the Accountability Pillar, are developed by teachers and are validated to ensure that they are age appropriate and validly test the subject matter. We have had them in place for over 20 years which means we have longtitudinal data on students for over 13 years. With that data we can demonstrate that the tests are entirely predictive. What it also tells us that we have not used the results well to change that predictability. We often get lobbied to change the percentage that diplomas count on a students final grade twelve mark (currently 50%). But the data also demonstrates variability based on identifiable criteria which suggest that a provincial standard is necessary (I admit to a bias — my English 30 mark was significantly improved by the diploma result).
Test results are not intended to be used to measure or rank schools and teachers, yet we are forced to release the results under FOIP and they are then published and used inappropriately to rank schools. Many districts and school principals use the results well to discuss with teaching professionals identifiable characteristics and trends which can then be improved upon, but anecdotally I have heard that some principals have used them to justify teaching assignments. I think that we all, as educational stakeholders, would agree that we do not want these tests to be "high stakes" for teachers and schools, but rather to be used at a higher level to inform practice.
Finland, as Pasi Sahlberg points out in his Real Learning First presentations, does not use a standardized curriculum or standardized testing, so arguably testing is not needed to be number one — but what have they done? Well, as a start — they have put a strong emphasis on valuing teachers and making the teaching profession a highly respected calling. We could learn from them on that account. Maybe we can share with them curriculum and testing. Or not.
What do you think? How can we fit all this into the broader question asked during Inspiring Eduction: A Dialogue with Albertans, "What does the educated Albertan look like 20 years from now?" How do we know we are there? While in London I met with Dr. Barry McGaw from the University of Melbourne who is engaged in a three year program to develop appropriate assesment tools to assess 20th century skills. Maybe the prudent thing to do is engage with him and his study to help develop new methods of assessment and then replace our PAT 3s.
After thinking hard about this for years, as both a parent and a journalist, I think well-written and designed standardized tests are an essential educational tool -- not so we can grade eight-year-olds, but so that we can measure how well our schools and teachers are doing.
A couple of years ago, I wrote a story about Inglewood school, a not-quite-inner-city school where HLAT and PAT scores had soared. Why the sudden surge?In large part because the school began an excellent hot lunch program. When kids were well-nourished, they were able to learn. How could we have measured that without the tests?
Another time, I wrote about how aboriginal children's test scores went up -- because they were put in classes with better student-teacher ratios, and better-trained teachers. The improvement had nothing to do with cultural sensitivity, just with better teaching. Again, how could we have measured that without tests?
I didn't write a column about 503 at the time because it seemed a meaningless motion. But perhaps I do need to weigh in, in a broader forum.
Now, if only we could make MLAs take tests every three years, to see what they've learned.....
Standardized testing is a measurement tool, as Paula says, for the schools and teachers. While the results can be (and possibly are being) abused, this doesn't invalidate the tool itself. If it isn't taken as an absolute way to rank schools against each other, it simply gives data for us to use to evaluate changes in our educational system.
I think it also benefits the kids themselves. I value that I was taught to strive for excellence at school, and "no competition/cooperative" academic environments make me wonder if that same challenge is being posed to students today. At the very least, we will be able to evaluate the educational techniques used today against the results of yesterday and determine if we like the direction we're headed.
Unfortunately, Paula and Christopher misconstrued the point that the test is still the indicator of success. If we push to do well on "the test" then we can skew those results (see Dr. Berliner's book, mentioned above), thus failing to recognize the other areas of learning not covered under the PAT.
What it important is that Paula demonstrates how social factors can affect scores. The risk is to blame a lack of "better teachers" when even one simple social factor (ie lunch, lower teacher/pupil ratio) can significantly impact learning. Yes, summative assessments can serve a purpose. Just remember, lies, damn lies and PAT tests, to paraphrase an old saying.
Excellent blog, Dave. You state the challenge very succinctly.
I don't think the real issue is with Achievement Testing. I think the concern is the one time nature of the exams. Students who "freeze", are sick that day, had a bad week due to a bully, any number of issues can have a significant effect on a one-time test. Rural areas with small class sizes can be especially affected by this, where one student doing poorly is worth twice the percentage of students as in a larger urban school. Cultural differences can make a tremendous difference; some groups start reading later than others, have no opportunity to read at home, and so forth.
We need to account for variables better. We need to spread out the PATs so their impact is minimized; don't place all the emphasis on the results of just one test.
Better, we need to put assessment in the hand of professionals. Teachers. While there are 14 other measures in the Accountability Pillar, most others are qualitative; we need a good quantitative measure. Utilize teacher driven and teacher produced marks to provide another more applicable quantitative measure, and relegate the PAT to being just one measure of success, not THE measure.
First, thank you Mr. Hancock for this outreach to the public and succinct articulation of the host of pressures impacting assessment (what kids know and can do with that knowing) and evaluation (the assigned grade). I read all the comments above as having value in this debate. My own take is to evaluate all the crucial social actors whose decisions impact student achievement (e.g., politicians, various professionals and their organizations, parents etc). After all, to reduce the assessment/evaluation issue to a single young person denies the obvious - they are embedded beings in complex social networks and geographical environments. my own paper- where my co-author and i try to identify 'indicators' for various people impacting student learning - and albeit in the context of the US's NCLB legislation, but equally applicable here, can be found at
http://www.uofaweb.ualberta.ca/secondaryed/pdfs/wdrupolicyfutures.pdf
for what it is worth. thanks again!
What necessary words... super, a brilliant phrase