It is a familiar judgment of this or that policy or program that it encourages “teaching to the test”. This is invariably pronounced as if teaching to the test is a bad thing. It is not.
Call it “testing to the teaching” if it sounds more palatable but it’s the same thing and it is a good thing. One wants to be crystal clear on what students are expected to learn, to determine afterwards whether or not they have done so, to reward those who have and to work hard with those who have not. It is therefore obvious that we should be testing, and that we should align teaching and testing. The only semi-plausible argument to not teach to the test is if the test sucks. And, yes, most tests suck. But the first and most important question to ask is: why does the test suck?
The VCE mathematics tests, for example, suck, and the “why” is a no-brainer. Literally. The destruction of the curriculum, the perversion by CAS, the aimless and idiotically fusspot SACs, and the superficial and excruciatingly badly written exams are all the product of decades of arrogant brainlessness. Still, a VCE teacher must teach to the test.
The thoroughly perverted culture of VCE mathematics guarantees that the teacher is going to be teaching a hell of a lot of nonsense. She is continually going to have to shortcut proper explanation because the explanation is not examinable and there is simply no time to spare. She is continually going to have her students perform black box rituals rather than thinking mathematical thoughts. She is continually going to remind her students that they absolutely must keep in mind some anal-retentive nonsense that VCAA will fuss over in November. But that’s the game. The game sucks and you should tell your Head and your Principal and VCAA, loudly and often. But in the end, if you cannot stomach the game, and we are aware of a few smart teachers who cannot, then yell about it and get out. If you’re in the game then you play by the rules of the game, which means teaching to the test.
A VCE teacher that is not teaching to the test should be sacked, and probably would be if they were not first lynched by their students. But even for lower-years and/or low stakes, run-of-the-mill tests, teachers should be teaching to the test. Do these tests also suck? Yes, typically. But, again, we must first ask why.
NAPLAN sucks. Hugely. It sucks because ACARA has idiotically arranged to test “numeracy” rather than mathematics and cannot even do that properly. But NAPLAN also matters. Private schools, exhibiting all the intelligence and moral maturity of exclusive golf clubs, demand NAPLAN results from applicant students. VHAP, exhibiting all the discernment skills of the Robodebt thugs, uses NAPLAN to select students for its “high ability” extension classes. God knows what else, but NAPLAN matters. Why, then, would it not be proper for teachers to teach to NAPLAN? ACARA says teachers, and parents, are “not expected” to help students prepare for NAPLAN, that NAPLAN is just one itsy bitsy test, that “excessive coaching” doesn’t help, that it’s all about the schools, and so on. ACARA is lying through its crooked yellow teeth. Of course practising NAPLAN tests will improve NAPLAN scores, and of course NAPLAN scores matter for the student.
What about regular lower year tests? Yes, most of these tests suck as well. Most schools, guided by ACARA’s thin-gruel curriculum, and immersed in an “Everyone’s A Winner” culture, discourage proper testing and will whack any teacher who attempts to do so. As we wrote elsewhere,
In junior secondary school there are tests, but the tests are gamed. A parent may believe that 70/100 on a Year 8 maths test is pretty good, but the parent is typically unaware that the test likely consists of 60 per cent that is trivial, 40 per cent that checks basic facts and skills, and 0 per cent of anything requiring deeper thought; that 70/100, therefore, likely indicates 10/40 on the basics, which should ring the alarm bells but does not.
Most school tests suck and there is a decent chance that your co-teacher who wrote the test or simply grabbed the test from two years ago, and the Head focussed on your co-teacher’s precious feelings, and the Principal focussed on the threat of whiny parents, don’t care or are too stupid to realise. Still, you gotta do it. You can and should yell that nonsense is nonsense, but you still have to teach to the test.
It is simply unsustainable to expect to teach good and solid mathematics but to then not test it. Students learn what you teach them, and if you teach them that the hard and solid stuff will not be tested then they will very quickly learn to not learn it. Students are busy, too, and they are not dumb, even if we are doing our best to make them so.
The fundamental issue is not the testing but the teaching. If one is permitted to and does teach well then one can test well, aligned with the teaching, and one should. Do not horse and cart this. Do not blame testing for bad teaching.
These days, many tests are set by firms that provide testing software to schools. In my experience, the quality of such tests leave much to be desired. For example: Simplify the equation (sic)
+ 
.
I don’t understand. Who purchases these tests and for what purpose?
Cambridge (and I assume other publishers) have “test builder” features teachers can use.
There are a few other “question bank” type programs out there as well. Some offer an annual subscription where new questions are added each year.
Some schools use them, some schools just recycle old papers and some teachers write new stuff each year.
The teachers who write new tests each year get paid the same as those who source tests elsewhere so… yeah.
There is nothing wrong with recycling a test unless the test sucks. How many teachers care if the test sucks?
Epsilon.
Exactly. The problem is not recycling, or buying, tests, the problem is if and that the tests suck.
As opposed to other aspects of Mathematics education in this country…?
It all seems rather connected to me.
Of course v
Think Terry is talking about things like ‘Education Perfect’, basically online tests (practice exercises and pre-post tests as well) on topics in the curriculum., and they allow teachers to track student progress. Apparently the companies that supply these platforms are making a killing from schools. Here’s a question I saw once.
Ugh!
I have a few queries on this topic. Are these examples of teaching to the test?
A) How do you feel about tests that use prior knowledge e.g a Yr12 methods trig question that may use geometry knowledge from years 8 or 9?
I tend to think that anything done previously is fair game but you should let students KNOW the expectation prior to the start of a topic.
B) Close on the heels of A); how do you feel about test questions that incorporate several aspects of mathematics into the one question e.g. solving a differential equation but using mechanics ( is any mechanics left in VCE??) to set up the ODE.
I am OK with questions that engage 2-3 topics in a question but I tend to draw a line at that number. Too many topics in a question will result in students falling off the cliff in a hurry
C) How do you feel about test questions that have implied knowledge of what is required to solve them? Almost all questions now have multipart answers leading to the final result e.g. rather than asking for the volume enclosed by the solid of revolution of the area formed by two intersecting curves; you may get several parts before the final answer (drawing the curves, finding intersections and only then finding the volume)
I am personally OK with implied knowledge questions. However many colleagues now consider them to be a no-no because students complain vociferously of unfairness and the fact that so many tests that students have previously attempted are of a multipart nature.
D) Should test questions all be of exactly the same format, structure and configuration comparable to what they have experienced previously?
I am loathe to repeat everything exactly the same as you want students to be resilient enough to read what is required rather than a rigid test structure
Thanks, Simon. I have thoughts, of course, but I’ll leave off replying for now, and let others have a go.
All of these questions/queries/thoughts are important. I doubt many educational leaders in schools think much about them and therein lies the heart of this problem (in my very unimportant opinion).
Starting with the most fundamental question – why do we test? There are no doubt a bunch of ideological reasons schools and teachers can give, but ultimately I suspect the answer is that we, as teachers/whatever are expected to test.
Is it to rank students by ability? Maybe at the end of school this is true, but in the other 12-and-a-bit years that argument feels a bit redundant.
Is it to find out what students can and cannot do? We don’t need a test for that, surely. And even when we do test, do we then go back and re-teach everything a student gets wrong? Doubtful, we move on to the next topic.
Is it to provide “feedback on learning”? Now that is starting to sound edu-babble enough that it may well be the case in many schools. If so, tests are not a great way of doing it.
I would test whether or not I was expected to. In general we test for many reasons, but particularly as an incentive to learn! Marks matter to kids (as they should). I’ve also found that individual test results have surprised me on student ability assessed through walking around and tests are essential guides as to how a kid is really going. Life has many tests (where you are required to make decisions under time pressure) and preparing a student for the real world is one function of tests.
However for an individual assessment, it is important that it be tailored to the particular outcomes you want to achieve with a group of students. eg I taught a streamed class of kids who were about 6 years below expected in mid high school. A very important goal was to give them some confidence in maths so everyone who worked in class passed (which was the whole class).
I should also emphasise how hard it is to write a test from scratch at the right level. I often had to scale marks for this class as I made the test too hard or too long etc. A good reason why just grabbing a previous year’s test which has itself been ‘tested’ and refined over the years is not such a bad thing.
Thanks, JJ. I totally agree that a test must keep the objectives, of both the teacher and the students in mind. As I replied to RF, I’m fine with recycling tests, unless the tests suck. I’ve never seen a teacher care whether or not the recycled test sucked.
Why do you think a test is not a good method of determining (and providing evidence for) what a student can or cannot do?
Whether a test can be such and whether the majority of tests I have seen this year are such are two different questions and have opposite answers.
So, to be clear – Yes a test can be a good (possibly the best) method for determining what a student knows/understands/needs but No I do not think a lot of the tests (I am going to go so far as to say the overwhelming majority) I have seen this year achieve that aim.
Why not? A few reasons:
1. Much like a good lesson, a good test requires some thought and planning. Why are you writing (or copying from elsewhere) that question? What skill are you actually wanting to test and is it written in such a way to actually test that skill?
2. Once you get beyond the most fundamental “can a student do this skill?” questions and into the world of actual problem solving or deeper thinking, how do you make sure a student who answers correctly is not one who has been shown a similar problem by their tutor (or weekend school)?
3. If you allow students to bring in CAS calculators and notes, what are you really testing?
You are saying tests suck. I agreed, multiple times in the post. That has nothing to do with whether proper testing, after and aligned to proper teaching, is valuable.
Do people find the multi-part questions unfair because each step of the way is marked? While, if it was just a single question there wouldn’t be as many marks allocated and hence the marking not focused on minute details.
I don’t teach in VIC but NSW and there hasn’t be an aversion to multi-part questions here (That’ I’m aware of). Usually the questions are in two parts (four at most) and are seen as scaffolds. Also, any errors are not penalised again in proceeding parts.
Perhaps the parts in NSW are larger and/or deeper than the parts in Victoria. Without question, the multiparting in Victoria leads to a focus on trivial matters of form at the expense of reason and solid technique.
Hi, Simon. I agree with what RF and JJ write below (and to which I’ll also reply).
A) and B)
It all depends upon context. There is obviously an issue with students (and all humans) forgetting and/or siloing information. So in principle I’m fine with testing broadly, including techniques and information from previous topics, but the “fair game” should be properly fair. Beyond students being told what is testable, the broader range of the topic should have reasonably have been covered in class and/or practice exercises.
C) I’m not quite sure I understand what “implied knowledge” means, but I hate multipart questions, at least at the senior school level, and probably generally.
D) At lower year levels I think mostly “same format” is ok. The material is, or at least should be, solid and standard and mechanical. Can the kid factorise quadratics or not? At the senior school level, I would naturally expect a mix. There is still plenty of straight-forward questions to be asked: solve thsi integral, graph this function, and so on. But there are, or at least should be, depth ni the topics, and enough overlapping topics and techniques, that natural non-formulaic questions can, and should be asked.
No, they shouldn’t. They should teach what is necessary (at least) and (moreover if time allows) what is also good to know. And then they should test what part of the stuff arrived at the recipients and is internalised by them. (In my country, there were times when that was possible.)
teachers should test only what students have learned?
That’s not what I wrote and meant. They should test what they taught. Not what the bureaucrats want to be tested. Teachers should plan their lessons first, and then design their test.
I’m not sure how you’re disagreeing with me, extent maybe to the extent that the teachers are free to ignore bureaucrats (and their co-teaches and Head and Principal) in order to teach what they want and should.
For year 11 and 12, I refer to HSC questions to give me ideas about the extent to which a topic can go (e.g. difficulty or how it links with other topics). Often I use past HSC questions for revision or make quizzes from them. I don’t have a problem with teaching to the test because the HSC questions are generally decent. As Marty pointed out, teaching to the test gives clear expectations to students about what to learn and the rigor expected.
Exactly this.
Re Victoria – I wonder if it might not be time to switch back to full exam assessment in maths (and perhaps more generally given AI and the virtual impossibility of detecting cheating in unsupervised assessments in many subjects). My impression is that SACS seem to proliferate throughout the year and create heightened arousal and stress across a fair bit of the year, whereas mid and end of year exams written by VCAA were suggested to me as a means of reducing teacher workload as well as student stress and anxiety.
If there are SACs out there creating “heightened arousal” I might have to rethink my views.
Of course it is time to bring back full exam assessment. The current system is utterly mad and utterly maddening. Everyone is going nuts.
In VIC we no longer grade tests by marks for Years 7-9 at least . A student’s result is At expected level; above/below expected level; well above/below expected level.
Since when? By whose orders?
Google “Five point scale – Education | vic.gov.au”
No.
That does vary quite a bit in the independent system.
The blanket requirement in Victoria is that students are graded on a minimum of a 5-point scale but beyond that it is up to the individual schools to decide.
Teaching to a specific Test sounds like leaving any potential joy of math out of question. Also it excludes exposing students to the concept of “math as an art”. So statistically or pedagogically it may be great, but such motto officially says that anybody who wants to consider pure math as a carrier path should look somewhere else. At least it’s honest.
I don’t see why you cannot test fun and beautiful and hard mathematics.
You can, but it requires a level of planning and forethought that you won’t find too often.
Compare the recent TNDOT posts to the VCAA exam posts and it becomes evident that what once was is now not (for many, intertwined reasons).
Of course it’s not easy in the reality of stupid constraints. But I think the question Solange is raising is if it’s even theoretically possible.
I see no reason why it is not theoretically possible. Most of the comments here seem to be arguing about the practicality of various settings though.
The most memorable moment for me this year is when my year 11 Specialist class completes the test on Functions and graphs (trig inverse, reciprocal…) on the last lesson before the term break, instead of rushing out of the class, they all hang back and discussed. One of the best kid exclaimed: “I really like the test, it’s different from the textbook question and challenged, but really doable and fun.” And it made me extremely proud and please because I had spent so much time and effort to write that test, changed them each year for the cohort I taught. For me as a teacher, the process of writing assessment is a never ending learning curve, doesn’t matter how long I have been teaching, every time I write a test or an investigation, I learn something.
I used to have a colleague, he is very very good and he often wrote a SAC with one general question like “Investigate this and that…” without any further instructions/ scaffolding. It suited well with one or two best students as they would be able to freely expand their wings. But it really a nightmare for the the lower kids who sometime felt directionless. With regards of Vic system, I think more flexible assessment can be done at junior levels, but for VCE, I just need to play the game, otherwise I will do my students a disservice.
Thanks, Victoria. It is unfortunate (?) for teachers that the textbook questions are so awful. What is required is a very large bank of questions from which teachers can choose. That doesn’t mean that writing a test is then simply selecting such questions, but it’d certainly speed things up.
“The fundamental purpose of assessment is to establish where learners are in their learning at the time of assessment” (Masters, 2013).
I remind myself of this principle every time I set an assessment task.
What was interesting to me was that when I looked for reference material on tests (including such as making the first question easy and gradually increasing in complexity), I couldn’t find anything much online. I also looked for an expected range of marks for my class due to the fact that I had to use poorly constructed tests over which I had little control for my streamed class and needed to scale the marks to an appropriate level to ensure the students (who were working hard) did not lose confidence.
I was surprised at how little I could find out there.