Grading and testing go together in some ways: both are too often conceived of in narrow ways and too often seen as a kind of punishment. It seems almost proper for students to dread tests, for kids to hate report cards. But this comes out of a tradition and we have to decide if we want to follow that tradition.
There are two big issues with grades or two uses: motivators and evaluators.
Motivators: I’ll have to lay my cards on the table here. I think good grades or high grades tell the student what he already knows and, following Alfie Kohn, can actually undermine the motivational motive. And haven’t we all know people who made very good grades but showed no sign of having learned anything? If you haven’t, perhaps that’s a sign of where you stand on this issue.
Poor grades, OTOH, demotivate. Often, most often, they confirm what the child (notice my drop into the more emotive “child”) already suspects, that he is just no good. Today I heard a caller to a talk show on NPR discussing cheating say that he cheated only in Spanish class b/c he was no good at Spanish. He wouldn’t dream of cheating in his other classes (see my blog post “Just No Good At Languages” of this date)! Pretty amazing.My take on this is probably very different from that of a lot of teachers: I think “just no good” is a popular way of saying that a fl as a list of rules which were to be memorized and then used to produce sentences in a fl the student had no knowledge of – a memorized rules are not knowledge – was so utterly confusing that he felt JUSTIFIED in cheating. Yeah, I know, I have no moral fiber.
Evaluators: grades are not evaluators, they should reflect evaluators. Evaluators are the various ways in which the teacher evaluates the effectiveness of the instruction. No, I am not leaving the student out of the equation; it’s all a package. But what we should be doing is figuring out whether and how much and how well students are learning. That can be done in many ways, incl tests. But no matter what the teacher employs to evaluate the learning going on in the classroom, the grade is not the evaluator, although it is often taken that way.
So before we start throwing grades around, we need to be very wary of conflating a lot of behaviors: compliance and defiance, studying and learning, learning skills and test-taking skills, motivated vs unmotivated (more code words), low intelligence and high intelligence, beauty and ugliness……. wow! Where did those two come from? Let’s be honest: tall, good-looking, self-confident people tend to be given the benefit of the doubt. We won’t go into bias of various kinds here but biases must be taken into account. For instance, I like kids who talk about things, so I must be sure I do not overlook kids who have been raised not to be “forward” or “bold”. Other teachers who may be shy have to watch they don’t get a negative attitude toward big-mouth, self-confident kids who are nevertheless respectful and thoughtful. Of course, giving the star quarterback a passing grade even though he doesn’t know what Hola means is OK.
Testing is a huge subject but there are certain principles: test what you taught, create tests to access a variety of things learned, use various modalities of testing, use tests as a means of more instruction.
Those are a few and I’ll elaborate on each. Textbook companies often provide tests or educational entities like districts provide them, or some other source e.g. tradition provides them. I would be extremely cautious about using other people’s tests. In the interest of testing what you taught, it would seem proper to adjust testing to the conditions each class, each year encounters. Personally, I’ve never given the same test over the same material. I use boiler-plate questions but each test emphasizes different parts of the topics covered in class.
This conflicts with the notion of standards, at least on the surface. How can we know we have a standard to teach to unless we test for that standard and how can we have those tests unless they are the same for everybody? That makes sense. However………
Standards are for everybody, not just the group they are momentarily being applied to. Teachers have standards to meet, schools, school districts, cities and towns. If a town allows a child to go hungry and without medical care, it is absurd to insist that that child meet a standard designed for a well-fed and healthy child. If you want to do that – and many people in America seem to want to do just that – then you must abandon the idea of equal opportunity. It’s not there.
What happens is that teachers and schools try to bring their students as close to the standard as possible. I know that if I were to teach a grammar-based syllabus, 90% or more of my students would be unable to meet the standards of such a syllabus b/c I would have to start from scratch teaching them grammar. I’ve surveyed my students, so I know what I’m talking about. So we have a mish-mash of teachers desperately trying to get students up to par (or savagely slashing them from their classes as being “unprepared”) but with no consistent manner of doing so.
How do you test in such a situation? You devise tests which give a broad range of students an opportunity to show they have learned. What if you gave an oral test in a fl class, who would shine? The out-going children, of course. What if you gave a grammar test replete with formulas to follow in order to produce the correct grammatical form? You would get the logico-mathematical crowd but precious few others.
In other words, you have to give tests that allow students to show they have learned; exactly what they have learned will vary. If you want to, you can insist that a student master a certain ability e.g. writing paragraphs, labeling pictures, telling about his day at the zoo, whatever. But if you are observant and – dare I say? – sensitive, you will notice that students you KNOW have learned will not do well on the test. That is b/c you have addressed only one channel of learning. You can insist, if you want, that everyone conjugate verbs in the present tense and know the capitals of countries and match TL words with English words, but you are going to miss testing a lot of knowledge students pick up.
I would study testing, not just fl testing. Lots of stuff has been written so make sure you don’t read only literature produced by the test makers. They have an agenda. So does everyone else, but read them, too.
Now consider this: it took the state of Arizona many, many tries to get the state’s high-stakes test right so they could keep kids who couldn’t pass it from graduating (worthy goal, that). So do you really think that you, who have probably never taken even one full class on test construction, can do better? Dare you put so much weight on your tests? Even less just, can you put so much weight on a publisher’s test or a district test written by a committee? Should a kid’s grade hang on that?
So what to do? Here’s what I do with tests: if they show a kid has learned, I figure the test grade in. If they don’t show that, I look at what else the student has done and decide whether the test is an accurate assessment of what this kid knows. Remember, you’re not testing just knowledge but the ability to display that knowledge in a highly restricted setting.
Let me give you a simple example. Some scholars put a lot of stock in IQ tests. They assume that, given decent testing conditions, the test subject will perform up to his/her ability. What they fail to take into account is that test situations are quite foreign to many people. One wonders what would happen if someone thrust a puzzle at one of the test writers and said, “Here, solve this” what would happen. My guess is that the test maker would try to solve the puzzle, try really hard to solve the puzzle b/c that is the mentalityof the test writer. Solving pointless puzzles is fun!!!
Now tell a 12 year old kid who plays outside all the time to solve a puzzle. He might wonder why you were doing this silly thing no one in his life has ever done. His dad never does puzzles, his brother works on the farm and never does puzzles, what good is it? Such an attitude completely baffles test writers b/c they grew up valuing this sort of aimless behavior; they consider it a fun game. Moreover, by engaging in this fun game, they have learned a lot about how to solve these puzzles.
Quick, what does “fetch” mean in “the fetch of the wave”? Hey, I thought you were an educated person? OK, another chance: tarnation is to frustration as “land o’ Goshen” is to wonderment – true or false? Don’t know those expressions? Well, I do.
Now I realize that test making is a lot more sophisticated than that, but at bottom, we have a problem that a lot of students simply do not have basic test-taking skills and that must either be addressed first or the teacher’s tests must take that into account.
I find test-making fun. My district gave us a mandate to give tests in various formats. I did that. Most teachers used bubbled scantron sheets. My tests prepared students for various formats; theirs didn’t. I also throw out a test as a whole or for an individual student if the test doesn’t seem to me (I’m the expert) to have allowed students to show what they knew.
One very good teacher I knew used to give grades out based on calculations down to several places – you know, eighty-seven point seven seven nine or some such. Very, very precise, very mathematical and therefore unquestionable, at least by the typical working-class parent. To top it off, it was figured and printed out on a computer! Unassailable.
But one day I noticed students giving presentations in front of the class and I stood in the doorway and listened a bit. Later, I congratulated the teacher on the neat projects and asked her how she graded them. A rubric. Really, and what generated a grade? The points given on each part of the rubric. Wow! Impressive. How many points? Oh, they can vary 5 points here or there!!! And yet a kid could drop a grade or even fail based on one one hundreth of a point, easily wiped out by a slosh of one or two points one way or the other on the rubric. Think about it. It’s all about power and safety, isn’ t?
I’ll leave you with this from the Passionate Teacher by Robert Fried (p. 212)
the work is important to student learning, I should assign it.
If I assign it and they do it, I must read it.
If I read it, I should critique it.
If I critique it, I must grade it.
If I grade it, I must enter that grade in my book.
If I enter it, I must average it with other assignments.
If I average it, it must count toward the course grade.
Now I have heard teachers defend this to the hilt right after telling me how they fudge in order to get a kid’s grade to come out the way they think it shoud OR, sadly, give the kid a low grade they don’t believe reflects what he really knows. And then we complain when kids cheat? See, again, the blog entry Just No Good At Languages.
What I base grading on is that the student DO the work. It is my job to see that the work is such that if they do it, they will learn. So while I give what I label “a good old-fashioned vocab test” for fun and grade it as a discrete item test, my real tests ask students to do as much as they can e.g. label pictures, write a poem, make a narrative from a series of pictures or a series of phrases, make a story about several characters or change a story. A lot of what I do is have them break a story down into its component parts (characters, scene, conflict, turning point, plot, outcome, etc. – one English teacher told me the last week of school she likes getting my students b/c they know how a story is made up).
So it’s kind of hard to make a low grade in my class if you do the work. I do “count” grades in the student’s strong areas so the final grade might be based on a couple of things for one kid but on a couple of other things for another kid. The outcome should be that they can read Latin. One teacher whose kid was in my second year class relayed his daughter’s comment that she had learned nothing that first year but in my class, we did goofy things like color pictures but somehow, she realized, she could now read Latin.
Several stories from other parents like that and from teachers give me feedback that the students are carrying Latin outside the classroom as well (reading Latin in church, etc.). But you can’t say that every kid will do such and such, just as you can’t say every teacher will do such and such. So that makes it hard to get into the “data-driven” stuff. I think that comes from the same people who brought us NCLB and charter schools.