I awoke far too early this morning and as a result spent quite a bit of time lying in bed thinking of random stuff. One of those random thoughts was about my time in university and about one of the nastier types of exams we had to undertake. From memory the class was entomology.
For those who don’t know, a punitive exam is different from a regular exam in that a correct question would be +1 while any incorrect question is -1. So if you got 50% of the questions correct and 50% incorrect you mark would be 0.
As you can imagine everyone scored lowly (and I felt so bad for the folks with a minus score) I think I scraped in a pass.
I really cannot think of any purpose for such an exam other than to put dent in peoples self confidence.
If anyone else has had to undergo one of those horrors please feel free to give your story. Anyone who has actually set such an exam for you poor students feel free to give your reasoning.
It is to discourage guessing. For example, most standardized tests (SAT, GRE etc…) do this. They are multiple choice, with 4 possible answers. An incorrect response is worth -0.25 points. On average, guessing is a wash.
If formulated correctly, leaving an answer blank shouldn’t be punitive, just wrong answers.
Also, I’m an Entomologist During my training, we weren’t allowed to use keys for identification - I was trained in sight identification (at least down to family, but down to species for anything local). Still comes in handy when I need to know (generally) what something is, but don’t have an ID key handy.
And my prof was a real joker. Did things like pulling the antennae off of adult antlions and putting it on the exam (in this case, without the antennae, it looks almost exactly like a damselfly - completely different order - unless you looked REALLY carefully). I think we had a punitive component as well. Doc didn’t want us to guess. he wanted us to KNOW.
Concur. It’s been a long time since my education classes, but I seem to recall that the purpose of an examination is to prove that the student understands the coursework, not to enhance their self-confidence.
If you don’t know the answer, leave it blank and you won’t get the negative effect.
I had a Paleontology professor who was like that. He loved to throw in all sorts of odd bones into otherwise uniform “deposits,” just so we wouldn’t get complacent about what we were categorizing.
As long as it gets translated into some sort of reasonable grading scale at the end of the class, I don’t think it matters if individual tests are given scores in regular percentages, logarithmic scales, or number of smiley stickers. I do like the general idea of punishing wrong answers though. If you don’t know the answer to a technical question in real life or on the job, you really should just pass rather than make something up hoping it is right.
It would be interesting to see how exams played out if they were REALLY punitive. If its plus one for being right and minus one for being wrong its only a penalty if you are making a wild assed guess (and even then not really). If you have any inkling of the right answer its worth taking the risk.
Wonder how people would react if it was plus one for being right and say minus 10 for being wrong?
I once got in a fair argument with a professor of mine. He claimed you needed to know everything. I pointed out how many PhDs in different scientific fields (and variations in the same field for that matter) the school offered and that it was impractical. My position was it was more important to know what you really knew and know what you didnt.
They actually used something like this in a downstate Illinois prison. Prisoners who had good behaviour got ONE point each day. If they had an infraction they lost FIVE points. And everyone in their block lost an additional point as well.
My daddy, a Civil Engineering professor, would give negative grades without batting an eye. His exams were all word problems, however. You might have one test which was just one or two problems.
I remember making the highest grade in the class on one of his Structures III tests one time.
My osteology professor was the same way on practical exams where we had to identify human bone fragments and, if possible, the age, sex, side, etc. It wasn’t unknown for her to put in something like a fragmented calf sphenoid. It was more fun later when I was her graduate assistant and got my chance to haze in turn.
Again, the whole point of this was to force the student to really look, be able to defend his answer, and not assume. You didn’t want to classify a hydrocephalic calf cranium, for instance, as a human cranium (a cautionary tale which was actually done by one anthropologist).
Mathematically, unless there is a third option (i.e. 0 for a blank answer), there’s no difference between a normal 1-point-for-correct test and a punitive plus-or-minus-1 test. The former has all questions originally worth 0 and by answering correctly, you change it to +1, thus gaining one point. The latter has an original score of -1 and you change it to +1, thus gaining two points. The former has a range of 0-100 while the latter has -50 to +50.
There’s no difference between the two other than the arbitrary number that results. It’s the same as if you made each question worth 10 points, or 5, or a million. Having a negative number as a possibility changes nothing, and thus, we shouldn’t be talking about ‘punitive’ tests, but rather where blanks are different than wrong answers.
Because professors are sadistic pricks? We had a physics class FULL of questions like:
That exam was graded on a curve. A score of 30% passed, 35% got a B. One of my dorm mates had just enlisted, but hadn’t moved out yet. He went into the exam and answered every question with a ‘C’, except number 7, which he answered ‘B’ (Why? I dunno, he just did.)
He turned in his exam before the professor was done handing it out to the rest of the class…and got a ‘B’.
I had a wash-out Intro to thermodynamics class that all Engineering students had to take. It was given by a guy that had worked previously in an Ivy League College, and was called by some ‘The God of Thermo’, and Professor Hitler by others (Hitler roughly resembling his last name.
I took the course from him and dropped it when it was obvious I wasn’t going to pass…Took it AGAIN and failed. Took it from another instructor and missed a A by 2%.
Makes me wonder, if he was a God of thermo, why was he teaching a 100 level class at a State University?
I also knew a perfectly reasonable EE student. Smart…friendly…rational. That all changed when he became a Grad and started teaching.
I’m not following this. If we’re taking a 100-question test then the first range would be 0 to +100 and the second range would be -100 to +100 wouldn’t it?
But I agree with Chessic Sence in that it is only a rescaling. Why do students get so upset if the average score is 30 out of 5000 if the letter grades will be comparative anyway?
Now if a blank answer gets a higher score that an incorrect answer, then clearly the purpose is to discourage guessing, or when I do it, it’s explicitly to make the expected value of a guess the same as a blank answer. For example, if I have a True/Flase question, I make the scale:
Correct: 1
Blank: 0
Incorrect: -1
That way the expected value of a guess is zero. If you don’t know about expected value, pretend there are 10 such questions and you guess on all of them. Then you would expect to get a score of 0 (half right and half wrong.), which is the same you would get for not answering at all. Which is as it should be.
Usually, however, I don’t have T/F or multiple choice questions, as I teach math.
Yes, that’s exactly our motivation for everything. You got us.
I don’t understand the point of this story. If the questions were too hard, then of course you shouldn’t need 90% to pass, right? I mean, you seem to agree with the rescaling, right?
A tenure-track position at a large state university is a pretty good job. I know mostly Math people, but the top people in a lot of specialties are at state institutions. Also, that particular department might have had a lot of other good people in his particular specialty.
In high school, we took a nationwide math test with 30 questions, which gave 0 points for a wrong answer, 2 points for a blank answer, and 5 points for a right answer. So if you just turned it in blank, you’d get a guaranteed 60 points. Our teacher stressed this fact, and I imagine so did every other math teacher in the country. And the nationwide average was still 58.
Dad’s reasoning is that engineers need to know what they’re doing fully, and wild-ass guesses will get people killed. So, no wild-ass guesses on his exams.
On every test like this I’ve ever seen, the third option is in fact 0 for a blank answer.
It makes sense to grade like this if you want to remove the benefit that “partial” knowledge provides.
On the SAT, a random guess has expected value 0, but if you can eliminate even one of the options, the expected value is positive guessing between the remaining. This makes sense for the SAT, because someone who can eliminate an option does in fact know more.
But lots of examinations aren’t really like that. Either you are correct, or you are wrong. There might be multiple ways of being wrong, but unless you know the right answer, you are equally wrong. Making the expected value of a guess always be negative will weed out everyone who doesn’t know the right answer (and give you an accurate picture of which students don’t understand basic probability, too).
During my training, we weren’t allowed to use keys for identification - I was trained in sight identification (at least down to family, but down to species for anything local). Still comes in handy when I need to know (generally) what something is, but don’t have an ID key handy.