Speak for yourself. New York state has had “Regents” tests that are required for graduation for years.
While not traditional high school exit exams, students had to pass a standardized test that everyone in New York took in Math (9th and 10th grade), Science (Earth Science and Biology), Social Studies (all four years), English (all four years) and Foreign Language (taken after four years of foreign language instruction).
By my count, I took 13 of them, and if I failed one, then I failed the class and could not graduate.
I’ll pull out an exception then. First, I’m a 99th-percentile scorer on every standardized test I took in high school: SAT, PSAT, SATII, and so on. I nailed 'em, to the point where I could walk out and tell you which two questions I got wrong.
But once, just once, I was unable to take a test. I was taking my Honors Physics final in high school, came in to sit for the test, and started shivering. Sat down, opened the test booklet, and discovered that (a) I could no longer read English, and (b) 45 minutes had just disappeared. Freaked out by this weirdness, I approached the teacher and said “I think I’m going to throw up.” He excused me to the bathroom, where I did in fact throw up. I went back in, told him I was having serious problems reading the test and he said “If you’re sick, go to the nurse. But tell her not to have your books sent up to the infirmary - you have to finish the test in quarantine.” On my way to the nurse’s office I became convinced that one of the trees was following me.
It turns out that dehydration, over-caffeination, a little exam stress, and lack of sleep had combined to give me a panic attack. The nurse said I was barely coherent when I got to her office. A few hours later after a nap and some water, I finished the exam and managed an 85, to which my physics teacher replied “Wow, you really were sick.”
If what happened to me that one time is what happens to bad test-takers every time, then we need to figure out how to coach them through it before we can start requiring students to pass a bubble test.
Some people (I’m still conflicted on the issue) think that students should pass a standardized comprehensive exam *because *I and my classmates didn’t. And when asked questions by David Letterman’s man on the street, we look like flaming idiots who can’t find Canada on a map. And when tests of high school and college students were compared to their age mates in other countries, we were starting to look like a whole nation of flaming idiots. And when skilled jobs started moving out of our country and to those countries with higher test scores, some people started to (rightly or wrongly) connect those dots and decided that we need to improve education and make sure our high school graduates really know what it is we think they know so we don’t end up a nation of people who can only work at The Gap.
This was a huge issue when I was a kid. I did several debates in high school on the topic of Standardized Curricula and Testing, and some of the points I used were things like grade inflation, wretched test scores and also the growing movement of students from one school district to another. It made sense, back when people didn’t move much, for the local school to decide what its own students needed to know. But that meant that a 5th grade curriculum in Detroit might be very different from a 5th grade curriculum in Peoria. Which, you might say, doesn’t matter as long as all the important stuff is covered by graduation. But what about that kid who moves from Detroit to Peoria, and so never ever covers some of the fundamental units?
We saw a problem: kids were falling through the gaps, in what they were learning, in when they were learning it and in how they were evaluated on it. Graduation testing and NCLB are both attempts to identify and prevent students from falling through those gaps. NCLB, is, I think, a disastrously stupid way to do it, for reasons **LHoD **mentions and others, but I do think it came from a well meant position. Graduation testing might not be the perfect answer, either.
If I was to wave a magic wand and order the World According to Me, I’d offer a portfolio option for graduation. But portfolios can’t be standardized, and require greater money and time to evaluate, and they still leave one open to the same biases and conveniences that make grade inflation so tempting.
I was thinking that just guessing, even without elimination, tends on average to give you zero points. Since there are four wrong answers worth -1/4 each and one right answer worth +1, I figured that with large numbers of questions it all basically tends to cancel out. (I hadn’t thought of the fact Indistinguishable pointed out–that since there are more wrong answers than right answers, your chance of getting a negative score tends to be higher. But I take it he agrees that with large numbers of questions, this effect becomes negligible.) Based on this, I was thinking you might as well guess in every case when you don’t know the answer, whether you’ve eliminated answers or not.
Oh man, did I love those. When I went to school there was a policy that if your score on the Regents exam was higher than your class average they would substitute it for the average. If it was lower the Regents grade would be averaged in as 1/3 or so of your final grade.
I abused this system to the utmost, blowing off assignments, skipping classes, carrying a good solid F through the entire school year. I’d show up for the Regents, pull off a 90%, and go home on the honor roll. My senior year I missed 43 days of class out of 180 and finished with an 88% average. They changed that policy the subsequent year to say that missing more than a certain number of classes was grounds for failure, regardless of class average.
I felt a little guilty about spoiling it for those who came after me, but I didn’t make the rules, I just lived under them, and if I could get a 90% in a class without attending, the instructor was clearly not providing me a service. I was never a discipline problem except in a Bartleby kind of way, I’d sit in the back of the class and read quietly. After the first couple of interactions the instructors would usually decide to let me continue my meaningless rebellion, lest I find a meaningful one.
Are you saying that a student who has the competency to do well on a** reasoning** test wouldn’t be able to pass a high school competency test? Granted, I haven’t taken either test in a decade, but they seemed pretty similar to me at the time (do some algebra, read some graphs, read and interpret a story passage, etc.). I find it hard to believe that a student could do well on a college entrance exam and not be able to pass a high school competency exam.
Ah, and that was the point I made earlier - if there was that level of test anxiety on every test, it certainly would have manifested itself much, much earlier than the fifth time the kid took the AIMS test.
Okee doke! Didn’t want to hijack much more…
Well, they’re just totally different tests that measure totally different things. Passing the SAT isn’t going to help a student on the Science portion of the AIMS any more than it would help on the AP French test.
I think I understand what’s going on now. You and I are speaking of two very different things.
You’re speaking of the number of negative scores, and I’m speaking of the overall score on the test.
If I can narrow every question down to two potential answers and then I guess, I’ll get half of them wrong, but my average score will be 0.75 points per question. If there are three potential answers, I’ll get 2/3 of them wrong, but I’ll still average 0.5 points per question. With four, I’ll get 3/4 wrong, but I’m still going to average 0.25 points per question.
So, you are quite correct that guessing will get you many more negative scores (on individual questions) than leaving questions blank.
I am also correct, in that even .25 points per question beats the heck out of zero points per question (what you get if you leave it blank).
entirely different types of exams. The Scholastic Aptitude Test, for example, is not a test of understanding, but rather a test of aptitude for learning. An IQ substitute test, if you will.
A person can do well on the SAT (and to some extent on the ACT, which is more of an achievement test), and still not have the knowledge to pass, say, the Ohio Graduation Test, which asks some very tough questions (do YOU know the ins and outs of the industrial revolution and its effect upon workers in Great Britain?? - I sure didn’t in high school).
I suppose he wanted to take this hijack outside the thread, but, judging from the PMs I’ve exchanged with him, I don’t believe that’s what ArizonaTeach’s point is. His claim is still that the probability of getting a negative score overall is significantly greater than that of getting a positive score overall, randomly choosing 1 out of 4 possibilities for each question. His claim is grounded in the assertion that the probability distribution of answers to distinct questions are not independent; that is, there are certain statistical patterns linking them. For example, he asserts, the probability of four answers in a row being the same letter is extremely low, much lower than would be expected if separate answers were independently distributed. In fact, he believes, this has only happened twice in the last 30 years of the test. Similarly, he has also been told that the test designers ensure that answers are approximately equal in number, each letter being constrained to a frequency of 20% plus or minus 3%. If separate questions’ answers are not statistically independent, then, I agree, it is mathematically possible for the probability distribution over answer keys to be such that random guessing to be very probable to hurt overall.
However, I find such claims about the lack of independence of separate answers hard to swallow, as they suggest the tests are susceptible to being “gamed” in unfortunate ways, through more “sophisticated” guessing (have three answers in a row that you know are As? Don’t even look at what choice A is for the next question; strike it out. Filled in all the answers you know? Find the letter you used the least often, and fill it in for everything else, guaranteeing (or almost guaranteeing, thanks to the 3% wiggle room) a score at least high as you would receive had you left them blank, and very probably much higher). But I lack authoritative information from the test designers on what exactly they are doing, and some people in the business of test preparation seem to be passing these claims along.
Though I continue to be skeptical of what ArizonaTeach has been told about the design of the tests, it is not my intent to badger him about this, which I fear I may be doing by continuing to voice my disbelief/skepticism/argument, particularly given the view that it is a hijack of this thread. He did give me some links, though, for others making similar assertions about the distribution of answers; for example, here and here. Quoting that last link may be informative, in seeing once again the kinds of premises he has been taught on which his claims about guessing can be grounded:
The particular SAT preparation company which he has experience with uses slightly different numbers, but the same basic idea. Anyway, there you go.
I’m Martini Enfield and I have crippling exam anxiety.
It started in 7th Form and got worse from there, to the point where I’ve had to put my university studies on hold because I just can’t handle being in an exam. I’m not usually an anxious or nervous person by nature (at least, I’d like to think I’m not), but when it comes to exams I completely freak out and my brain usually decides to take a holiday just after the exam starts, clearing off to go Tiger hunting in Africa* or being a Secret Agent for MI5 or whatever the hell my brain does when it leaves me in the lurch during exams.
In short, I can swot for weeks before an exam and know the material back to front, but as soon as I get into the exam, I either get the mental version of that black & white “Please Stand By” test pattern with the Indian Chieftan on it, or I completely freak out and have to resist the urge to write something like “I Am A Fish” 500 times on the exam paper before passing out in a cold sweat.
I’ve tried everything from relaxation CDs to a double bourbon right before the exam to calm down, and it doesn’t work- I either get a mental blank (“Carlill v Carbolic Smoke Ball Company*… nope, doesn’t ring a bell.”*) or a panic attack in Imax 3-D with Dolby DTS surround sound.
What’s more, once I fail an exam, my anxiety gets worse- making it less and less likely that I’ll pass anything. And what’s worse is that because I’m a smart person (so I’m told, anyway), people expect me to do well, and end up being Very Disappointed in me because I failed so badly. Which makes it worse.
In short, it is entirely possible to be smart and still fail exams spectacularly, despite knowing the material thoroughly. In fact, the whole thing is very distressing for me and this is probably the first time I’ve really talked about it.
*Yes, I know there aren’t any Tigers in Africa
**Famous case in contract law about unilateral offers; wikipedia article here if anyone’s interested
Yeah, that’s the basic gist of it. I guarantee that any SAT prep company you encounter has done the same research and come to the same conclusions…there is a structure to the SAT family of tests (although not necessarily a pattern) and it is entirely possible to game the test…not to get a perfect score, but to improve the score significantly. Interestingly enough, I came across this link which suggests that in the LSAT, they intentionally change up the rules to throw people off - but not in the SAT. They don’t mess around with the SAT.
Thing about the College Board is that it’s all “neither confirm nor deny” so the only way to check up one what they do is to get your hands on previous tests, and that’s what these reputable companies do, and then spend a lot of time and research studying the test itself.
You know, I hadn’t realized that tests had become such a big part of the American education system. It really wasn’t very long since I was in public high school… when did all this change?
And, for that matter, if your standing really is dependent on these tests, then what’s the justification for grading the kids on anything other than the test? Why not allow the bright bulbs to test right out of high school classes completely? Why penalize them for truancy, late homework, etc. if they still end up scoring high enough?
Absolutely. I worked for one of those companies for a year, and I can take the SAT right now and score a perfect 1600–just by doing exactly what I was told to teach students to do. I won’t even have to read the questions in depth. Skimming will be sufficient. I know that sounds ridiculous and impossible to believe, but it’s absolutely true. There is a very specific logic to how the SAT is designed, and if you know the way it works, you can nail it. I never taught the students a single thing besides how to take the test. That’s ultimately why I quit. I didn’t care that I was helping people game the test, but I was annoyed that they weren’t learning anything except how to game the test.
It works on the GRE, too. I did horribly on the math section the first time around and didn’t score high enough to be given a TA position in grad school (though my verbal and essay sections were quite high). I took the GRE class (for free since I worked there) and retook the test and improved by about 300 points–and I didn’t all of a sudden become a genius in math.
I read the entire thread, and I don’t think anybody mentioned one very key point–it’s possible for high school seniors to simply not care. I’ve been teaching a college level course at a HS all year. These students know that if they fail, they will have to retake the class next year. They know if they fail, those marks will be on their transcripts and bringing down their GPA for the next four years. They just don’t care. I’m dead serious. It’s very, very difficult to make them understand that the decisions they’re making now are going to impact things down the road, and in unexpected ways. You can have all the achievement tests you want, give them ample opportunity to pass, teach to the test, threaten their prospects of graduation, etc and if they don’t care, there’s absolutely zero that can be done. I’ve begged, cajoled, threatened, warned, contacted the principal–but I swear to God, there’s nothing more stubborn than a teenager who thinks they’re about done with HS. And the added threats to the school and district by NCLB won’t make them care, either. I sure as hell didn’t when I was a senior.
I think test anxiety does exist. I’ve been fortunate enough to never experience it. But I think a far more likely culprit than test anxiety is dontgiveafuck-itis. That might be difficult for a logical, thinking person to comprehend. Why would somebody choose to fail a class or a test if they know it could ruin their chance of graduation, make them lose scholarships, etc etc? Well, if somebody knows the answer to that question, please PM me. Because I don’t understand why I have to go talk to the principal tomorrow about a student who decided to plagiarize her final assignment, and two other students who just never turned anything in at all.
I don’t want to bore you with the details, specially since they involve a completely different school system, but I’ve flunked tests and even done things like forget to take an essential tool due to anxiety over paternal threats to make me drop out; and, a few days later (when everything was already officially lost, therefore I had nothing to lose), aced an exam for which I hadn’t studied, on a subject I couldn’t have cared less about, because I was so relaxed you could have used me for pillow stuffing.
While I agree that there is only so far we should do to accomodate differences (I definitely have no desire to drive over bridges designed by an engineer with zero ability at math), and if you flunk you’ve flunked and have to eat up whatever that sends your way, saying that a well-prepared student can’t flunk because of nerves… takes some nerve, it does!
