Why are people so opposed to standardized tests? (the ACT, school, interpretation)

[mitsguy2001]

Originally Posted by ddm2k

Was "regatta" material to the outcome of the question? Or could it have been replaced by any noun and the test taker still be able to solve the problem?

The story I was told was that it was an analogy question, with Yacht : Regatta. I don't know what the other half of the analogy was. And I was told this 25 years ago, so, assuming it was true, it was an exam from a long time.

I was also told that they threw out a question that used the word "low" to mean the sound that a cow makes. Even though we were told that getting the question thrown out was somehow discriminatory against minorities, I'm sure the real reason why it was thrown out was because "low" has other meanings, and the sound a cow makes is not likely to be the first thing that comes to mind, whereas "yacht" and "regatta" only have one meaning. I have no idea why throwing out the "low" question is discriminatory against minorities anyway.

[mitsguy2001]

Originally Posted by ddm2k

The inconsistencies I'm talking about would be educators taking liberties to make the following assumptions:

- Spelling error, or missing hyphen in a name = 100% wrong ("Guillain-Barré" as "Gillian Barr")
- Two-word answer, words are out of order = 100% wrong ("Myers Lake" as "Lake Myers")
- Answer written in pen instead of pencil, or "illegible" (highly subjective) = 100% wrong (strikeouts and erasures aren't harmful to a teacher's health, as frustrating as they may be to read around or read through, you still have to grade the paper and consider the answer, that's your job.)

I disagree with all of the above being marked as incorrect. Unless the spelling error or transposition result in a full match to another term being discussed in that chapter, it should not be marked incorrect. The instructor knows what was meant by the student.

And the real problem is inconsistent policies. Suppose a high school has 4 teachers teaching the same class. Teacher A marks any spelling error, no matter how trivial, as being completely wrong. Teacher B counts spelling errors as correct, as long as he/she knows what the student means. Teacher C gives half credit for mis-spelled but correct answers. Teacher D counts mis-spelled answers as correct for the football quarterback, cheerleading captain, and PTA president's child, but marks them wrong for other students. The students who get Teacher B for that class are going to be at a huge advantage, and the students who get Teacher A are going to be at a huge disadvantage. The SAT and other standardized tests, even if the rules are flawed, at least we are all playing by the same rules.

Again, I use the football analogy. Would it make any sense for one team to get 6 points for a touchdown and for the other team to only get 5 points and to not be allowed to attempt 2 point conversions? Obviously not. But we seem to accept this when it comes to teachers grading students.

[mitsguy2001]

Originally Posted by ClaraC

I always did well on standardized tests, although I am glad colleges are getting away from using them.

They really don't seem to be a predictor of future success as a student, and as an adult.

They just point out who is a good test taker.

Rarely, if ever, in life are you presented with a problem and four possible correct answers that you have to pick from. Life doesn't work that way.

I disagree. What if you have to choose between 4 jobs and can only accept one? Or considering 4 houses to buy and can only buy one? Or considering moving to 4 different cities and can only move to one? Or you can choose between 4 different routes to work and can only choose one. Most of life is choosing between multiple choices and with limited knowledge.

[mitsguy2001]

Originally Posted by ncole1

I'm not saying that teacher-written tests are intrinsically ”superior” to standardized tests, but at the very least if the teacher writes the test, then the test is based on the teaching rather than the teaching being based on the test. A mediocre teacher will be a mediocre teacher regardless, but a system with fewer standardized tests (1 per year in primary school and 2 per year in secondary school) gives good teachers more of a chance to really help the students without being too restrained.

Ok, but not every student gets the "good" teachers. Why should students be penalized for getting stuck with a "bad" teacher?

[mitsguy2001]

Originally Posted by steiconi

A friend taught grade school in Hawaii. One standardized test asked, "What season is it when the leaves turn red?" Answer: "Mango Season!" Because young mango leaves are red, and Autumn doesn't look the same in the tropics.

That's a good example of how some standardized questions may have multiple right answers, but only one that gets counted.

I took an IQ test that asked money questions that any child could answer--if they were familiar with pounds, pence, and shillings.

And that reminds of another point. Often, the instructions will say to choose the "best" answer. There will be no correct answer, but the teacher will argue that you were told to choose the "best" answer, not the "correct" answer. Which answer is "best" is a judgment call, and the teacher will only accept one as being correct. What I wonder is, are these questions intended to see how students think outside the box? Or are they because the teacher wants to cover themselves when they screw up and write a question with no correct answer, and allow them to hide behind saying they said to choose the "best" answer, not the "correct" answer.

As a real life example: suppose yourself and 2 others (3 total) are ordering pizza, and you each want 3 slices of pizza. Suppose the pies at the local pizzeria have 8 slices. How many pies do you buy? Do you buy 1 pie, and one person can only get 2 slices? Or do you buy 2 pies and waste 7 slices? No perfect answer, yet I'm sure teachers will insist that one and only one of those options is correct. And it ignores other options, such as asking the pizzeria to cut the pie into 9 slightly smaller slices, or buying 2 pies and having leftovers (albeit with no "fair" way to decide how to divide 7 slices among 3 people), or have some other item for the person who only gets 2 slices.

[tnff]

I wish CD would allow answers within quotes to carry forward. Would be so much better for the conversation. Sorry if this gets hard to follow.

Originally Posted by mitsguy2001

So then why aren't students who underperform on the SAT and other standardized tests just told that it "doesn't mean jack"?

For most people, in the long run, it means less later on. But it is a high stakes test for getting into college and scholarships. The SAT and ACT have a disproportionate impact. So for such an impact, the key point is whether they, and similar tests, actually measure what the purport to measure and predict what the claim to predict?

Originally Posted by mitsguy2001

In my case, my high school had many students who were equal except for which teacher we had for AP Bio. Myself and others who were stuck with the hard teacher did not get into Ivy League schools, while many of those who had the easier teachers did get into an Ivy League school. It also cost me at least one scholarship, since that one grade alone knocked me out of the top 5%, and this school automatically gave a full scholarship if you were in the top 5%. I've explained in the past my high school's strange ranking system. So, yes, this one grade alone did make the difference.

I think a point can be made that grade inflation is a big reason for this. With so many students getting A's, rank in class is basically determined by the "noise" (which teacher you get for AP Bio) rather than the "signal". The problem is, when you get that one teacher who doesn't practice grade inflation, you are at a huge disadvantage.

I don't doubt what you say about your high school or the teacher. I suspect most of us have had at least one, if not several, teachers who were just hard graders, regardless of what you did.

Originally Posted by mitsguy2001

I may be wrong, but I believe you implied that you are African American, and that your youngest daughter plays soccer. I'm guessing that her reach schools either needed more African American women, or needed more soccer players, while her zone schools either had enough African American women, or their women's soccer team was full. I remember in grad school meeting a woman who was at MIT because it was the only school that accepted her! The other schools she applied to were less selective than MIT, but they focused less on affirmative action than MIT did, and they probably had enough women anyway. Even if I'm wrong about you being African American and your daughter playing soccer, most likely she fit some demographic that the reach schools wanted that the zone schools already had enough of.

Sorry for the confusion. Let me clear up. I'm a middle class white guy. My dad was born a sharecropper and he picked cotton growing up. While both kids played soccer in high school, neither did for college. The oldest was offered a couple of opportunities by small LACs, but choose to focus on her degree at an R1. Did get some scholarships, but most of that was paid for by working through school and loans. Youngest did the typical 2 reach, 2 target, 2 safe schools. We were kind of surprised that he was accepted by both reach, but neither target schools.

Both picked out of state schools, even at higher cost, because the in state schools didn't have quality programs in what they wanted so there was better ROI out of state.

Originally Posted by mitsguy2001

'

I disagree. At least when I was in high school, you could take the SAT multiple times, and it was offered nearly every month. So, if you have one bad day, you can take it over. But in a difficult high school class, one bad day can really ruin your grade. Plus, even if the SAT is a bad test, the entire nation is on equal footing, since the entire nation took the same bad test. But a single teacher giving a bad test puts that teacher's students at a serious disadvantage compared to teachers who give better tests.



That was always a pet peeve of mine. At the very least, every teacher should be required to use the same system. Would it make any sense to have a football game where one team gets 6 points for a touchdown, and the other team gets only 5 points for a touchdown, and is not allowed two point conversions? Obviously not.

Yes, you can take the SAT and ACT multiple times (though not so with many state end of course exams and the like). In fact, not only can you take the SAT and ACT multiple times, your score is expected to go up the first couple of retakes due to learning the "quirks" of the test. Schools will also accept your highest score and many will use either your SAT or ACT score that gives you the best advantage. Not only that, many will superscore, combining your highest score from the separate sections to maximize your score.

To use your football example, that would be like playing the same game three times, then using the highest number of touchdowns, then the highest number of field goals, then highest number of PATs to determine the winner. Which may not match the actual winner of the individual games.

The fact schools do this should be an indicator of the quality of these tests as a predictive tool.

Originally Posted by mitsguy2001

.
Not sure what you are saying, to be honest.
.

You asked if it would do random testing to determine strength of material. I answered yes. A big part of my career was as a test engineer. There is a whole science to designing tests. Frankly if I were to design a test that looks and acts like most standardized tests, I'd be sent back to do it over. Let's look at what we talked about just above.

Scores get better with repeats. That's an indicator of either a bad test design, burn-in period, or training and user error. Those are all things that create error in a test or mask flaws under the noise they create. Training for example. My state has gone through several versions of annual tests so the school district my kids attended started teaching them how to take standardized tests in Kindergarten.

Superscoring: AKA cherry-picking data.

Originally Posted by mitsguy2001

So why can't you just say the same thing to somebody who underperformed on the SAT?

Unfortunately, I'm an engineer making less than what teachers make in my area, plus they get a pension and retiree health insurance, neither of which I get. But I realize that's unique to the area where I live, where teachers are overpaid compared to the rest of the nation and compared to other professionals.

I wouldn't say the same thing because the scope and impact is different. So much so that, as we discussed above, you can take the SAT again and superscore. You and i are approaching this from different perspectives. You are looking for ultimate fairness. I'm looking for validity and accuracy as either a measure of knowledge, a measure of ability, or as a predictive tool. If it's valid and accurate, then it is fair. If it's not valid and accurate, then fairness doesn't matter because it shouldn't be used.

You must live in Long Island. Only place I've heard that can claim teachers are paid more than engineers.

[tnff]

Originally Posted by mitsguy2001

And that reminds of another point. Often, the instructions will say to choose the "best" answer. There will be no correct answer, but the teacher will argue that you were told to choose the "best" answer, not the "correct" answer. Which answer is "best" is a judgment call, and the teacher will only accept one as being correct. What I wonder is, are these questions intended to see how students think outside the box? Or are they because the teacher wants to cover themselves when they screw up and write a question with no correct answer, and allow them to hide behind saying they said to choose the "best" answer, not the "correct" answer.

...

That's actually an argument against standardized tests since they are full of "pick the 'best' answer" or "pick the 'most correct' answer" type questions. What the heck is a "most correct" answer? It's either correct or it isn't'.

[Lillie767]

Based on all the comments were about tests and testing, it seems that the only unbiased method of choosing college students is a national lottery. All who are interested in attending a college, submit their names into a blind lottery and the winners get to go to that school.

Since no one has come up with a fair way to choose students and everyone has dissed the "non" standardized tests, why not just use chance?

Or are multi-part applications, combined with school and standardized tests a better alternative, albeit not perfect?

[mitsguy2001]

Originally Posted by tnff

I wish CD would allow answers within quotes to carry forward. Would be so much better for the conversation. Sorry if this gets hard to follow.


For most people, in the long run, it means less later on. But it is a high stakes test for getting into college and scholarships. The SAT and ACT have a disproportionate impact. So for such an impact, the key point is whether they, and similar tests, actually measure what the purport to measure and predict what the claim to predict?

Again, I disagree. When I was applying to colleges, it seemed that rank in class was the main factor that schools focused on. Maybe it has changed since then.

I don't doubt what you say about your high school or the teacher. I suspect most of us have had at least one, if not several, teachers who were just hard graders, regardless of what you did.

And students who get stuck with such teachers are at a huge disadvantage.

Sorry for the confusion. Let me clear up. I'm a middle class white guy. My dad was born a sharecropper and he picked cotton growing up.

The reference to share cropping and cotton were why I assumed you were African-American.

While both kids played soccer in high school, neither did for college. The oldest was offered a couple of opportunities by small LACs, but choose to focus on her degree at an R1. Did get some scholarships, but most of that was paid for by working through school and loans. Youngest did the typical 2 reach, 2 target, 2 safe schools. We were kind of surprised that he was accepted by both reach, but neither target schools.

Again, even if not race, gender, or sports, your son likely fit some demographic that the reach schools were looking for but his zone schools were not.

Yes, you can take the SAT and ACT multiple times (though not so with many state end of course exams and the like). In fact, not only can you take the SAT and ACT multiple times, your score is expected to go up the first couple of retakes due to learning the "quirks" of the test. Schools will also accept your highest score and many will use either your SAT or ACT score that gives you the best advantage. Not only that, many will superscore, combining your highest score from the separate sections to maximize your score.

But again, everybody has these same opportunities, whereas teacher-written exams do not give the same opportunity to every student.

To use your football example, that would be like playing the same game three times, then using the highest number of touchdowns, then the highest number of field goals, then highest number of PATs to determine the winner. Which may not match the actual winner of the individual games.

But if that is the rules that every team follows, then it is “fair”. Is it fair that an undefeated team can have one bad playoff game and be eliminated by a weaker team? Whether it’s “fair” or not, it’s the same rules that every team has to follow.

I wouldn't say the same thing because the scope and impact is different. So much so that, as we discussed above, you can take the SAT again and superscore. You and i are approaching this from different perspectives. You are looking for ultimate fairness. I'm looking for validity and accuracy as either a measure of knowledge, a measure of ability, or as a predictive tool. If it's valid and accurate, then it is fair. If it's not valid and accurate, then fairness doesn't matter because it shouldn't be used.

If my AP Bio teacher gave every student a B, and a different teacher gave every student an A, is the grade in any way a valid measure of knowledge or ability? Obviously not.

You must live in Long Island. Only place I've heard that can claim teachers are paid more than engineers.

Yes I do live on Long Island.

Originally Posted by tnff

That's actually an argument against standardized tests since they are full of "pick the 'best' answer" or "pick the 'most correct' answer" type questions. What the heck is a "most correct" answer? It's either correct or it isn't'.

But teacher-written tests tend to be the same “pick the best answer” or “pick the most correct” answer types, even more so than the standardized tests.

[phetaroi]

Originally Posted by lhpartridge

How did we measure schools before NCLB when the students typically took one test per year such as the ITBS or Stanford Achievement Test before their college entrance exams late in high school?

Basically, we didn't. Schools told us, "We're doing a great job", and sometimes that was true and sometimes it wasn't. Then along came the general public demanding proof. Testing began. Just like many other aspects of our society. We want evidence that schools are doing a good job.