So whats is really wrong with standardized tests?

[hobbsyoyo]

I'm coming at this from the angle of having to compute - by hand - multiple flyby trajectories, course change burns and deltaV calculations in a single problem (on top of other problems) in an hour and fifteen minutes.

Just completely unrealistic and so far removed from any real world situation where you would have ample time to get it right that this kind of testing is a bad measuring stick in my opinion.

I've had tests that were so crammed people couldn't finish not because they didn't know how to do it but because they couldn't get it into and out of their calculators fast enough. These tests are called 'calculator Olympics' and they seem to be pretty standard fare in my field of study.

I don't at all discredit your time management on Law School exams - to the contrary, I acknowledge that time management is an important part of test taking. I just don't feel that forcing artificial time management into a testing situation is actually relevant to any real world situations or a good yardstick of aptitude other than your aptitude at that particular style of testing, in my field at least.

I did have a couple of professors that would book rooms in the evening and give us unlimited time. They were the exception.

 

[Flying Pig]

In most lines of work, matching your level of detail to the time available is an important skill. I spent most of my life having to make plans and then explain them to small groups of people, with various amounts of time available to do so. If I've got five minutes to give a briefing, I need to have the skill of knowing which details I should keep to myself and which I need to tell the troops to make sure that everything works. I imagine similar situations exist in law: if you're a barrister, you need to know how to condense your argument so as to tell the court the essential details to carry your case without overloading them with information that they don't need. Working out what fits into which category is a learned skill.

 

[hobbsyoyo]

I can believe that. It just doesn't apply to my field.

No one expects you to crunch out new trajectories for a spacecraft on the fly because if you did that, billions of dollars (and possibly lives) get burned up.

The closest scenario I can think of even approaching the madness of our academic tests is the Apollo 13 disaster and even then a very large group of people with essentially limitless resources worked on the problem on the ground. Oh and you bet they were collaborating too, which is not how tests work.

 

[JollyRoger]

I imagine similar situations exist in law: if you're a barrister, you need to know how to condense your argument so as to tell the court the essential details to carry your case without overloading them with information that they don't need. Working out what fits into which category is a learned skill.

Exactly . . . and also, courts have crowded dockets so you are only going to get so much time before the judge cuts you off. Some judges will severely limit the time you have for voir dire, opening and closing statements, etc. Gotta keep it like a miniskirt, short enough to be interesting, but long enough to cover the essentials.

 

[Flying Pig]

I can believe that. It just doesn't apply to my field.

No one expects you to crunch out new trajectories for a spacecraft on the fly because if you did that, billions of dollars (and possibly lives) get burned up.

The closest scenario I can think of even approaching the madness of our academic tests is the Apollo 13 disaster and even then a very large group of people with essentially limitless resources worked on the problem on the ground. Oh and you bet they were collaborating too, which is not how tests work.

There is a practical need for time-limited tests - namely, that the invigilator isn't going to sit there indefinitely while the students do the exam - but it sounds like your tests tried to cram too many questions into that time.

 

[hangman]

I'm coming at this from the angle of having to compute - by hand - multiple flyby trajectories, course change burns and deltaV calculations in a single problem (on top of other problems) in an hour and fifteen minutes.

Just completely unrealistic and so far removed from any real world situation where you would have ample time to get it right that this kind of testing is a bad measuring stick in my opinion.

I've had tests that were so crammed people couldn't finish not because they didn't know how to do it but because they couldn't get it into and out of their calculators fast enough. These tests are called 'calculator Olympics' and they seem to be pretty standard fare in my field of study.

I don't at all discredit your time management on Law School exams - to the contrary, I acknowledge that time management is an important part of test taking. I just don't feel that forcing artificial time management into a testing situation is actually relevant to any real world situations or a good yardstick of aptitude other than your aptitude at that particular style of testing, in my field at least.

I did have a couple of professors that would book rooms in the evening and give us unlimited time. They were the exception.

Yeah, in atmospheric science, a couple of places are starting to experiment with more open-ended written quals. Having to solve a problem in the space of an hour simply never happens in research. The math is complex enough that it can take a week just to cover a single derivation. We had a number of take-home tests and finals.

 

[Smote]

I don't buy the entire idea behind rigid curriculums. The idea that students should know this material at this age is just going to cause problems. We ought to have teaching be as individualized as possible, with tests only for when its absolutely necessary to demonstrate a certain level of proficiency.

 

[Zack]

I don't buy the entire idea behind rigid curriculums. The idea that students should know this material at this age is just going to cause problems. We ought to have teaching be as individualized as possible, with tests only for when its absolutely necessary to demonstrate a certain level of proficiency.

Not really feasible without an absurd number of teachers.

 

[Flying Pig]

An absurd number of extremely skilled (and therefore expensive) teachers, as well. The advantage of standardised curricula is that you can check on the teachers' performance as well as that of the students - you know what a substandard teacher looks like because their students can't tick the right boxes, and if they're new or not all that Dead Poets' Society then they have a list of things that they can teach, rather than being left to make up their own list, which might be monumentally stupid.

 

[Kouvb593kdnuewnd]

Are people that put high score on such tests also the ones that tend to be best at video games or being very creative

Well standardized test is probably not the thing I can do to well however do it matter

Better too beat games on the hardest difficulty in your first playthrough if you ask me

For many others, other things are also more important then being good at something like math

But Im sure everyone can beat everyone at something

 

[uppi]

Having to solve a problem in the space of an hour simply never happens in research. The math is complex enough that it can take a week just to cover a single derivation.

Yes and no. In research you do not have a strict time limit very often. However, someone who can solve the problem in one hour will be much more successful than someone who needs two weeks for the same problem. So how fast one can solve a problem is a valid measure to test.

But there is obviously a limit: it should not become a contest who can write or push buttons on the calculator the fastest.

We had a number of take-home tests and finals.

Take-home test have the disadvantage that there is no way to verify that it was actually you doing the work and that you did not just post it on the internet and wait for an answer.

An absurd number of extremely skilled (and therefore expensive) teachers, as well. The advantage of standardised curricula is that you can check on the teachers' performance as well as that of the students - you know what a substandard teacher looks like because their students can't tick the right boxes, and if they're new or not all that Dead Poets' Society then they have a list of things that they can teach, rather than being left to make up their own list, which might be monumentally stupid.

If you use the standardized tests to evaluate the teachers, you will end up with the teachers teaching how to tick the right boxes instead of what they are supposed to teach.

 

[hangman]

Yes and no. In research you do not have a strict time limit very often. However, someone who can solve the problem in one hour will be much more successful than someone who needs two weeks for the same problem. So how fast one can solve a problem is a valid measure to test.

But there is obviously a limit: it should not become a contest who can write or push buttons on the calculator the fastest.

...which is usually what winds up happening in practice. I very often did not have time to complete problems that I had planned out. I'd have to write out my algorithm and hope for partial credit so I could get everything in before time was up.

Take-home test have the disadvantage that there is no way to verify that it was actually you doing the work and that you did not just post it on the internet and wait for an answer.

You're assuming someone online would want to answer these kinds of problems But usually, the problems are written in a way that it doesn't matter if you have to look something up online. If you can't do it, you can't do it. I've had some profs give a 24-hr time limit, which I felt was pretty reasonable, and it should preempt most online forum answers.

 

[cardgame]

The best way to game these tests is to have a high reading speed. The tests are timed. There is a lot of material to read. I always had a decent amount of time to re-check my work or to take a break in between sessions (it's very helpful to have a minute to breathe).

I scored above the nat'l average with middle-school knowledge of mathematics (read: almost none) due to the science (with lots of reading, decent mix of comprehension and actual solving) and english portions.

BTW I'm specifically talking about the ACT but I can't imagine the SAT being too different.

 

[Zack]

The science section didn't really even require any scientific knowledge, iirc it was basically just interpreting graphs.

 

[Smote]

Lessons from Finland:

http://www.cnn.com/2014/10/06/opinion/sahlberg-finland-education/

 

[Flying Pig]

If you use the standardized tests to evaluate the teachers, you will end up with the teachers teaching how to tick the right boxes instead of what they are supposed to teach.

Yes, but in a well-designed test the two are indistinguishable. If the test is a good one, the only way they can teach their pupils to tick the right boxes is to teach them to understand the questions. If the teachers themselves are gaming the system, the problem is in implementation, not principle.

 

[warpus]

I'm coming at this from the angle of having to compute - by hand - multiple flyby trajectories, course change burns and deltaV calculations in a single problem (on top of other problems) in an hour and fifteen minutes.

Just completely unrealistic and so far removed from any real world situation where you would have ample time to get it right that this kind of testing is a bad measuring stick in my opinion.

Interviews for programming positions have a reputation for being incredibly silly and useless too - not testing the programmer's problem solving skills and programming fundamentals - but rather his/her ability to memorize function names and proper syntax and spit them out on the fly - as well as the ability to solve "gotcha" style and/or clever programming riddles on the spot.

It doesn't help you figure out who might or might not be a good programmer... Neither do HR style questions.

 

[Archbob]

Lessons from Finland:

http://www.cnn.com/2014/10/06/opinion/sahlberg-finland-education/

Its also interesting to note than Japan and Korean school are the exact opposite and score better than Finland.



I think our educational system should be federally run, not state run.

 

[Smote]

Its also interesting to note than Japan and Korean school are the exact opposite and score better than Finland.



I think our educational system should be federally run, not state run.

I think what many schools have failed to do is really commit to helping students achieve a healthy outlook on life. This is what may be the single factor that I most applaud in Finland - it manages to balance academic excellence with well-being. What do you prefer - Korea's 14 hour school day, or Finland's 5 hour school day?

Many young South Koreans suffer physical symptoms of academic stress, like my brother did. In a typical case, one friend reported losing clumps of hair as she focused on her studies in high school; her hair regrew only when she entered college.

Students are also inclined to see academic performance as their only source of validation and self-worth. Among young South Koreans who confessed to feeling suicidal in 2010, an alarming 53 percent identified inadequate academic performance as the main reason for such thoughts.

http://www.nytimes.com/2014/08/02/opinion/sunday/south-koreas-education-system-hurts-students.html

Japan, on the other hand, seems to have a severe bullying problem. http://www.japantimes.co.jp/opinion/2014/10/25/editorials/school-bullying-increase/#.VLJCnY7F98E

In America, it's really difficult to teach kids a mature worldview, because parents don't want schools indoctrinating their kids with some brand of Christianity, or some philosophy that makes belief in Christianity impossible. Schools tend not to give reasons for why kids should follow rules. Rather, they focus on teaching the material, and let kids "decide for themselves" as to how to approach life.

But this decision process is not guided. Schools don't teach kids reasons to be moral, other than the idea of "that's how you would want to be treated, isn't it?" They teach kids to follow the rules, because that's what makes society work. But they don't teach any of the underlying philosophies in an objective way. Rather, they just teach you to "work hard to succeed so you can go to college and get a good job." But why should we go to college? Why should we get a good job? Why should we want to make a lot of money? So you can do whatever you want. So you can have options. (Sarcasm)

But ultimately, there's no good discussion of the underlying reason to do anything (Hint: there isn't any.) All there is is what we make of it, and schools do a dreadful job of teaching this. They might say it, occasionally. But they don't explain it. They don't really dig into philosophy. The closest thing to this is discussion of literature in higher level English classes that deal with philosophy, but it is too little too late, and only brushes the surface. They ought to teach students outlooks on life by various famous philosophers. They ought to teach students our best knowledge of human psychology. They ought to teach students about the "ego" - the drive for social status which causes lots of bullying and social conflict. They ought to teach students stress-coping mechanisms like mindfulness. This could all be done with some sort of Philosophy/Psychology course in early Middle School (which is when bullying comes most into play).

These are all "sticky topics" that teachers try their utmost to avoid, and yet, they are some of the most important topics for any student. They provide the best basis for internal motivation and later success. Why is it okay for us to cop out on teaching our kids our best guesses to their biggest questions? Simply, because it's easier? Or because it's not right for schools to teach students how to live their life? I'd argue that schools ought to teach students an availability of options of ways to live their life, rather than saying the same old "Follow the rules, work hard, and get a job" without explanation as to why this is actually important.

So I could support a federal plan, if it would include this type of instruction for students. But I know it wouldn't. Teaching kids philosophy and psychology would be spun as "indoctrination" by conservatives. The Christian Right has a hard enough time with Evolution. They'd have a heart attack if people were taught philosophy.

 

[Samson]

En masse, it'd be highly correlated that those who score high compared to others in non multiple choice would score well compared to others in multiple choice (sat math also has small section of "free answer" where you report up to 8 numeric characters long, bubbled in. So a step up from only multiple choice)

In a well designed test, the format does not matter

The sat is curved anyways

I do not buy this argument. The image below is a pretty highly correlated (and completely unrelated) graph (R^2 of 0.67 I think). One could think of every dot in the converse quadrants (those high in one and low in the other) as someone a bit thick who has loads of money spent on their education to little result, or someone clever who does not get the education to fully exploit their potential. Given this, does not not make sense for the state to spend a few minutes of someones time to mark a test where marks can be given for working?

Yes, but in a well-designed test the two are indistinguishable. If the test is a good one, the only way they can teach their pupils to tick the right boxes is to teach them to understand the questions. If the teachers themselves are gaming the system, the problem is in implementation, not principle.

Unless the principle is to rate teachers on their pupils results in said tests.

Gotta keep it like a miniskirt, short enough to be interesting, but long enough to cover the essentials.

I like that simile and want to use it more in conversation