Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

An alarm about the influence of standardized tests on American society

scientificamerican.com

Post reply on HN
Print view View original

91–100 of 198 posts

Re: An alarm about the influence of standardized tests on American society

#91

Post by kyleschiller »

cperciva wrote: post #74

Standardized tests are a lousy way of deciding things. But at least they're better than non-standardized tests. At my university (SFU, in Canada) admission to most faculties is based on high school grades. That's grades awarded by high schools -- we used to have province-wide grade 12 exams, but for almost all subjects those were abolished a few years ago. It's widely recognized at the post-secondary level that a 90%…

Yeah, I don't think Sternberg is advocating for ending standardized testing:

"My colleagues and I developed assessments for creativity, common sense and wisdom... we doubled prediction [accuracy] for academic performance, and with Kaleidoscope we could predict the quality of extracurricular performance, which the SAT doesn’t do."

It sounds like he just wants to reform them, and has some strong initial results to suggest that we could do much better.

Re: An alarm about the influence of standardized tests on American society

#92

Post by bluGill »

shmageggy wrote: post #66

> I think it’s hurting everything. We get scientists who are very good forward incrementers—they are good at doing the next step but they are not the people who change the field. They are not redirectors or reinitiators, who start a field over. And those are the people we need. I doubt the problem is with the scientists. I know and work with lots of really smart and creative people in an academic environment, and I'm…

Most of the need in science is in advancing another step. For every major advancement we need tens of thousands of small advancements to refine and work out all the details. Physics was 99% solved in 1880 with just the work of finishing out the constants. In that 1% we were missing was relativity and quantum mechanics, and we know that there is something more but not what. However as big as those two are they are still Yes there are areas of complete unknown in science (all over). However they are complete unknowns: you can spend a lifetime in false starts.

In Computer science the P=NP problem has been an important unknown for years and many have spent a large part of their lifetime on it and we still are not close to a solution. This is a known unknown, must of what remains is not even that well defined, just a case where "something funny" happens and we have no idea why and often can't figure out how to replicate it on demand.

Re: An alarm about the influence of standardized tests on American society

#93

Post by kyleschiller »

germinalphrase wrote: post #90

We need reliable ways to measure student achievement, knowledge and skill level. If we're not satisfied with the cost and negative externalities of standardized testing, then IMO we should be seeking ways to make the assessments that teachers are using in their lessons more accurate and trustworthy. We're already paying people to do the work, but we just don't trust the results. That's the problem - so let's fix it.

Yeah, isn't this exactly what Sternberg is saying? It sounds like he's done a lot of work precisely on improving the assessments.

Re: An alarm about the influence of standardized tests on American society

#94

Post by cperciva »

chrisseaton wrote: post #86
cperciva wrote: post #74

Standardized tests are a lousy way of deciding things. But at least they're better than non-standardized tests. At my university (SFU, in Canada) admission to most faculties is based on high school grades. That's grades awarded by high schools -- we used to have province-wide grade 12 exams, but for almost all subjects those were abolished a few years ago. It's widely recognized at the post-secondary level that a 90%…

> At my university (SFU, in Canada) admission to most faculties is based on high school grades Do you not interview the applicants though? I think the best thing to do is to set a basic requirement for grades from their highschool, then get people in and ask them to take your own tests and give them an interview with one of the teaching staff to make your own judgement about whether they're suitable or not. Accepting…

We're a public university admitting over 7,000 students per year. There simply aren't the resources to interview students... not to mention the fact that SFU competes with other institutions and would likely lose a large number of students if we required them to come in for interviews.

Re: An alarm about the influence of standardized tests on American society

#95

Post by barrkel »

chrisseaton wrote: post #86
cperciva wrote: post #74

Standardized tests are a lousy way of deciding things. But at least they're better than non-standardized tests. At my university (SFU, in Canada) admission to most faculties is based on high school grades. That's grades awarded by high schools -- we used to have province-wide grade 12 exams, but for almost all subjects those were abolished a few years ago. It's widely recognized at the post-secondary level that a 90%…

> At my university (SFU, in Canada) admission to most faculties is based on high school grades Do you not interview the applicants though? I think the best thing to do is to set a basic requirement for grades from their highschool, then get people in and ask them to take your own tests and give them an interview with one of the teaching staff to make your own judgement about whether they're suitable or not. Accepting…

As we all know, people never judge based on looks or accent or verbal word choice. We select for talent even if it's from someone with a different culture and social mileau to us, right?

Re: An alarm about the influence of standardized tests on American society

#96

Post by pzh »

cheetos wrote: post #84
cperciva wrote: post #74

Standardized tests are a lousy way of deciding things. But at least they're better than non-standardized tests. At my university (SFU, in Canada) admission to most faculties is based on high school grades. That's grades awarded by high schools -- we used to have province-wide grade 12 exams, but for almost all subjects those were abolished a few years ago. It's widely recognized at the post-secondary level that a 90%…

I knew a classmate who openly admitted to putting made up volunteer activities on her college applications, arguing there was no chance the school would fact check her. She was right!

The problem with these is that they couldn't have been significant enough to make a dent in the outcome. Pretty much all significant achievements are fact-checkable. E.g. you may falsely claim that you were doing karate or basketball, but if you did win any trophies, then most selective schools wouldn't care about you.

Re: An alarm about the influence of standardized tests on American society

#97

Post by petra »

nostrademons wrote: post #56
neogodless wrote: post #49

Earlier quoted context omitted.

I see what you're saying about the virtues being "different based on the situation and/or other variables." However, that doesn't mean you can't "improve upon" the ability to make virtuous choices. In your dinner example, if someone has only math and finance education, they might simply say that fast food is the best choice for dinner. But if they get some nutrition education, suddenly they're wondering if maybe some…

Sure, and this type of moral education is actually done in schools. I grew up in the 80s and 90s, and we had a unit on water conservation; a unit on energy efficiency; a play we wrote about rainforest destruction, endangered species, and global warming; recurrent reminders to recycle; programs for anti-drug and anti-smoking initiatives; sex ed; and pervasive lessons on respecting each other and authority. I haven't h…

That's a good analysis , particular the part about defectors .

And under that analysis, it seems that wisdom education , even of the deeper style that Buddhists or Christians do , will likely to fail, because a single defectors amplified by tech could do so much damage.

So are there any solutions to that?

Re: An alarm about the influence of standardized tests on American society

#98

Post by pnathan »

triangleman wrote: post #5
pnathan wrote: post #2

The ancient Western word for this is virtue : to cultivate virtues as part of an education. It's not a novel concept, but it rather got shot in the head in the postmodern revolution in education, and we're a lesser society for it.

Do you have links to old educational material where this sort of thing was part and parcel of the pedagogy? I am interested in getting into homeschooling, especially focusing on character development. I suppose an emphasis on the humanities would be a good first step, but I don't know where to find the "leading lights" in this kind of old-school education.

The keyword is "Classical" education when you're looking for homeschooling materials.

Re: An alarm about the influence of standardized tests on American society

#99

Post by dragonwriter »

grandalf wrote: post #59

While standardized tests are criticized for rewarding general intelligence at the expense of other strengths, it seems that college admissions has been steadily devaluing general intelligence over the years. Correspondingly, the alphabet tests increasingly reward rote memorization, diligence, and repeated practice at the expense of general intelligence. When I was in college the most disturbing trend was students who…

> When I was in college the most disturbing trend was students who expected all tests to be things that they could study for using the technique of rote memorization and diligence. To a large extent, pre-law and pre-med curricula offer this

I can't speak to pre-med, but IME pre-law curriculum tends (unsurprisingly like law school itself) to include courses for which rote memorization and diligence is necessary but not sufficient to exam success; specifically, where there is a mass of detailed material (rules and their context, mostly) which must be memorized fairly exactly, but which must be applied to novel circumstances in free response essays in exams.

Re: An alarm about the influence of standardized tests on American society

#100

Post by chrisseaton »

barrkel wrote: post #95
chrisseaton wrote: post #86

Earlier quoted context omitted.

> At my university (SFU, in Canada) admission to most faculties is based on high school grades Do you not interview the applicants though? I think the best thing to do is to set a basic requirement for grades from their highschool, then get people in and ask them to take your own tests and give them an interview with one of the teaching staff to make your own judgement about whether they're suitable or not. Accepting…

As we all know, people never judge based on looks or accent or verbal word choice. We select for talent even if it's from someone with a different culture and social mileau to us, right?

Well there's no need to be so sarcastic and snarky.

I did say some universities administer their own tests, so that's as objective as a test normally can be (not perfect), and probably more objective than one administered across different schools as people have already said.

I don't think it would be unreasonable to combine that with an interview. Teaching is a personal thing - academics will want to check that they can teach you and you're able to learn from them. The best university interviews I did were like a short teaching session and I came away with an opinion about who I wanted to learn from as much as I presume they had an opinion about who they wanted to spend their time teaching.

Post reply on HN