|
On June 14 2012 04:16 Cirqueenflex wrote:
For me being smart is trying to learn out of everything you do, finding mistakes by connecting patterns and constantly improving. This is like infinite possibilities and can never be truly measured by any kind of stupid IQ test where you suddenly seem more intelligent if you had the luck to already have heard such a question and its solution already.
You probably don't know much about IQ tests. It's a wide array of tasks that you have to complete, and the result is calculated considering the time it took you and how good were your answers. From what I remember there were logic, calculus, some general knowledge, vocabulary and even some moral issues (like, if you find a wallet with the owner ID in it, what do you do). Of course factors such as mood, weariness and maybe even luck to an extent can change the outcome.
The trick, which is what I think most people don't know, is that the result isn't strictly how intelligent you are. It's more like trying to know what time is it by watching the shade of a stick in the ground. Indirect measurement, and with only 70% accuracy if I remember correctly. Still the best way we have to measure it so far, as far as I know.
So, being intelligent isn't how fast/good you perform those tasks, but being intelligent helps in doing them better/faster.
What you describe in the first sentence I quoted would be the smart way to use your intelligence, not being intelligent.
Also, please stop saying that IQ tests are stupid. Some people have suffered being "gifted" with high IQ, and knowing why can help them.
|
5 cents
Also I feel like the whole article is based on a so called "cognitive bias". Higher education, even scientific one does not equal smarter people.
|
Gosh I'm so bad at maths, I reread the proofs on the first page for the past 10 minutes at least before I could understand how it's 5c haha omg... 
in my defence it's 6am and ive been awake for over 3 hours and I feel like absolute shit
|
5 cents.. and I'm surprised college grads missed the lily pad question, it seems rather simple.. if something doubles in size every day it was obviously half the size the day before.. so it filled half the lake one day before..
Are you saying having a better education makes you more prone to miss these types of questions? I find it hard to belive if you gave the same questions to a group of people with no college degree a small percentage of them would miss the question..
|
5 cent ;p
But when you say "answer as as spontaneously as possible."
This means the whole research is about cases where you have to think quickly?
Because if you take like 30 seconds to think about it, it's just a simple question.
And if you know these questions are about bias, you would take that extra second to think about the answer.
The one about the lake is also pretty easy too, if you think about it logically.
|
"The Smarter People Get, The More Prone They Are to Cognitive Bias" isn't what the article says. More like the press getting their hands on an article like this.
When faced with the problems in the study, being 'more intelligent' did not relate to getting more items wrong or right for cognitive bias questions. Being smart did however relate to an overestimation of a persons own ability to not be fooled compared to other people (bias blind spot). Which I don't really think is a "cognitive bias" in itself, since it's reasonable to think that someone who is smarter would be better at not being tricked. At least it was reasonable until I read the article, but turns out it just doesn't matter.
Also that overestimation was weak and like the title says "Cognitive Sophistication Does Not Attenuate the Bias Blind Spot" which is as they say at least the 'conservative' way of interpreting the results.
Anyway, I think the OP is still interesting just wanted to point it out. I think the fact that being aware of bias or being more open-minded (at least based on self-report), or scoring higher on SAT doesn't matter is pretty interesting. At the same time it's not impossible to get rid of cognitive bias. I and I'm sure many others looked an extra time or two at the questions because I knew there would be something strange about them and that the answer wouldn't be so obvious.
|
Ball is 5 cents because bat is $1 more and total price is $1.1. So 1.1= X+X+1.0. The initial reaction is 10 cents but then the bat would only be 90 cents more than the ball.
|
On June 13 2012 13:25 Infernal_dream wrote:Show nested quote +On June 13 2012 13:19 Galaxy1again wrote:
Asking ambiguous questions with poor punctuation doesn't make intelligent people unintelligent. The punctuation is perfectly fine. They are short sentences, but that doesn't make them incorrect.
Futhermore, this question is about as specific as a question can be.
|
Zounds, it seems almost nobody is still getting what this paper was about or have picked it up from reading other responses. To reiterate and make it very clear:
If you got the tricky-ish little math questions (referred to as the CRT) correct, then you are smarter.
If you got the tricky-ish little math questions wrong, then you are stupider.
That's not the issue. That was never the issue, and this result is consistent with previous studies. Better students and people with better cognitive performance on other measures get the CRT questions correct more often.
The authors' result is that supposedly "smarter" people underestimated their own bias on this OTHER GROUP of "bias questions."
I can see how at first everyone thought the discussion was different, both due to the OP, and apparently journalists' presentation of this. Since then it's been clarified but people keep missing it, there's a misunderstanding of what the paper was actually meant to be about. You're not "supposed" to get the CRT questions wrong. It's good to get them right. The real point the authors attempt to make is on this other group of bias questions. Now the brilliant part is, ok, now you're thinking the comparison is between the students' results on the CRT, and results on the bias questions. The authors don't report that data either. They drew the bias conclusion based on calculating correlations to SAT scores. Understand what the paper is actually about, though I'd say it's pretty rubbish on those grounds too, and not sorry for anyone who still didn't find access because then you won't have wasted your time.
+ Show Spoiler +
So in essence, for another summary of what the paper presents, what happened was the authors conclude that better self-reported test scores do not correlate to a lack of bias on the bias questions. The particular point is on the bias-blind spot, or how people "underestimate their own bias" in a meta-sense as a few others have mentioned. None of that is news or interesting, the shock value of this paper is supposedly in how this relates to people being "smart" aka cognitively sophisticated. For whatever reason, cough, they decided to only use the self-reported test scores from the students to decide which students were smart. Also that comes with such deliciously weasely statements like, self-reported test scores usually correlate to students' actual test scores pretty well.
And ffs, the authors don't even address the use of total SAT scores instead of just SAT math which makes way more sense. Anybody with half a brain in these sorts of social sciences in American universities should know that SAT math and verbal scores do not correlate perfectly, especially due to foreign/ESL students who can have much higher math scores. Again, it's possible that the vast incompetence of all the students involved was such that it turns out this didn't matter and the authors just didn't mention it, but there was no explicit reason for why they don't simply take the much more intuitive SAT math scores only to compare to their other metrics)
Remember that almost all the student participants in the study generally have terrible scores as well. So the conclusion is reasonably questionable from the very nature of this study right away. It would be like, because researchers studied 500 bronze and silver level players and concluded that greater skill had no correlation to reduced bias, therefore more skilled Starcraft players are biased, therefore professional players are more biased. And then the responses in the thread are everyone saying "lol I can split my workers/solve this math problem, silly bronze level noobs," but the point was the set of "bias questions" in the first place, and professional players weren't even tested.
What is frustrating, to elaborate on from before and especially given the thread discussion, is that the authors do not actually report the correlation between performance on the CRT - those tricky little math questions - and performance on the bias questions. They decided not to publish that. In essence, the CRT results are rather irrelevant to most of the paper.
At that, I do think it is rather disturbing that the paper in its current state is supposedly being published in a well-reputed journal within the psychology/sociology field. Maybe this is not so different from other papers in the same social science domains to which I am also unfamiliar, but that did surprise me. If this lack of rigor happened in another field, say climate science, it would be the sort of thing where Fox News would be running 24/7 headlines on "Climate Scientists Refuse to Publish Data, Liberal Conspiracy Confirmed." One of the so-called biggest "scandals" in all the climate science controversy is debating the interpretation of ice core data going back to thousands of years ago and that's orders of magnitude less suspicious than this paper.
This is not the 1950s, where people had to sit around with slide rules or whatnot. The chance they wrote any of their own code or anything like that is vanishingly small. There's no possible way for one to believe the authors could not actually calculate these other correlation coefficients, that they literally couldn't press enter and print a table from a computer which is almost certainly all they had to do. Yet for some reason these correlations are omitted from publication.
It is possible to give the authors the benefit of the doubt and suspect no intentional wrongdoing, but still, there's either sheer idiocy at play or a clear revelation of inherent flaws in the study methodology and its sample. No possible reason not to publish the other tables of data that I can think of.
So for one last time, whether you got the tricky little math questions - the CRT - is not the issue here. The paper itself fails to present the necessary data on the CRT results compared to the results on the bias measures. A better title which reflects some of the actual results on the bias measures, might perhaps be:
"Female JMU students are imbeciles who think Africa has 10 countries, trees are 1000 feet tall."
|
|
|
This study definitely makes a whole lot of sense. People who are more intelligent are better at making their brains follow certain patterns to solve problems, as in the long run it saves time and cognition power for more intricate tasks. This study isn't trying to say 'intelligent people are unintelligent' or whatever it is you all are claiming. It just exposes some of the shortcuts the human mind can take in approaching problem solving. Very interesting.
|
Wow I'm really mean, my bad
|
my slightly tipsy brain screamed 10 cents followed within seconds by 'NOO it's a trap'. I assume that afther a few seconds of thought everybody will get to the right answer. Otherwise i can't really get interested by this research, it rieks of Diederik Stapelesc assumptions.
|
One problem I've always had with tests like these is that they are done under the premise of trying to trick people. And worse, people usually know that.
Which is why, if you ask an ambiguous mathematical question that uses exact semantics to try and fool people (who are already overthinking the question), you will get results you ask for.
|
On June 14 2012 07:06 Crazy Eddie wrote:
At that, I do think it is rather disturbing that the paper in its current state is supposedly being published in a well-reputed journal within the psychology/sociology field. Maybe this is not so different from other papers in the same social science domains to which I am also unfamiliar, but that did surprise me. If this lack of rigor happened in another field, say climate science, it would be the sort of thing where Fox News would be running 24/7 headlines on "Climate Scientists Refuse to Publish Data, Liberal Conspiracy Confirmed." One of the so-called biggest "scandals" in all the climate science controversy is debating the interpretation of ice core data going back to thousands of years ago and that's orders of magnitude less suspicious than this paper.
This is not the 1950s, where people had to sit around with slide rules or whatnot. The chance they wrote any of their own code or anything like that is vanishingly small. There's no possible way for one to believe the authors could not actually calculate these other correlation coefficients, that they literally couldn't press enter and print a table from a computer which is almost certainly all they had to do. Yet for some reason these correlations are omitted from publication.
The numbers are there. I can't say I'm used to these kind of results but for example Table 3 for SATs. For example:
"The next row of the table indicates that the slope of the response as a function of SAT score was steeper for the negative outcome condition (.254) than for the positive outcome condition (.072). The direction of the effect is indicating that participants of higher SAT displayed ...."
Regarding some of the other stuff criticism, you kind of responded to it yourself in the beginning. The article is about "Bias Blind Spot" and "Cognitive Sophistication". It's not about how well people perform on the cognitive bias tests, that's just the part that the press liked. Maybe really high math SAT people do well on this stuff, maybe they don't, this article isn't about it.
If you don't find it interesting at all that is fine but not really a need to be so hostile about it. It's not like you would have made a better study and there are plenty of good points like including a sample of non-college students.
|
The numbers are not there for the CRT scores of the JMU participants. (Nor for the NFC or other cognitive measure they mentioned, I believe)
It's remarkably lazy at best to only check the correlation to SAT scores and the self-responses on bias blind spot. I don't see how they get away with not even publishing the other data, even an appendix table, unless I'm missing something or the version of the paper I had access to wasn't complete. But I don't see it there, which is silly considering the "study 2" only tables CRT, not SAT scores, and the reader then can't even compare that to the first study.
It's quite possible that due to the low number and skewed sample size of students with high scores on the CRT, it wasn't really possible to calculate the correlation, but that alone highlights a huge flaw in this study and would also be worth a mention.
|
But if your asked to answer as spontaneously as possible, wouldn't it be natural to answer with something wrong, unless you've answered a lot of similar questions in your life? I mean, doesn't being smart mean you'd find your error sooner, so if you think about your answer, you'd find out it's wrong, but the question specifically tells you not to think about it, right?
edit: my point being, how could anyone belief being smart (if we presuppose that degrees mean you smart, which i can tell you is idiotic at least in austria) means having better intuition?
|
On June 14 2012 08:41 blarkh wrote:
But if your asked to answer as spontaneously as possible, wouldn't it be natural to answer with something wrong, unless you've answered a lot of similar questions in your life? I mean, doesn't being smart mean you'd find your error sooner, so if you think about your answer, you'd find out it's wrong, but the question specifically tells you not to think about it, right?
edit: my point being, how could anyone belief being smart (if we presuppose that degrees mean you smart, which i can tell you is idiotic at least in austria) means having better intuition?
This is exactly what I was thinking. If you ask someone that is good at math the question and tell them to write it down, I bet they would score far better than if you told them to verbally tell the answer. However, if you ask someone that is bad at math the same thing, they would probably get the answer wrong both times. This is all conjecture, but I think it really means what you said, that smart people don't give great spontaneous answers (which I don't think many people would, in this scenario).
This is all assuming that the questions were given verbally, or given in written form but with some form of added urgency.
|
On June 14 2012 08:32 Crazy Eddie wrote:
The numbers are not there for the CRT scores of the JMU participants. (Nor for the NFC or other cognitive measure they mentioned, I believe)
It's remarkably lazy at best to only check the correlation to SAT scores and the self-responses on bias blind spot. I don't see how they get away with not even publishing the other data, even an appendix table, unless I'm missing something or the version of the paper I had access to wasn't complete. But I don't see it there, which is silly considering the "study 2" only tables CRT, not SAT scores, and the reader then can't even compare that to the first study.
It's quite possible that due to the low number and skewed sample size of students with high scores on the CRT, it wasn't really possible to calculate the correlation, but that alone highlights a huge flaw in this study and would also be worth a mention.
Study 2 didn't use SATs.
The study is NOT about how "intelligent" people perform on cognitive bias tests. You yourself pointed this out. Yet you continue to criticise the paper based on the assumption that it's about something it is not.
The exact scores for everything just isn't relevant to the focus of the paper. They present enough data to suggest that there isn't a difference, and if someone else wants to pick it up from there and make a study about how "Cognitive ability and performance on cognitive bias testing" then that study will report all the data. Because then the data is not OT. In this study it is a side note as a test to see if it is a "real" blind spot or not.
|
5 cents.
I hesitated though, instinctively I went 10 cents because the question is like a riddle. 0.0
|
|
|
|
|
|
EPT
