IQ

[Reach]

Re: IQ

Your test is already worthless on these grounds alone.

Ok. Provide evidence to back this up?

Whether you punish people for wrong answers, like I think the SAT used to do, or don't, doesn't make any statistical difference. I actually prefer not punishing people for wrong answers, because then people don't worry and not answer even though they are right.

It just gets incorporated into the standardization. It means the average score on the test is slightly higher than it would be if you penalized wrong answers, but the curve ends up being the same (because everyone's score goes up).

[xingyguy]

Re: IQ

Mine is 1337.

Darn it! You stole my line.

[travman301]

Re: IQ

Reach, can you explain the scoring on the second test you posted? I didn't really understand it.

[Reach]

Re: IQ

Reach, can you explain the scoring on the second test you posted? I didn't really understand it.

Oh, sorry.

http://www.iqcomparisonsite.com/IQtable.aspx

You need to click on the icon at the end of the test that matches your raw score with a percentile. http://www.cerebrals.org/tests/tri/Classeur1.pdf

Look for your percentile score (e.g. 98.8) and match it with the IQ. Use the column on the left.

I should note that this test is suited more for individuals that scored very high on the first test I posted. I have no idea what you scored, so I figure I should point that out.

If I were to make an educated guess, I'd say your score was 940+10.

I scored 930 after approximately 2 hours, IIRC.

Interestingly, on very difficult tests like this, how long you take doesn't seem to matter (assuming you continue until you start having problems). It seems that once you reach your capacity cap, you start to get too many wrong for it to matter how long you spend.

[travman301]

Re: IQ

Ah, thanks. And I got a 700 on the that one.

I've never taken a real IQ test, but all of the "nonofficial" ones have ranged from 120-135.

And personally I enjoyed the 2nd one much more than the first one. The patterns were much more interesting to me.

[devonin]

Re: IQ

Ok. Provide evidence to back this up?

Scoring full points for a lucky guess means that the results have a much larger potential deviation based around something you can't possibly measure or predict for: The degree to which individuals taking the test happen to be lucky or not.

The only way to account for that is to somehow try and deduce roughly how many answers people will guess at, and then your results are +/- MORE points, so you fall into a range bigger than you ought to if the test forced a greater degree of precision.

I've seen tests (thinking particularly of highschool math contest tests) where you got full points for a correct answer, no points for an incorrect answer, and a fixed smaller number of points for the first X questions you left blank.

So you got full points for knowing the answer, and rather than reward you for potential lucky guessing, gave the same degree of wriggle room for "I don't know the answer to this" to everybody. Since there was a fixed number of potential points you -could- get by admitting ignorance, it was easier to measure for and correct than trying to account for who might be how lucky just guessing.

[OrganisM]

Re: IQ

All this intellectual stimulation is giving me a mindgasm.

[bluguerrilla]

Re: IQ

All this mental-circlejerk is giving me a mindgasm.

Even though I haven't contributed anything I have found all this interesting.

I like the GRE -> IQ estimator, I have like a 145 IQ from that and I'm sure it's more accurate than taking any other test right?


I have to sit down and take one of these because they're always fun.

[devonin]

Re: IQ

I also question just what useful conclusion you're supposed to draw from your ability to analyze and manipulate patterns, when that ability is the only thing being measured.

[Izzy]

Re: IQ

Fun stuff. To lazy to take a test. I'm usually around average on everything so i'll just say 100.

[devonin]

Re: IQ

Random clicking on the second test scored me an IQ of 109. Sweet.

[Izzy]

Re: IQ

Got a 97 by clicking all C's on reaches first link.

[Reach]

Re: IQ

While you bring up some valid things Devonin, some things need to be cleared up.

Random clicking on the second test scored me an IQ of 109. Sweet.

The second test is not designed to measure average IQ, let alone low IQ. It's a high range test.

Take the first test. Even the first test is not designed to measure low IQ. It's a fair test for FFR members, given anyone that's going to take it is probably reasonably intelligent. However, the test is still not designed to measure IQ below about 85.

The Mensa.dk test is based on Ravens APM, or advanced progressive matrices. This means it's a harder version of the original, designed to have higher accuracy in the higher IQ range. The Progressive Matrices test is much better at measuring low IQ.

I also question just what useful conclusion you're supposed to draw from your ability to analyze and manipulate patterns, when that ability is the only thing being measured.

I've been through this before, so I won't get into detail, but to summarize:

1. you are using a variety of intellectual abilities in order to solve these problems. Even verbal skills, in many cases.

2. Matrices are known to be highly g-loaded. This means performance on them correlates very highly with performance on all other mental tasks. If you are good at them, you're probably also good at general knowledge questions, vocabulary, analogies, processing speed, sequences, memory and working memory, etc.

This is because the g-loading is tapping into your general intelligence factor, which influences performance on any task.


That's not to say there aren't downsides to a test like this. There are a few. Tests like the WAIS (probably the most commonly administered in america) test multiple things, and can give a better picture of your intelligence. Your WAIS score and score on this test won't differ much though. Also, the reason tests like this are used is because they're culturally unbiased and can be given to just about anyone that is aware of what you're supposed to do.

Scoring full points for a lucky guess means that the results have a much larger potential deviation based around something you can't possibly measure or predict for: The degree to which individuals taking the test happen to be lucky or not.

The only way to account for that is to somehow try and deduce roughly how many answers people will guess at, and then your results are +/- MORE points, so you fall into a range bigger than you ought to if the test forced a greater degree of precision.

I've seen tests (thinking particularly of highschool math contest tests) where you got full points for a correct answer, no points for an incorrect answer, and a fixed smaller number of points for the first X questions you left blank.

So you got full points for knowing the answer, and rather than reward you for potential lucky guessing, gave the same degree of wriggle room for "I don't know the answer to this" to everybody. Since there was a fixed number of potential points you -could- get by admitting ignorance, it was easier to measure for and correct than trying to account for who might be how lucky just guessing.

Sure, scoring a test like this increases the error of measurement.

It's just that, proper tests actually account for this through what is known as the standard error. You normally will receive your score as a range, e.g. 120-136, rather than an exact score. This accounts for things such as luck.

This particular test doesn't list the standard error, which is a downside of the test. It's probably about 16 points though.

[BARMK_RULEZ]

Re: IQ

doesn't age not have an effect on IQ?

so yeah

So you saying that my IQ will stay the same forever? I was born with an 137 IQ? I will die with a 137 IQ????

[robertsona]

Re: IQ

Correct.