So...

I just wanted to scrutinize the earlier claim about IQ testing and what the results warranted. Someone claimed that doing well on an IQ test only showed that that person was good at taking IQ tests, but it does not necessarily show any measure of intelligence...

Am I the only one here that finds that objectionable?

It works from the presupposition that intelligence isn't needed to do well on an intelligence test. It also works from the presupposition that one may be intelligent and yet not do well on one.

Our disagreement seems to be based upon your more narrow definition of the word 'to measure', You seem to apply it as something that only is about determining quantity. Perhaps if you reconsider my statements under this definition of measure : the act or process of ascertaining the extent, dimensions, or quantity of something;
In case of determining wether an 's' or a 'z' is spelled, you made a measurement of what you saw. Otherwise, how could you tell the two apart? — Tomseltje

This, I must wholeheartedly agree.

I refer you to my final reply to FLUX23 - I believe you must have a very specific theory of mind-brain identity involving a very specific definition of "measurement" and probably some representational view of perception such that merely in seeing something, a measurement is always being made. Under any run-of-the-mill notion of "measurement" seeing and "s" on a page and measuring an "s" on a page are entirely different kinds of activities. — MetaphysicsNow

Well you didn't mention anything about my reply after your final reply, which I assume you just simply missed.

However, the measurement thing is a well established theory. Your brain is basically a computer with machine learning capability ("machine learning" is a term btw), although not as precise. It is probably your definition of "measurement" which is too specialized and specific compared to my definition of "measurement" which is more general (and is more often used).

Human brain is quite dynamic and smart in a sense. The fact that you don't remember every single details of what you see, for example now, is because your brain automatically filters out unnecessary information the moment you look at something, because when you look at something you have an objective to look at that certain "something". A great deal of measurement and processing is going on in your brain. Interestingly most of the time it is done instantaneously and unconsciously.

For example, because of scotoma, certain detail of your vision is not available. However, your brain automatically processes the blind spot with surrounding details. This is pure measurement and is a fascinating fact.

Sure it works. It works for bodylenght, so why not for intelligence. If we want to determine whether someone is short or tall, we compare them to the average height. Next to this we can express their height in cm or inches, the latter doesn't tell whether someone is tall or short without a known average.

In case of children we even correct the measured lenght for age, same as with iq tests. Why assume it won't work if the same appraoch clearly works for other things we measure? — Tomseltje

So how about if we measure weight with a thermometer? (We'll just call it a "weight-measuring device"... for good measure.) We are measuring something, we can do comparisons, calculate average, etc. We should be good, right?

Where am I going wrong with this? Your entire argument is that intelligence is adequately measured by intelligence tests because "intelligence tests" measure "intelligence" - what else could they be doing? Boom, done!

Don't you see how empty and useless such talk is? Look, you can't contribute meaningfully to a conversation about IQ tests if you don't want to get into the substance of the matter. What is intelligence? Is it something that can be measured on a scale? How can it be measured? Are existing tests adequate for the purpose? And what are such tests good for, anyway? These questions cannot be answered with wordplay alone.

However, you confused "accuracy" with "precision". — FLUX23

You are probably right on this part, however, that doesn't mean it's an indication of my level of understanding statistics. More likely it's just a translation error, since I'm not a native english speaker.
My mathematical training was in my native language. So thanks for pointing it out.

My apologies for causing confusion on this, I didn't think the difference was that relevant to the subject in this case. Since on a single measurement the precision influences the accuracy. Wich applies in this subject, since generally the individuals only get tested once for iq.

Your entire argument is that intelligence is adequately measured by intelligence tests — SophistiCat

Nonsense, in order to consider something to be adequate, you will have to state your reference. Adequate for what?
I don't recall mentioning the word adequately. Nor did I mention an application of what an iq test can be used for. My argument is that its the best way to measure it that is available, not that it can't be improved. Nor that it's adequate to substanciate conclusion x. Whatever x you may think of. Not to say that there are no conclusions to be derrived from iq tests, I just didn't make any claims about them here.

My final reply to you was that you had a very specific and itself philosophically contestable theory of mind, and that discussing that was for another thread (of which there are plenty, by the way) not this one.

See my example of a violin player who at the beginning has to concentrate very hard on the exact positioning of fingers on the fingerboard, but who - when fully proficient - no longer needs to concentrate on the exact positioning of his or her fingers, they just hit the right spot. — MetaphysicsNow

That is what we call muscle memory, wich is separate from what we attempt to measure in iq tests. The measuring is only required when training muscle memory (when the violin player still has to look at his hands).

Scientific theories are supported by empirical data. You have not data (I suppose) of your claim. — Belter

wich claim I made are you referring to? At best you could make a case that I didn't present the data, you are merely assuming I don't have it since I didn't present it yet. Since I don't know wich claim I made you wan't me to defend, I don't know wich data you are asking for.

It is possible, but again you do not show any evidence. You should cite any actual IQ tests — Belter

Why do you want me to cite an iq test as evidence? Even if I could post the symbols used in iq tests here (no idea if that is even possible here) I don't see how that is any evidence for my statements. What I'm talking about are statistical results of great numbers of filled in iq tests. Now I don't have those piles of filled in iq tests, I have to rely on the scientists who do, and published their research. I could post some links to their papers if that satisfies you.

OK, but the general point holds: activities required to gain a skill are not required to maintain a skill. Recognising an "s" is a skill, perhaps measuring is required to gain that skill (although I am still unconvinced of that, you seem to have a model of human cognition that is generally contestable), but even so it does not follow from that that measuring in any way shape or form is required to maintain it.

If the IQ scores are always lower (or higher) than the perfect test, and the values vary greatly between individuals compared to the perfect test, then the IQ test is both "inaccurate" AND have "deviation". — FLUX23

Alas there is evidence that this is the case. I assume we don't disagree on the deviation. But since there are several tests for iq, we have a certain iq test that has results that are on average 12 points lower than another iq test. So either one of the tests is inaccurate and off by 12 points, or both of the tests are inaccurate, off by 0-12 points.

wich claim I made are you referring to? — Tomseltje

You said that rejecting IQ tests shows low IQ, and to question its validity shows high IQ.

Why do you want me to cite an iq test as evidence? — Tomseltje

Because you are questioning them, but you do not cite anyone. It is such as to question that thermometers are not reliable, but without to give an example of it. Your criticisms is not critic.

You said that rejecting IQ tests shows low IQ, and to question its validity shows high IQ. — Belter

As an improvement of what I considered to be a joke you stated:

The questioning of IQ validity is an evidence of low IQ. — Belter

I merely attempted to point out the difference between questioning something, and dismissing something. Where scientifically it's always ok to question something, as in applieng scrutiny. Wich is different from dismissing something, as in stating it's validity is 0.

Though if you didn't intend it as a joke, I would say:
At best they are an indicator, like all the questions in an iq test are mere indicators, having a single question on an iq test right or wrong sais nothing about that persons iq, you can't apply statistics to individual cases.

So how about if we measure weight with a thermometer? (We'll just call it a "weight-measuring device"... for good measure.) We are measuring something, we can do comparisons, calculate average, etc. We should be good, right? — SophistiCat

Nice strawman.
You missed the point. The point being is that you can measure body lenght, or body weight, but just the measuring result in meters or kg doesn't tell you wether the person you measured is tall or heavy. In order to make such a determination you have to compare the measured result with the average. We have units for such measurements that are quite fixed.

The difference with iq tests is that the unit isn't fixed, but gets updated each year to compensate for the flyn effect.

Because you are questioning them, but you do not cite anyone. It is such as to question that thermometers are not reliable, but without to give an example of it. — Belter

Very well, here is an example, the precision of an iq test has the 95% reliability interval at measured iq + or - 15 iq points. 15 iq points is also the standard deviation in measured iq in a population.

I'm not saying they are not reliable, their reliability is dependant on what they are used for. A thermometer that may give a value 10 K more or less than the actual temperature is quite unsuited to measure the temperature of an incubater for bacteria, but it's adequate for measuring metal temperatures when forging in a coal fire.

OK, but the general point holds: activities required to gain a skill are not required to maintain a skill. Recognising an "s" is a skill, perhaps measuring is required to gain that skill (although I am still unconvinced of that, you seem to have a model of human cognition that is generally contestable), but even so it does not follow from that that measuring in any way shape or form is required to maintain it. — MetaphysicsNow

Recognizing something using your eyes includes measuring, once you can recognize the s with your eyes closed you don't need to measure it any more. If you have to use your sense(s) to make an assesment, you are measuring, no matter how quickly you do it. Measuring is a skill by itself. It can be trained and improved.

Give the violin player a violin where the strings are twice as much apart, and he/she won't be able to play by just using muscle memory any more. Since the 's' you read, isn't the same shape and size in each instance, you can't rely on muscle memory either, and must measure the 's' in order to recognize it as an 's'.

Very well, here is an example, the precision of an iq test has the 95% reliability interval at measured iq + or - 15 iq points. 15 iq points is also the standard deviation in measured iq in a population. — Tomseltje

It is another abstract example. I do not understand what want you say with this. In my view, a rational skepticism would question the validity of the Raven Test, or any other IQ test (Mensa, etc).

Recognizing something using your eyes includes measuring

Let's recap where we are on this particular line of thought.
I claim that people can recognise intelligent behaviour without having to measure anything.
I point out that recognising something and measuring something are entirely distinct kinds of activity.
You admit that, but you say that nevertheless recognising something always requires measuring something.

I then go on to question this, since on a normal understanding of what measuring is, in so far as it is an activity engaged in by human beings, it is an intentional activity engaged in knowingly. For example, measuring the height and length of a wall in order to work out how much paint I'll be needing.
No such intentional measuring activity is going on when I recognise an "s" on a page, and I doubt very much that it is going on when you do either. I can certainly imagine circumstances where it would be going on - for instance, if I want to make sure that the "s" I have picked out is of exactly the right size to fit into the space I have on the document I am trying to forge, then recognising and measuring an "s" are going on together. However, note that in this kind of case, it is the measuring that depends on the recognition, not the other way around. You seem to be suggesting now, however, that there is some arcane notion of measurement which nevertheless is always going on when I visually recognise anything.

What is that notion of measurement? It is certainly not the usual one. I expect you will be tempted to say something along the lines, well, when I recognise anything at all my brain is measuring stuff.

My reply to that is that you are making a category error: brains are not things that measure. Human beings measure things and by analogic extension, we have created devices that also measure things, but stricly speaking even those devices do not measure anything, we measure things with those devices. It makes sense to say "thermometers measure temperature" since we measure temperature with thermometers, for instance, but strictly speaking, what thermometers do is react in characteristic and usually predictable ways to certain kinds of environmental change, such that they can serve as devices for measuring temperature.

So, now, in what sense do brains measure anything? Brains do not engage in intentional measuring activity themselves, and we certainly do not use them as we do devices such as thermometers. I can probably come up with a scenario in which I might use a brain to measure something, but it would be a truly bizarre scenario - perhaps I want to see if a box I have is big enough to carry a brain in, so I see if a brain I have to hand fits in the box. Arguably there I have used a brain to measure the size of a box.

Now I think you might be tempted to say something like, "but look at MRI scans that are recording what is going on in the brain when people are doing stuff like recognising "s"s: there's specific brain activity going on!". Well, I suppose we could try to use an MRI scan to measure brain activity when a person is recognising an "s" - I am not denying that in the least, although as far as I am aware nobody has been able to pinpoint specific types of brain activity corresponding to recognising anything, let alone something specific like an "s", although its early days of brain science. Regardless, what you seem to be wanting to say is that not only do MRI scans measure brain activity (in the sense that we can use them to measure brain activity), but they are measuring themeasuring activity of the brain. But there we are again back at the same category error. After all, if the brain is actively measuring anything, what is it measuring that thing with?

You probably think I don't understand what you mean by "measuring" and you are probably right. I supsect what you mean by "measuring" is not in fact measuring at all.

activities required to gain a skill are not required to maintain a skill — MetaphysicsNow

It would depend on the skill. When getting skilled in archery for example, you look at the target. Once you are a skilled archer, you still have to look at the target in order to hit it. Even if you were training to shoot blindfolded, in both cases there is still the activity of holding the bow. So in this case your assesment doesn't hold.

It is another abstract example — Belter

iq is a rather abstract concept, what were you expecting?

In my view, a rational skepticism would question the validity of the Raven Test, or any other IQ test — Belter

All scientific findings should be questioned, it's called scrutiny. When it comes to science it can't happen too much.

For example, measuring the height and length of a wall in order to work out how much paint I'll be needing. — MetaphysicsNow

To measure the wall in your example, one could use a measuring device. However a skilled painter can do it without, just by looking at the surface and making an estimate of the surface area. In both cases the width and height of the surface area has been measured.

No such intentional measuring activity is going on when I recognise an "s" on a page — MetaphysicsNow

That's what you think, I didn't say you should be concious of it.
When a preditory animal encounters another animal, an estimation gets made: Is this something to eat, is this something to run away from. In order to make the estimation, the size of the other animal gets measured. Is it smaller then it's something to eat, is it bigger, then it's something to run away from.

What is that notion of measurement? — MetaphysicsNow

I gave you the definition I used, what more do you want?

brains are not things that measure. Human beings measure things and by analogic extension, we have created devices that also measure things — MetaphysicsNow

This is where you go wrong, reread the definition I gave, a device is not needed to measure, the definition doesn't mention its requirement, though it can still help to obtain more accurate measurements.

It makes sense to say "thermometers measure temperature" since we measure temperature with thermometers — MetaphysicsNow

We measured temperature with our heat sensor cells in our skin way before we discovered how to make a device like a thermometer. You may think its an uncommen way of applieng the word to measure, but actually it's the more common way, we have done so for millions of years, thermomenters only exist a few hundred years.
If we weren't able to measure without such devices, we wouldn't be able to determine whether the water we were boiling is hot or still cold when putting our hand into it.

Stop applieng your narrowed interpretation of the word measure, and start applieng the definition given if you want to make sense in your responses.

That's what you think, I didn't say you should be concious of it.

So we can unconsciously measure things?
Here is your definition of "measurement"
" the act or process of ascertaining the extent, dimensions, or quantity of something;"
Well acts and processes are distinct things, but presumably based on this definition, to measure something is to act in order to ascertain the amount of something .....So, how can I act to ascertain anything - let alone extents and dimensions - unconsiously? NB, this is not a general question about whether some of my actions can be unconscious or not, this is specifically about actions aimed at ascertaining results.
Perhaps if I somehow managed to use a tape measure during an episode of somnabulance I could be said to have performed an act of measuring whilst unconscious, but it would be tough to say that I had actually ascertained anything in doing so.

We measured temperature with our heat sensor cells in our skin way before we discovered how to make a device like a thermometer. You may think its an uncommen way of applieng the word to measure, but actually it's the more common way, we have done so for millions of years, thermomenters only exist a few hundred years.
If we weren't able to measure without such devices, we wouldn't be able to determine whether the water we were boiling is hot or still cold when putting our hand into it.

Well, there are probably some fine-grained distinctions to make between testing the temperature of something with ones hand and measuring the temperature, but sure perhaps we can use bodily parts as instruments to measure things in general: an experienced painter may indeed size-up a wall with his eyes and then estimate how much paint he needs. Nothing I've said rules any of that out. I think the main point is the intentionality of measurement, not whether the instruments we use to do so are natural or artifactual.

You are probably right on this part, however, that doesn't mean it's an indication of my level of understanding statistics. More likely it's just a translation error, since I'm not a native english speaker.
My mathematical training was in my native language. So thanks for pointing it out. — Tomseltje

I understand. English is also not my mother tongue so I can understand it happens.

My apologies for causing confusion on this, I didn't think the difference was that relevant to the subject in this case. Since on a single measurement the precision influences the accuracy. Wich applies in this subject, since generally the individuals only get tested once for iq.

I am still unsure if you are correctly understanding statistics.

Precision and accuracy are independent parameters. That means the value of one of them will not influence the value of the other. Simply put, precision DOES NOT influence the accuracy. You still seem to confuse these two terms. It doesn't matter how many times the measurement is done (even if zero measurement is done). Any test has its own intrinsic precision and accuracy.


For example, let's say we are trying to measure the radiation level of a copper-made statue using a Geiger counter.

Geiger counter works by having a electromagnetic wave (of sufficient photon energy like radiation from radioactive material) come into the probe where the gas inside the probe is ionized and it results in electrical current in presence of an applied voltage.

Now, we put the Geiger counter in a perfectly dark nonradioactive box. What would the meter say? It will still have some non-zero number. But why? The Geiger counter is put inside a box where no radiation exists so it shouldn't be detecting anything. This is called "dark counts". There are several reasons why this happens and I won't get to that, but severely high dark counts (for some poorly constructed counters) will suffer high signal-to-noise ratio.

We take the Geiger counter out and go outside. There are tons of radiation source (sunlight, radiation from the ground, etc). This will also worsen the signal-to-noise ratio. Now you go measure the radioactivity of a copper-made statue. Copper naturally have radioactive isotopes, so its decay will emit radiation. Picking this radiation up with Geiger counter gives you a value.

Now, you measure just once (let's say 2 seconds). How accurate is that value? Well it should be inaccurate. That is, the noise level (due to dark count and background radiation) causes a shift to a higher value than it is supposed to. Lack of accuracy due to noise is often processed by taking something called "background measurement". You take the average value of the background and subtract it from the signal value you got. Also, quantum efficiency of Geiger counter is not 100%. Most manuals provide how efficient Geiger counter is at detecting radiation. So you also account for this by dividing the efficiency from the value.

Good, now we should be getting the accurate value. Then how about precision?

If you have done only one single measurement, then you have no way of knowing the precision of the Geiger counter. In statistical terms, this is called "level of confidence". You need to do numerous measurement until the value (in average) converges to a certain value. Most of the time, 95% level of confidence is good enough statistically, but this depends on the exact measurement you are trying to do as well as personal preferences. If you have done only one measurement, then the level of confidence should be extremely low.


So to apply this to what you said about "IQ test only being performed once for each person", this refers to the level of confidence. It is not about precision (although not unrelated)or accuracy. Please be sure to understand these concepts before claiming that you understand statistics. The more you talk about it gives me higher level of confidence that you don't understand statistics.


(To the moderators, I apologize for the slightly off-topic post but the knowledge of statistics is crucially important for understanding IQ tests)

Well, what I said wasn't really philosophy but science. It is scientifically well established that brain performs measurement in processing information. Just that it is not the sort of things you are imagining "measurements" to be.

But yes, it is slightly off-topic so let's leave it at that.

Alas there is evidence that this is the case. I assume we don't disagree on the deviation. But since there are several tests for iq, we have a certain iq test that has results that are on average 12 points lower than another iq test. So either one of the tests is inaccurate and off by 12 points, or both of the tests are inaccurate, off by 0-12 points. — Tomseltje

I indeed agree that there is a deviation.

Can you be a little more specific about "several" test? Are they test with completely different problems? Or are they test with similar problems? This is particularly important when discussing statistics.

Well it kind of is and kind of is not off-topic. Do you have a link to a paper where it is scientifically established that brains perform measurements? I think there is some co-opting of this term "measurement" going on in psychology and it is being given a very specific technical sense when experiments are being carried out, and then when conclusions are being drawn the usual everyday sense is being ushered back in through the back door.

Well I am not a neuroscientist, and to be honest I can only refer to lectures from neuroscientists, the most famous out of them was Edvard I. Moser in a conference meeting.

So short answer, no. I don't know specific scientific papers.


However, it should be kept in mind that the only reason we have a disagreement is because we do not share the same definition of measurement. Your definition of measurement seems to be very instrumentally and computationally quantitative, compared to me and Tomseltje's definition which is more general.

But brain do process information in a neuro-network in an extremely complicated but interesting way, which is, from a neuroscientific standpoint, can be effectively called "measurement" and "processing".
I understand well that these words are ill-defined, but this is merely a terminology problem and does not change the fact that the brain process information in a way that involves "measurement".

I'm not convinced it is just terminology. The point is scepticism about there being a thing called intelligence that comes in amounts and can be measured, in the ordinary sense of measurement that goes on when people take themselves to be measuring IQ. When I doubt this, I am told that it must be a measurable thing because we can recognise intelligent behaviour. I claim that recognition and measurement are entirely distinct things and recognising something need not involve measuring anything - in that same perfectly ordinary sense of measurement, and also - as if it were any different - in the sense defined by Tomseltj. Then I am told there is a special technical sense of measurement, but no one seems able to tell me what that is, nor why it needs to be going on when I recognise intelligent behaviour. If all you mean is that there must be brain activity going on when I recognise intelligent behaviour, sure, maybe that's right, but that does not entail that what that activity is is measuring something called intelligence.

Well acts and processes are distinct things — MetaphysicsNow

not from where i stand. I can move my arm as an act, wich includes several chemical processes within my muscles. It's impossible for me to perform that act without those chemical processes taking place.

You sound alot like someone acknowledging that he/she can move his/her arm, but denies the chemical reactions taking place within your muscle tissue.