Re: Working Memory improvement! Oh joy.

[Pontus Granström]

One of my "dreams" if I can call it that, is that people did pre/post
"bip testing", for scientific reasons. Since n-backing in some sense
improves the span speed and surely
one parameter in the BIP equation it might very well be dnb that is
behind the increase.

On Thu, Apr 7, 2011 at 10:29 PM, D <damian....@gmail.com> wrote:
> I apologise if I've been posting too much in recent times, but this I
> had to share.
>
> On
> http://www.mental-aktiv.de/mental-aktiv/Mentaltraining/Tests/flash_1.html
>
> I tested myself four days ago (on the 3rd April) and achieved a score
> of 57.71bps.
>
> Since then, I've been training consistently on dual n-back, doing
> exactly twenty sessions each day, with enough trials to put my focus
> to the test (each session always lasts longer than 2 minutes).
>
> My score today is 145.6bps.
>
> Sure I can attribute some increase to getting used to the game, but
> this is only the second time I've played it. Today I discovered that
> when the stream of letters sped up, I could recall them more easily.
>
> Surely dual n-back has had a say in my improvement?
>
> --
> You received this message because you are subscribed to the Google Groups "Dual N-Back, Brain Training & Intelligence" group.
> To post to this group, send email to brain-t...@googlegroups.com.
> To unsubscribe from this group, send email to brain-trainin...@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/brain-training?hl=en.
>
>

[Ashirgo]

No evidence for this kind of improvement was found, if my memory is
worth anything.

[Pontus Granström]

If you want to improve your reading span in terms of recall you need
to train with the CWS protocol if I recall it correctly. However the
mere speed of reading
a line of letters might be improved from n-backing. I broke my
personal record at the "speed part" two days ago with a bit/s of 28.6
(155 IQ).

On Fri, Apr 8, 2011 at 9:25 AM, Ashirgo <wielo...@gmail.com> wrote:
> No evidence for this kind of improvement was found, if my memory is
> worth anything.
>

[Działo, Christopher]

Pontus,

I took the speed part this morning and scored 30.8 bits/s. (http://i51.tinypic.com/2d0xmhz.png)

Does this imply IQ 155+ based off of your above comment?

[Działo, Christopher]

Pontus,

A strange phenomena has occurred with my BIP scores. I've mentioned this before, but several weeks ago I was unable to surpass no more than 11 bits/s, and now score in between 28-30.8bit/s. My dyslexia still troubles me at times, resulting in accidental reading and typing of the viewed letters, and throughly believe it's holding back even greater BIP performance.

I'm unsure who to ask, and you appear very active in these groups -- can you further recommend any exercises that can improve various aspects of G? I do daily training at mental-aktiv.de and play several rounds of color, letter, and number match, along with routine n-backing. I also experienced a jump in performance, finally surpassing 4-back, 5-back, and now scoring in the 50s range on 6-back, all within a week's time.

Thanks,

Christopher 

2011/4/8 Działo, Christopher <chris...@gmail.com>

[Pontus Granström]

If you look at the graph for the correlation between BIP and
psychometric IQ you will find that each bit in general corresponds to
5 points on the IQ-scale. 25 bit is a 140 IQ,
and 30 bit would be a 165 IQ, as I got 28.6 as my highest score that
would correspond to 155. As for improvements, I was stuck at level 4
for quite some time but
now I can hardly miss any trial at that level, it seems that
improvements come when least suspected and often in unusual ways.
Perhaps it all depends on why
one struggles at a level, might be that proceeding to higher levels
are associated with a build up of white matter, similar to as the
thickness of the muscle fibers increases
when doing workouts! By who knows.. Nice that you have progressed though.

[Działo, Christopher]

Can you explain the CWS protocol you mentioned? I did a search on Google but was unable to find anything.

[Pontus Granström]

It's called CWM (my bad), you can find the description of it here
http://www.jtoomim.org/brain-training/expanding%20the%20minds%20workspace.pdf
It's more of store/recall (memory operations for reading in other
words) rather than an update compare (reasoning). This protocol
transferred to tasks that
n-back did not.

2011/4/9 Działo, Christopher <chris...@gmail.com>:

[cev]

Is there a mental speed test similar to that on the mentalaktiv site
that does not repeat certain strings of letters? Results on the
mentalaktiv version may become less valid with practice.

[Ashirgo]

My scores are: (in no particular order)
23.5
28.3
and 30.8 BIP.

I wonder why such high variability. One way to explain it would be
that a brain is not a very reliable machine, and it has its ups and
downs :)

Ash

[Pontus Granström]

You got similar progression as I've got, I scored 30.8 today for the
first time. But as you point out, there will be some variation, maybe
even because you get used to the sequences that repeat.

[Pontus Granström]

Took the test again scored 33.3, don't know if it's 100% "fair" though.

[cev]

Regarding the IQ correlations, does not the letter reading test that
Lehrl mentions involve reading letters out loud?

In fact, Lehrl details the procedure for this 'letter reading' test in
one of his papers: read out 20 unconnected, monosyllabic letters as
quickly as one can. Best result of four attempts counts, 100/time
taken gives the bits/s.

[Pontus Granström]

But they are measured in the same unit bit/s so I guess they are equal?

--

[cev]

As far as I remember, he provides a table comparing mean bits/s on
various tests and vocalised letter reading is a little more difficult.

33 bits/s would mean reading out the 20 letters in just a smidge above
3 seconds - tricky!

Possibly to put another spanner in the works, recent laymanly research
of mine suggested that high performance on letter reading correlates
much more strongly with verbal reasoning than non-verbal (0.8 vs <0.5,
c.25 subjects).

Letter reading performance also correlated very strongly with English
language and literature exam results - more strongly than verbal
reasoning test scores did! (0.89 vs 0.83, as far as I remember). This
is a nice result, as our verbal reasoning tests take 30 times longer
to administer than the letter reading task does and are probably more
susceptible to retest effects.

I would thus conclude that the test outputs one's verbal IQ, not
general IQ.

On Apr 9, 12:09 pm, Pontus Granström <lepon...@gmail.com> wrote:
> But they are measured in the same unit bit/s so I guess they are equal?
>

[Ashirgo]

I have found some free Raven test. Can any of you take it and post
your results?

http://www.clipsite.com.ar/HOME/Salud/Test/Raven/Principal.asp
It's important to give age (Edad) to have your results calculated
correctly.

[Pontus Granström]

Scored 134 on that test, however I were extremley tired, got 57/60.

--

[cev]

The scale is Cattell-like, I think, as I got 134 (57/60) and it gave
my 'percentil' as 95th.

So 147 is a mote short of the 98th percentile?

On Apr 9, 2:39 pm, likeprestige <plastic...@live.com.au> wrote:
> Does anyone know what the scale of this IQ test is? Would like to be
> able to compare this results with a couple of other online tests.
>
> Thanks.
>
> On Apr 9, 11:32 pm, likeprestige <plastic...@live.com.au> wrote:
>
> > Highest IQ score I've achieved.
>
> > IQ - 147
>
> > I nearly couldn't be bothered when it took so long for the problems to
> > get harder.
>
> > I don't think it's an accurate reflection of my 'real' IQ though,
> > practice effect with Raven's type questions. Somewhere between 125-135
> > is more realistic I think.

[Pontus Granström]

Seems like the questions in the end are easier than some early and mid questions.

[Thomas Johnson]

Sorry, can someone explain to me how this test works? The instructions are all in German, unless I'm missing an English version of the site

[whoisbambam]

thomas,
this is better than nothing...
select that text and paste it into here:
http://translate.google.com/

you will get:

Measuring the amount of memory as a self-test
Thank you for your interest to determine your work capacity. The
memory capacity is an important fundamental variable to describe the
brain power.
In the test method, first the memory span of numbers and letters are
collected and analyzed. Subsequently, the IVG (information processing
speed) were measured. The memory capacity is the product of memory
span and IVG. The evaluation is done automatically.
You can start the 3 parts of the measurements independently.


then:
Memory span measure
When you are first memory span of 3 characters (first numbers, then
letters) are shown. You will be prompted to sign in the correct
sequence the text box to enter. When entering the correct test to the
next stage will be continued.
An incorrect entry, the step is repeated once more. After a second
incorrect input is clear your results and begin the next phase of
testing can.

On Apr 9, 3:03 pm, Thomas Johnson <thomas.j.john...@gmail.com> wrote:
> Sorry, can someone explain to me how this test works? The instructions are
> all in German, unless I'm missing an English version of the site
>

> On Sat, Apr 9, 2011 at 1:59 PM, Pontus Granström <lepon...@gmail.com> wrote:
> > Seems like the questions in the end are easier than some early and mid
> > questions.
>

[A.Four.Sigma]

Gentlemen,

Why go out to buy juggling ball sets and equipment when you can simply
ball up a few pair of socks? Too much brain training and not enough
[creative] thinking!

A4S

[milestones]

On Apr 9, 10:39 pm, likeprestige <plastic...@live.com.au> wrote:
> Does anyone know what the scale of this IQ test is? Would like to be
> able to compare this results with a couple of other online tests.
>
> Thanks

I took this a while back and scored 58/60.

The test looks to be a knock off of the Standard Progressive Matrices,
a test geared for those in the normal range of (fluid) intelligence.
This is not the sort of test to administer to above average subjects,
but surely it's a good test for those between 75-125. I'm not sure
whether it's Wechsler, Binet, or Cattell scale, but if it's sd 24 then
it really is an incredibly low ceiling. I think, however, it's
probably sd 15/16 because it's in the tradition fo SPM (60 questions).
If one scores at 125 or higher, then better to take another test with
a higher ceiling. The ceiling is too low to be of much discriminating
value for those in the above average range. That said, if you score in
a range above 125, then you probably have a 125+ (fluid) IQ. I'd say
that the GIGI test is probably more appropriate test for those with
the above average range.

Cev: ("Letter reading performance also correlated very strongly with

English
language and literature exam results - more strongly than verbal
reasoning test scores did! (0.89 vs 0.83, as far as I remember). This
is a nice result, as our verbal reasoning tests take 30 times longer
to administer than the letter reading task does and are probably more

susceptible to retest effects.)

I would thus conclude that the test outputs one's verbal IQ, not
general IQ.

That's interesting. I find it fascinating to think that this sort of
test would equal the G loading of say, an English vocabulary test. I
don't doubt what you say about verbal IQ correlations, but this test
seems to be capturing processing speed more than verbal reasoning. I
scored 33.3 on it when I first took it, scored same again when I took
it yesterday. I think that this test might provide a baseline for
130-98%ile + (or verbal, as you say) IQ but then -- similar to most IQ
tests -- not tell us really tell us much above that....so equating
performance on this measure to 4 sigma IQ scores+ is a joke.

[bofu]

I got the concept besides not knowing any German. Tried it once - it's
midnight. IVG 165 bit/s Is that good? Not sure what IVG is.

On Apr 7, 3:29 pm, D <damian.amen...@gmail.com> wrote:
> I apologise if I've been posting too much in recent times, but this I
> had to share.
>
> Onhttp://www.mental-aktiv.de/mental-aktiv/Mentaltraining/Tests/flash_1....
>
> I tested myself four days ago (on the 3rd April) and achieved a score
> of 57.71bps.
>
> Since then, I've been training consistently on dual n-back, doing
> exactly twenty sessions each day, with enough trials to put my focus
> to the test (each session always lasts longer than 2 minutes).
>
> My score today is 145.6bps.
>
> Sure I can attribute some increase to getting used to the game, but
> this is only the second time I've played it. Today I discovered that
> when the stream of letters sped up, I could recall them more easily.
>
> Surely dual n-back has had a say in my improvement?

[cev]

Ah sorry, I wasn't especially clear: those correlations are with the
_verbalised_ letter reading test, not the register and remember letter
reading test on mentalaktiv.

I did run a test with the mentalaktiv version and found one
particularly interesting result: none of the 8-9 yr olds scored above
adult average bits/s but the very brightest (top 2% of the population
by reasoning) of the 9-10 yr old students _scored within the adult
percentile equivalent to their age-appropriate percentile_ - almost as
if the visual processing part of the brain had been upgraded all in
one fell swoop! None of the 9-10s' working memory scores were
particularly good - their average well below adult average - but they
were marginally better than the 8-9s'.

The most musically able student (reasoning scores of c.110) scored
highest on working memory (average span of 7).

Also, two other (very bright, top 2%) students I suspect of having
ADHD (v. impulsive) scored very highly on unverbalised letter reading
and very poorly on working memory. My half-baked theory is that ADHD
is not a working memory _capacity_ deficit but a deficit in conscious
information organisation and retrieval (working memory access, I
suppose), as otherwise I don't understand how they could do so well on
reasoning tests.

[Działo, Christopher]

First attempt: 54/60 IQ 120

Second attempt: 56/60 IQ 132 (under the influence of a small amount of cannabis)

Would it be considered cheating using pen and paper to write down and draw relations in the RPM questions? I can sense and see various patterns, but I have trouble holding the pattens in working memory and then performing manipulation operations. Though, I don't quite understand why getting 2 more questions correct would 'boost my IQ 13' points. When I take tests and exams, I often know -- at least with high probability, which questions I answered incorrect. I try and feel if the answer is correct, and assume if I were to review my answers I would score higher. I changed a few of my previous selections based on which I felt I answered incorrectly the first time, and tried to figure out the harder ones with pen and paper -- I often become apprehensive and begin to doubt my self. I believe I may be placing too much emphasis on this whole IQ frivolity. I've come to terms that I don't have an IQ of 200+, but embrace that I, and many other members here are at least in the top 2%.

Random question -- Does anyone have any research or experience on the effects of formal, continuous education and the influence on Gf? I completed all but 1.5 years of high school (dropped out) and was never enrolled in more challenging classes such as calculus, physics, and chemistry, and feel this has had a negative effect on the development of my intellect and overall problem solving abilities as I never stressed them as often as a student in a rigorous academic setting.

[Pontus Granström]

I think I prefer the Gigi over this one, as the gigi also takes the time into account.

On Sun, Apr 10, 2011 at 11:49 AM, likeprestige <plast...@live.com.au> wrote:

Also, I think my score must have been 137 not 147 because I can't seem
to get over 140, regardless of how fast or accurate I complete the
test.

If anyone wants help with any of the questions let me know.

 :-)

Three limitations about this IQ test:

1. It is not adaptive
2. The testing items are not difficult enough to discern differences
in the High IQ range
3. There is no reward for completing the test at a faster rate (this
point is admittedly debatable - however, when the testing items are at
this [low] difficulty level I think it should be taken into
consideration)

On Apr 10, 4:00 pm, likeprestige <plastic...@live.com.au> wrote:
> I'm wrong, It's not an adaptive style IQ test (same questions each
> time), which in my opinion makes it less valid.

>
> On Apr 10, 4:59 am, Pontus Granström <lepon...@gmail.com> wrote:
>
>
>
>
>
>
>
> > Seems like the questions in the end are easier than some early and mid
> > questions.
>

[Działo, Christopher]

I should take the gigi test, but after hearing it takes into account time this aspect of the test will be a challenge for me. I've always been a tad slow at test taking, especially when my performance will decide my IQ.

[hallu]

Heh, I've solved the gigi for the second time a few days ago with a metronome to see how much speed is involved in the result (both tests were taken across two days). The second time I was at least twice faster, but scored 5 points less.... LOL. I had to be benumbed and made some blind shots obviously.

[J.]

Does anyone know where I could find the CWM task mentioned earlier in
this thread? It seems like it would be a useful complement to n-back
training.

[whoisbambam]

i cant even figure out how to do math n back.
:(
oh well.
i would need a very super easy version of it too.


On Apr 11, 8:05 pm, likeprestige <plastic...@live.com.au> wrote:
> I wouldn't be surprised if -Crab arithmetic dual-n-back- tapped the
> same processes (partly why I train using this mode), otherwise I am
> unaware of any.

[likeprestige]

Ok, here it is...

---Arithmetic 1-back----

The least harsh form of the available variations in this mode.

1. Step 1 - You will see a number presented on the screen
2. Step 2 - After the number has disappeared, a second number will
present itself in conjunction with an auditory input that will
instruct you to perform a mathematical operation (division, plus/
minus, multiply) with the first & second presented number.

This pattern continues with a math operation being performed on - the
2nd & 3rd, 3rd & 4th, 4th & 5th and so on...

For example,

First number presented = 7
Second number presented = 9
Auditory input = multiply

This means that the math operation is 7x 9 which = 63

Let's continue...

The third number presented = 5
Auditory input = plus
Math operation is now 9+5 which= 14

The fourth number presented = 6
Auditory input = minus
Math operation is now 5 - 6 which = -1

And so it continues...


Two points to consider:

1. Maybe start off here (arithmetic n-back [no position stimuli or
other variants attached]) & then work your way up the ladder until you
reach crab arithmetic dual-n-back.

2. Turn of __division__ in the config. file & ensure that the numbers
__do not__ exceed 12 (otherwise numbers start two equal more than
syllables which adds to the verbal load in an inconsistent fashion)


Hope that helps

[whoisbambam]

likeprestige,

Wow. that definitely helps.

it seems hollow, but..........

thank you!!!!!!!!!!!!

saving to a text file!

i would definitely start with arithmetic nback 1 and mess with the
time

the pointer on 12 is very nice............

turn off division? well, if it divides in whole numbers only, that
would be cool..........but if there are remainders, etc, that would be
a problem.

[likeprestige]

Actually, I suggest the max number to be 9 'above' anything else. It
may seem a bit pedantic but although the number 10 has only one
syllable, it has a larger visual _value_ compared to it's
predecessors (9,8,7,6,5 etc). Some may say that it doesn't matter
because we just represent the number as a whole rather than taking in
its individual units one at a time, however this may not be completely
accurate. The idea has promise from the intellectual perspective,
however, psychologically it perhaps becomes a lot more problematic.
Just an opinion though...

On Apr 12, 8:05 pm, likeprestige <plastic...@live.com.au> wrote:
> "the pointer on 12 is very nice............"
>

> Yeah, lol, you might want to make that 10 not 12; I'm not sure how I
> mistyped that!

[whoisbambam]

ok. 9 it is.

but 'thirt teen' and thereafter (excluding the new tens place, 20, 30,
40) is two words and it is that idea i thought you were talking about.

ten, eleven, and twelve are still one word.

ntm the two digit place compounding issue.........

perhaps 0-9, then 0-12 in stage 2

[cev]

Just got a string of good scores on the Verarbeitunggeschwindigkeit
test at mentalaktiv: 66.67, 40, 40 - the first score double my
previous best.

I have been doing their recommended word search training exercise for
3 mins each of the last four days, but no increase in my usual 26-33
range...until now.

Bizarre.

http://tinypic.com/view.php?pic=301gbns&s=7

PS - World Record? :-)

[hallu]

nice score, for me this seems to be consistent with some of the iq tests (gigi seems to be 10 points too high, probably due to practice effects from other tests), I've trained Verarbeitunggeschwindigkeit for three days, it improved my score by 3 bits

[Pontus Granström]

I scored 152 on the GIGI, I guess that my IQ is around 140 (I have scored that on at least two measure of fluid intelligence tests) however since the GIGI is a world
test and the average IQ in the world only is around 90 (relatively the "developed world"), it might actually be correct.

On Wed, Apr 13, 2011 at 6:44 PM, hallu <hall...@gmail.com> wrote:

nice score, for me this seems to be consistent with some of the iq tests (gigi seems to be 10 points too high, probably due to practice effects from other tests), I've trained Verarbeitunggeschwindigkeit for three days, it improved my score by 3 bits

--

[enigmamachine]

I just tried this and scored 225.0, despite being very sleepy and
having a slight headache. Going off of the graph found here:
http://www.v-weiss.de/lehrl-full.html, the correlation is
approximately 1 bit to 4.1 IQ points. This gives me a modest IQ of
963, which is a mere 57.5 standard deviations above the mean. That
sounds reasonable, right? I will have to read over this study
tomorrow, but such an impossibly high projected IQ score, especially
in my current state, makes me doubt its validity. Can anyone speak to
this?

My screenshot: http://imgur.com/MSajA

On Apr 8, 9:29 am, Pontus Granström <lepon...@gmail.com> wrote:
> If you look at the graph for the correlation between BIP and
> psychometric IQ you will find that each bit in general corresponds to
> 5 points on the IQ-scale. 25 bit is a 140 IQ,
> and 30 bit would be a 165 IQ, as I got 28.6 as my highest score that
> would correspond to 155. As for improvements, I was stuck at level 4
> for quite some time but
> now I can hardly miss any trial at that level, it seems that
> improvements come when least suspected and often in unusual ways.
> Perhaps it all depends on why
> one struggles at a level, might be that proceeding to higher levels
> are associated with a build up of white matter, similar to as the
> thickness of the muscle fibers increases
> when doing workouts! By who knows.. Nice that you have progressed though.
>
> 2011/4/8 Działo, Christopher <chrisdzi...@gmail.com>:
>
>
>
>
>
>
>
> > Pontus,
> > A strange phenomena has occurred with my BIP scores. I've mentioned this
> > before, but several weeks ago I was unable to surpass no more than 11
> > bits/s, and now score in between 28-30.8bit/s. My dyslexia still troubles me
> > at times, resulting in accidental reading and typing of the viewed letters,
> > and throughly believe it's holding back even greater BIP performance.
> > I'm unsure who to ask, and you appear very active in these groups -- can you
> > further recommend any exercises that can improve various aspects of G? I do
> > daily training at mental-aktiv.de and play several rounds of color, letter,
> > and number match, along with routine n-backing. I also experienced a jump in
> > performance, finally surpassing 4-back, 5-back, and now scoring in the 50s
> > range on 6-back, all within a week's time.
> > Thanks,
> > Christopher
>
> > 2011/4/8 Działo, Christopher <chrisdzi...@gmail.com>
>
> >> Pontus,
> >> I took the speed part this morning and scored 30.8 bits/s.
> >> (http://i51.tinypic.com/2d0xmhz.png)
> >> Does this imply IQ 155+ based off of your above comment?
>
> >> On Fri, Apr 8, 2011 at 7:02 AM, Pontus Granström <lepon...@gmail.com>
> >> wrote:
>
> >>> If you want to improve your reading span in terms of recall you need
> >>> to train with the CWS protocol if I recall it correctly. However the
> >>> mere speed of reading
> >>> a line of letters might be improved from n-backing. I broke my
> >>> personal record at the "speed part" two days ago with a bit/s of 28.6
> >>> (155 IQ).
>
> >>> On Fri, Apr 8, 2011 at 9:25 AM, Ashirgo <wieloslo...@gmail.com> wrote:
> >>> > No evidence for this kind of improvement was found, if my memory is
> >>> > worth anything.

[Działo, Christopher]

You are incorrect.

Go to :http://www.mental-aktiv.de/mental-aktiv/Mentaltraining/Test_Gehirnleistung.php

And take the middle test (Test for information processing speed (IVG)

This will give you the correct value for your BIP/IQ evaluation. 

[Pontus Granström]

You got your bits score which is the capacity of your short term memory (C), 225 bits would be around IQ 142.

2011/4/17 Działo, Christopher <chris...@gmail.com>

[enigmamachine]

I took the middle test. It was poorly implemented and used a set of
four letter combinations that never change, so I wrote a better
version that uses randomly generated "words." Now I just need a place
to host it. I got 43.8 BIP, by the way.

On Apr 17, 3:56 am, Pontus Granström <lepon...@gmail.com> wrote:
> You got your bits score which is the capacity of your short term memory (C),
> 225 bits would be around IQ 142.
>

> 2011/4/17 Działo, Christopher <chrisdzi...@gmail.com>

>
>
>
>
>
>
>
> > You are incorrect.
>
> > Go to :

> >http://www.mental-aktiv.de/mental-aktiv/Mentaltraining/Test_Gehirnlei...

> > And take the middle test (Test for information processing speed (IVG)
>
> > This will give you the correct value for your BIP/IQ evaluation.
>

[D]

Did you say 'words'? Wouldn't it be easier to memorise words so long
as you know how to spell them?

[enigmamachine]

I'm just calling them "words," because it is easier to type than
"strings containing four random letters." The problem with the
original test is that it used a list of "words" and was unable to
create new ones. Play it through a few times and you will see what i
mean. The creators of the test were really fond of "orz" and "uae,"
for example.