Multi-Stim N-Backing

[Reece]

I didn't want to contaminate Jonathan's test release + bug report
thread with this post, however I thought now that Brain Workshop has
had Multi-Stim mode out for a bit that it'd be interesting to here
what other people think about this new mode.

I am finding interference + multi-stim to be a rather challenging
combination... I found even Triple Stim 1-Back (one-back, no typo!)
incredibly hard initially, as reflected by my failing scores for the
first couple days and am now performing around 70% on Triple Stim 2-
Back, though I am still struggling to obtain much over 50% in Quad
Stim 1-Back.

I've been playing with multi-stim mode almost exclusively since
September began -- I'll do my 20 minutes of regular DNB "just to be
safe" and then do 1-2 hours per day of multi-stim mode. I am enjoying
playing with this new mode much more than I've enjoyed any of the
other modes in the past. With all the image options enabled, there is
a surprisingly large number of different image combinations when
playing Triple/Quad Stim modes which should hopefully further help
keep the task both interesting and challenging.

I am finding some of the images much more difficult to remember than
others and thus my scores fluctuate a lot more in these modes
depending on what images I've been given. In multi-stim mode, I seem
to be "intuitively unintentionally chunking" (if that makes any
sense?) the positions, yet I am having great trouble associating which
images I saw with their respective positions when playing Triple/Quad
Stim modes -- for example, I will all too often input a match for both
a hexagon and pentagon when the row-column position entries having
images match those row-column position entries having had images N
trials ago, even when the hexagon and the pentagon have swapped
positions. This problem seems to happen much more frequently the
closer in shape the images are to each other and has been particularly
troublesome with rotated/mirror images of the like generated when
tetromino/pentomino modes are enabled. This is something which caused
me much less trouble when playing QNB (even with the same interference
settings), although I'm unsure whether to attribute that to superior
depth than breadth of visuospatial processing, to their being more
different shapes in QNB (eg. the polygon set includes the semi-
circle, triangle, diamond, square, hexagon, pentagon, octagon, star
versus the 2-4 shape subset from that set which will be used in multi-
stim mode), to a combination of the former and latter, or perhaps yet
still to other factors such as a difficulty rapidly and effectively
encoding spatial data and associating that data with positional
locations when multiple visual elements are presented
simultaneously... In complete honesty, I unfortunately lack the
requisite neuroscience background and thus have no bloody clue why
this mode is so much more difficult.

I am curious to see whether this may in the long run afford benefits
to matrix completion and/or other visuospatial tasks requiring rapid
discrimination between a large number of simultaneously presented
visual stimuli. I know we've had the breadth versus depth discussion a
few times before, however this may be the best mode yet to test the
merits of training breadth of visual working memory and see how it
stacks up against traditional low-breadth high-depth n-back
training...

I would imagine it's probably best to use sound + multi-stim to make
it more closely resemble traditional DNB and I plan on adding auditory
stimuli once I start consistently performing well on Quad Stim 2-Back
(hopefully in a couple weeks!). I'm hoping to add dual audio to this
mode eventually, something which I hope will continue to challenge my
auditory working memory while playing multi-stim mode.

[polar]

You know, Reece, I have this feeling, that more than 1 hour of n-
backing a day is a risky enterprise. Intuitively, it's too much of a
big strain in one direction, and physiologically, you could have too
much dopamine activity (you surely know, dopamine is tied to rewards /
dependence...). For example, I guess you probably did your session
before writing this post - am i right? Do you realize, that you use
sentences with around 100 words? I dont know if you have any
experience with medication, but semantics you are using seems to me as
a problem with inhibition (i.e. too much activation). Just a opinion.

[Reece]

Hi Polar,

Yes, I did my sessions (mainly quad stim position n-back which I seem
to find particularly... stimulating) prior to writing this post and
looking back on it, I can certainly see what you mean about it being a
lot wordier than my usual self. The only medication I am currently
taking is Dexedrine (d-amphetamine) for ADHD, which I strongly believe
n-back training has sensitized me to. I can't even take my prescribed
dose of 20mg t.i.d. anymore, it being much too stimulating and causing
many negative side effects I haven't until recently experienced in a
long time such as jaw clenching, anxiety, and thermoregulatory
problems. Thankfully, I've been able to cut back my dose to 15mg
t.i.d., although perhaps I should try going a bit lower still judging
by some of my recent posts, although I will of course hold off on
making any such changes until I get a chance to discuss said changes
with my doctor.

likeprestige, thanks for the laugh! You really have a wicked (I mean
that in a good way) sense of humour :-)

Polar mentioned rewards dependence -- it certainly seems to be my
experience that playing longer sessions today reinforces playing
longer sessions tomorrow (at the expense of less rewarding things
like... college homework).

[milestones]

Well, I am a proponent of trying to reach the highest N level
possible...one should do this starting with DNB. But, after that is
reached (or nearly reached), I think it's a good idea to move on to
challenge oneself with other variants (as others in this group have
exhorted). Personally I have now started playing Dual Crab-back on BW
4.8 and after 20 sessions I am at Crab 5 back. I hope to reach
slightly higher N before adding stims to this variant. I think moving
from, say, DNB right to Quad may not be the optimal approach; I
believe it is better move from DNB to Triple N back and then aim to
reach as high an N level on that as possible before moving on to quad
back. Conversely, continuing to practice a tried and true variant
(i.e.DNB) "to be safe" while also moving to other variants and/or
adding modalities is just erring on the side of too much caution.
Additionally, as Polar indicated, it's putting in way too much
training time with most likely little to no return. Others, however,
might disagree.

So imo, two things to guard against while N backing over an extended
period of time: 1) A low N level with too many stims and 2) a very
high N level with minimal stims. The tricky thing is that "low" and
"high" are relative to the user and one moves on according to their
own judgement of when it's best to add more stims or try another
variant of DNB.

[Arkanj3l]

I'm thinking that the only two reasons for this being the case is to
tackle a tendency to form crystallized strategies in DNB as well as
stretching EF to its limits, which can be done in other ways. Other
than that, if we're just setting out to improve WM, why try the other
variants?

[Reece]

Different modes like crab, variable, combination, and multi-stim might
bring about different benefits which cannot be obtained or are
obtained to a lesser extent than through regular DNB even if DNB is
played at a higher n-back level. Reverse digit span is more g-loading
than forward digit span despite the tasks being identical other than
the order in which the digits are to be recalled. Variable n-back
modes require much more manipulation of working memory than
traditional DNB, while combination modes require one to make
associations between stimuli.

[Arkanj3l]

So instead of just trying to increase the WM load, it's the overall g-
load that we're trying to increase. OK.

So reece, do you play Quad-Crab-Variable-Combo-N-Back? :p

[Reece]

I know there are quite a few DNBackers with ADD(ADHD-PI) / ADHD, many
older DNBers that want to preserve/improve their memory, many students
hoping to increase their GPA and/or cut down on study time, and some
extremely intelligent DNBers who are interested in increasing their
WMC, executive function, IQ/g, and certainly others with different
goals in mind. It's quite an accomplishment that a single program
(Brain Workshop) has developed so many variants to cater to such a
broad range of people trying to improve themselves.

I find combo modes extremely difficult -- it seems I have quite a bit
of trouble making rapid associations... It's something I shall
definitely work on in the future, as this ability is highly correlated
with g.

[Arkanj3l]

Do you do speeded n-back training? It might help with the
associations.

[A.Four.Sigma]

That's true, but I have the same issues with speed of combination/
variable nback. I only started a few days ago, but it becomes
intuitive the more time you spend doing it. I set all of my sessions
for 20 minutes and quit when I have grown tired, then start another 20
minute session after a break. By spending such large amounts of time
with the training, it becomes more intuitive. I expect that by the end
of this week, combo-variable nback should be more or less completely
intuitive (by virtue of continuous daily drilling), even though today
I experience great difficulties with it.

[Arkanj3l]

I don't think that change in one's latent inhibition is a neutral
thing, but I think that if it's good or bad varies on the context of
what one wants to get out of training. In addition, I'm not sure if
it's affected by n-back at all; latent inhibition physiologically does
have something to do with dopamine receptors, but it's also a matter
of which ones. As far as we know, DNB increases the amount of D1
receptors in the PFC, with a couple of other changes -- but I haven't
seen anything about changes in, say, the thalamus, which in my
understanding is like the immigration office for stimuli.
Schizophrenics, who have extremely low latent inhibition, have fewer
(or more?) D2 receptors in the thalamus, and N-Back doesn't seem to
affect these things (may be good or bad depending on your
interpretation). I haven't taken a close look at the McNab study on
latent inhibition to know exactly what kind of dopamine activity is
going on; is it uploaded in our files?

My questions are these: by doing this training, although you
experience consistent attentional shifts, has your working memory
capacity suffered? Do you experience more mental noise, or less? If
the ideal situation (at least in my opinion) is medium-low latent
inhibition + High IQ, and it turns out that WM is NOT affected by this
experiment of yours (doubtful, perhaps, due to the dependence of focus
for WM... of course if you can switch between "distractable" and "non-
distractable" modes of perceiving, that would be INCREDIBLY powerful
IMO), then practicing and training with both so called "depth" and
"breadth" (I think this metaphor is more appropriate in this
situation) is a major breakthrough.

We'll be GODS, man! :p

On Oct 4, 4:49 pm, likeprestige <plastic...@live.com.au> wrote:
> Hey,
>
> I've been playing "Quadruple-stim, position, colour, sound, sound 2
> (piano) 1-back" for the last 4 days, each session lasting 1 hour. On
> average I can currently manage 50% accuracy each session. However, I
> am not aware of what the potential consequences could be as a result
> of undertaking such a mode for this extended period of time. Will the
> consequences be positive or negative? Not completely surhttps://mail.google.com/mail/?shva=1#search/managecas/128aeecf392903b8e.
>
> Some speculations:
>
> Negative:
>
> 1. Attention problems
>
> - Explanation = constantly shifting focus of attention. Reminds me of
> the implications involved when watching too much television.
>
> Positive:
>
> 1. Improvement in visual pattern recognition/visual abstract reasoning
> processes.
>
> Neutral:
>
> 1. Change in one's level of "latent inhibition"
>
> - Explanation = Similar to the reason listed above. I've acknowledged
> it as a neutral consequence because I suppose such a change could be
> perceived as negative/positive, just depends on the activity that is
> being performed.
>
> ------>  I think it's important to look at the other side of the coin
> as opposed to always voicing the potential positives derived from
> training with particular variants.
>
> If anyone has any opinion on this or wishes to develop on what I have
> said, I strongly encourage you to do so.
>
> Thanks,
>
> likeprestige

[Arkanj3l]

Overall, though, I don't know enough about LI in order to make an
adequate analysis. It could be areas other than the thalamus also
responsible for LI AFAIK.

[αrgumziΩ]

LI does not change based on any form of training, as training has no
influence on the mesolimbic pathway. The only things that affect LI
are estrogen (by increasing it) and drugs that deal with psychosis
(i.e., that either promote or inhibit its occurrence, like
antipsychotics on one had and meth on the other).

(However, this is based on current data. Later research is most likely
to show decreased LI post-training.)

Besides, I seriously doubt attentional shifting, if one could infer
such a thing based on NB-training (per likeprestige), is a negative
thing; watching T.V. isn't remotely equivalent to NB-training either.
By my lights, I very rarely feel smarter after watching T.V.

argumzio

[Arkanj3l]

Yeah, I've been looking for a while on how to increase dopamine
activity in the mesolimbic pathway. The only thing I've found is a
study that showed rats with polyunsaturated fat deficiency showed
higher dopamine activity in the mesolimbic pathway, while lower
activity in the mesocortical pathway (imagine trying to figure out the
LI of rats :P). The only thing I can garner from that finding is in
order to improve LI, eat less fish.

Here it is:
http://www.ajcn.org/cgi/content/full/75/4/662

[milestones]

This is sort of an unusual topic. You might have a look at Steven
Johnson's "Everything Bad is Good for you" -- for a small book I found
it to be a pretty good argument in favor of television and video games
as that which has made people more intelligent over the years.

(By have a look, I mean scope out the Amazon Reviews).

http://www.amazon.com/Everything-Bad-Good-You-Actually/dp/1573223077.

His position appears to be contrarian against the "entertainment is n/
dumb" complaints, but he's mainly comparing television in 2000's
against its own history -- recent shows like 24 and the Sopranos that
have multiple plot lines versus the mono and often silly plots from
the 70's such as Starsky and Hutch and Threes Company. He does the
right thing comparing apples with apples and doesn't make claims that
television now has to be taken seriously. His point is simple and
pretty hard to refute: A mass audience has been manuevered (by dint of
sly corporate greed for back end sales) into more and more complex
story lines over tv's history. This is self evident but it's not an
argument that TV is a direct form of WM training or IQ
improvement...rather that complexity over time has had played a part
in the "Flynn effect".

One thing to note is that this book could just as well been a lengthy
blog entry on the internet...even though there are some footnotes.
It's taking a common assumption and then spinning it around in plain
sight in the style of Malcolm Gladwell. As for more weighty and
empirical fare that's been written on this topic, I have no idea....

[Reece]

Congrats likeprestige!

That really is incredible (you really have to try it to fully
appreciate the difficulty) !

I don't unfortunately yet have the skill yet to integrate multistim +
dual sound modes, though I have been having some fun making dual sound
mode harder by reversing which sounds go to the left/right ear after
each session using piano and word options and occasionally still use
2x piano when I feel like having an extra challenge. Once I get a bit
better with double stim Dual N-Back, I think I may replace the basic
DNB training I'm still doing each day by that -- I'm still struggling
with Double Stim D3B though and would like to ideally get at least to
decent scores on N=4 before I try that. Otherwise, I train without
sound using Triple Stim Position mode (N=2) -- I'd love to try Quad
some more but I find it too demanding, frustrating (my scores haha),
and surprisingly tiring -- reminds me of when I first jumped from
playing DNB about 15 months or so ago to playing QNB having never
tried TNB.

[Reece]

As to latent inhibition... Not objective in any way unfortunately,
however I do feel myself being "more creative" and having an easier
time associating things. I realize this is not necessarily a good
thing and almost got me in trouble yesterday when I started laughing
in my intro cognitive neuroscience class when my professor was talking
about Broca's aphasia and his patient "Tan" -- I asked myself "If
this were discovered today, would we have thought that Tan was having
an identity crisis and thought he was a Pokemon and then proceeded to
wonder what kind of Pokemon a Tan might be. I suppose this may have
been primed by the covering of material on Greebles earlier in the
lecture, though there was never any mention of Pokemon and it's been
10+ years since I've played a Pokemon game.
On Oct 20, 2:33 am, likeprestige <plastic...@live.com.au> wrote:
> A ___speculation___ on the potential implications of continuous,
> lengthy sessions of "Quadruple-stim, position, __image__, sound,
> sound2 (piano)".
>
> http://en.wikipedia.org/wiki/Latent_inhibition#Low_latent_inhibition
>
> Hmm... I am not informed enough to know where to stand on this.

[Arkanj3l]

It's not necessarily low latent inhibition, although associations do
seem to be a part of it. If your life is starting to feel more
"noisy", sensational-wise, then that might be closer to what we're
looking for. When we're talking about creative geniuses with Low-LI
and a high WMC, well, wow... just imagining it makes my head hurt.
Being able to have that kind of explicit recall amongst all the
supposed data noise almost scares me.

In addition, this anecdote seems like a pretty one-off thing, so as
evidence it's kind of hard to ratify. But, hey, I believe you.

By the way here's a cool article on personal biases:
http://www.skepdic.com/essays/evaluatingexperience.html

[A.Four.Sigma]

He's basically reporting the same thing as the rest of us (who noticed
effects). Any word on unique benefits of multi-stim vs other modes
yet??

[Arkanj3l]

Acknowledged. :P

On Oct 21, 7:45 pm, likeprestige <plastic...@live.com.au> wrote:
> Hey,
>

> Some of you seem to be misinterpreting what I said. The only thing
> that I admitted was my ___current average score___ on quadruple-stim,
> position, __colour__, sound, sound2 (piano), __not__ the __effects__
> that it has had in my daily life so far. This point needs to be
> understood very clearly. I only speculated what ____may___ be a
> consequence as a result of playing on this mode for extended sittings
> and days.
>
> Please acknowledge this.
>
> likeprestige

[A.Four.Sigma]

Can this be right? Quadruple Stim has 10 simultaneous stimuli? All of
my keys are mapped to a different stimulus from A, all the way to the
semi-colon. Is this what you are using?

On Oct 21, 6:45 am, likeprestige <plastic...@live.com.au> wrote:
> Hey,
>
> Some of you seem to be misinterpreting what I said. The only thing
> that I admitted was my ___current average score___ on quadruple-stim,
> position, __colour__, sound, sound2 (piano), __not__ the __effects__
> that it has had in my daily life so far. This point needs to be
> understood very clearly. I only speculated what ____may___ be a
> consequence as a result of playing on this mode for extended sittings
> and days.
>
> Please acknowledge this.
>
> likeprestige
>
> On Oct 21, 10:37 pm, "A.Four.Sigma" <davidsky...@gmail.com> wrote:
>

[A.Four.Sigma]

Finally figured it out, now there are 6 stimuli, which is more
manageable. It feels like my head will explode when trying to memorize
all of these. It's my bet that this will do something for pattern
recognition; it seems much different from any of the other modes... I
agree with Reece, this mode is good fun and extremely challenging..

[Jelani Sims]

I agree :) way to fun for something so beneficial.

On Oct 21, 2010 4:29 PM, "A.Four.Sigma" <david...@gmail.com> wrote:

Finally figured it out, now there are 6 stimuli, which is more
manageable. It feels like my head will explode when trying to memorize
all of these. It's my bet that this will do something for pattern
recognition; it seems much different from any of the other modes... I
agree with Reece, this mode is good fun and extremely challenging..

On Oct 21, 3:12 pm, "A.Four.Sigma" <davidsky...@gmail.com> wrote:

> Can this be right? Quadruple St...

[A.Four.Sigma]

I remember wondering if Quad 2 back was possible, but now Multistim
Quad 2back...! very interested to know if it is possible to do well
there, and eventually make it to 3 back.

On Oct 21, 3:38 pm, Jelani Sims <gouki...@gmail.com> wrote:
> I agree :) way to fun for something so beneficial.
>

[A.Four.Sigma]

Fullscreen mode is a great idea for maintaining focus against the
onslaught of Facebook. Will be implementing that.

A4S

On Oct 22, 1:43 am, likeprestige <plastic...@live.com.au> wrote:
> Not that it matters, but just for accuracy...
>
> "Interference -
> 1. Alter default 12.5% and increase to 70%."
>
> Not 70%, rather 75%**, lol.
>
> likeprestige

>
> On Oct 22, 4:47 pm, likeprestige <plastic...@live.com.au> wrote:
>
> > Hey,
>

> > Just for reference:
>
> > quadruple-stim, position, __colour__, sound, sound2 (piano)  __1-
> > back__            =
>
> > 4 position
> > 4 colours
> > 2 sound
>
> > = 10 stimuli = 10 different keys
>
> > Other modifications I make -->
>
> > Time -
> > 1. Sec/trial = default setting on this mode is 4.50sec. I always
> > change it to 3.0 sec. The default setting for traditional dual-n-back
> > 2. Trials = I change it to 35+1  --> which as a result modifies total
> > session time
> > 3. Session = 108 sec
>
> > Images -
> > 1. Cartoon faces and National park service images are turned off. All
> > other images remain turned on.
>
> > Sound -
> > 1. Letters turned off
> > 2. Piano turned on (this is the only auditory setting I maintain)
>
> > Interference -
> > 1. Alter default 12.5% and increase to 70%.
>
> > Config. file -
> > 1. Full-screen mode = Turn on
> > - This makes for easier game play.
>
> > Finally, not sure how long it will take me to get to _2-back_
> > (currently avg. 60% on n-1-back), but I look forward to when I do.
>
> > likeprestige

[αrgvmziΩ - σ>V]

It is objective, as such experiences have been seen time and time
again. DNB increases activation in the Dorsolateral Prefrontal Cortex,
which is more active than average in individuals with a high degree of
"openness to experience" in the Five-Factor Model of personality.
While there is a hereditary component, there is also an environmental
one which WM training may influence to some extent. However, it is
mistaken to view that as "creativity" per se, as (taking notes from
Paul Cooijmans) creativity is better understood as a synergy of
conscientiousness, associative horizon, and intelligence, all
interdependent on each other's respective range – a bona fide triadic
relation of the human kind!

More info here: http://www.paulcooijmans.com/genius/

And as for the DPFC and WM connection: http://www.jneurosci.org/cgi/content/full/27/8/1801
(a drop in the bucket)

As for myself, I see all sorts of associations, even before DNB, but
they increased in their fluency after training and are much more
precise in their relata.

argumzio

[Reece]

Thank you for the very insightful post αrgvmziΩ - σ>V. Time for me to
go read those links :-)

[αrgvmziΩ - σ>V]

Reece et al.,

You may be interested in Cooijmans' Personality Scales for Intelligent
Adults (which can be taken for $25USD, or free if you're a Glia
Society member).

This test, which I've taken, is extremely informative, much more than
the Five-Factor Model. It brings to bear certain personality features
that also further contribute to creativity (besides intelligence):
http://www.paulcooijmans.com/personalitytests/psia_scales.html

What I'd like to emphasize here is according to his analysis Deviance
is negatively correlated with Intelligence; the higher your
intelligence, the more rare it is for you to be deviant. In other
words, intelligence disposes for "normality", and genius (creativity
incarnate) becomes exceptionally rare for high intelligence levels.

argumzio


On Nov 1, 2:50 pm, Reece <rockyb...@hotmail.com> wrote:
> Thank you for the very insightful post árgvmziÙ - ó>V. Time for me to

[A.Four.Sigma]

Argumzio,

I also read some of Coojman's work. I started another post on various
intellectual profiles, partly due to questions brought up by his
research. Firstly, I'm wondering what STDev your IQ score was measured
at, and also your age, and hobbies. I am interested because of the
unique vocabulary that you use, and other reasons. My profile is
0sigma nonverbal, 4th sigma verbal, age 23. hobbies revolve around
learning, probably pertinent is that I keep social interaction to a
bare minimum because I find it fruitless in most cases). Do you share
any commonalities?

Next, is this second post of yours saying that genius requires a
degree of deviance, a trait which actually decreases with
intelligence?

[Reece]

Deviance doesn't decrease IQ. When we look at the population as a
whole, geniuses are so rare that no matter what traits may or may not
be correlated (be they positively or negatively) with genius, we are
highly unlikely to observe them when looking at the population as a
whole. We simply cannot look at population data and make the
conclusion that deviance, conscientiousness, etc are correlated with
IQ (or lack thereof) in the high IQ because this holds in "normal"
people. An IQ test with sigma = 15 has 97.5% of the results <= 130, so
regardless of whether we set the bar very low at 130 (I think only
Mensans consider 130 to be genius - ha!) or choose to use a much
higher number, I'm sure you can see why we can't rely on population
data -- at least not without throwing away 97.5%+ of it.

Googling "gifted bored" (no quotes) or search queries along those
lines should turn up ample anecdotal evidence that gifted students are
reported to be more likely to have "an attitude problem". We can
phrase that in a positive/negative light depending on who we decide to
listen to (eg. parent vs. teacher), however I don't think it's too
hard to imagine a highly intelligent child causing problems if he's
bored out of his mind or is fed up with his "ignorant" (at least when
compared to him) pupils and teacher.

I was by no means a genius, however I was scoring at a ninth grade
level in English when I was in grade 4, so I'd like to think on some
level I can appreciate why some gifted students choose to rebel and I
certainly wish I would have had the balls to do that as a kid (maybe
they could have found me some more challenging work to do if I had)...

[Samples]

Has anyone tried removing the auditory modality and instead just tried
2 position modalities?

My experience = Damn tough!! I can only barely get to 3-back on this
setting. It requires an enormous amount concentration because of the
interference chaos!

This mode is teaching me how short my attention span really is!!

Sameil Pleirz

[Reece]

Yes -- a MUCH harder version of "dual" n-back :-)

I have a relatively easy time remembering which positions had
stimuli, however I often struggle to remember which color/image goes
with which position (all that interference certainly doesn't help!!).