Training reading comprehension via complex working memory: Any CWM games out there?

[jttoto]

As some of you may know, the literature on training reading
comprehension is prolific, yet for the most part unsupported by
published studies and the scientific community. A study may have
stumbled upon a way to improve reading comprehension is by practicing
CWM: http://www.gwern.net/docs/chein2010.pdf

I do a lot of reading in my job, but my inquiry is not only helpful
for me. The practical benefit in improving RC is immense, with
schooling, the professional world, and standardized tests for higher
education requiring that we read efficiently and accurately.

A brief description of CWM is described above: "The hallmark
characteristic of CWM tasks is the interweaving of
storage (of the test items) and processing (the secondary decision-
making
task) components" I've search thoroughly throughout the web, yet the
closest thing I could find is Arithmetic n-back.

likeprestige has said in the past that arithmetic n-back mimics CWM
perfectly, which I mostly agree with. However, due to the minute
differences An-back and the regimen used by Chein, I am looking for
other CWM games out there.

[likeprestige]

"likeprestige has said in the past that arithmetic n-back mimics CWM
perfectly, which I mostly agree with"

Never said perfectly and I don't think it does mimic it 'perfectly' (I
think it resembles the same sort of qualities, that is all), it really
depends on what CWM task your comparing it too. Arithmetic-n-back is
similar but I think crab-arithmetic-n-back is more-so, regardless of
the outcome it does not mean that either mode is a better or worse
training vehicle in comparison to traditional CWM tasks.

[genvirO]

If anything I think Triple Arithmetic _Variable_ n-back (3sec per
interval, high interference e.g. lure trials at 60% or above) would be
most effective out of the available options that BW has to offer.

:-)

[genvirO]

Sorry, referring to Dual Arithmetic _Variable_ n-back, not triple.

[genvirO]

Check this out:
http://www.psy.unibe.ch/unibe/philhuman/psy/apn/content/e5616/e5621/e7836/e7942/files7943/Loosli_et_al_2011_ger.pdf

Bear in mind this article has already been discussed, however, it's
relevant to this topic.

[Jonathan Toomim]

BTW, I expect to be implementing the Chien CWM training paradigm in
Brain Workshop sometime within the next month or two.

[☉]

Exciting news!

argumzio

[Pontus Granström]

God bless you ;)

--
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-training@googlegroups.com.
To unsubscribe from this group, send email to brain-training+unsubscribe@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/brain-training?hl=en.

[genvirO]

Fantastic!

On Aug 17, 3:27 am, Jonathan Toomim <jtoo...@jtoomim.org> wrote:
> BTW, I expect to be implementing the Chien CWM training paradigm in
> Brain Workshop sometime within the next month or two.
>
> On 8/16/2011 2:56 AM, genvirO wrote:
>
>
>
>
>
>
>
> > Check this out:

> >http://www.psy.unibe.ch/unibe/philhuman/psy/apn/content/e5616/e5621/e...

[genvirO]

I just got my UMAT (Undergraduate Medical Admissions Test) results
back and anyway, reading comprehension was what let me down here big
time. I want to say that although my overall percentile wasn't
terrific, I 'feel' that my n-backing over the years has contributed a
lot to my performance, especially in respect to the non-verbal
reasoning section (see score below).

In high school I really struggled and as a result I didn't obtain a
very good final result, which meant that I had to do a preparation
program to get into University initially (bear in mind however,
although it was a childhood dream to become a doctor, based on my
academic career and as a result what people generally thought about my
potential in this respect (mostly parents - rhetorical question = "Are
you smart enough?" they'd say = killed my confidence), I certainly
held no delusions about "going to medical school"). During the course
I naturally encountered the same learning difficulties, however, I
pretty much commenced the course in-line with the time in which I
started n-backing, and within a few months I progressively got better
at 'learning' to the point where I completed the program with modest
marks which enabled me to get into science, where my ability to learn
continued to improve.

(TIME GAP - I commenced science, decided it wasn't for me and then
changed to engineering, decided that wasn't for me and then I went
back to science and here I am today!).

Anyway, I won't drag it on. Long story short, (after a few years past)
medical school is now starting to become a distinct probability
(reality) as opposed to something that I could once only read in story
books (fantasy); but I've still got a lot more work to do
obviously :-). I suppose I'll write the 'essay' of how much I 'owe' to
brain workshop once I get into medical school :-)

People may think that I'm eager to establish causality here, but I'm
not trying to do that, my purpose here is only to demonstrate some
level of appreciation and not anything more than this.

Overall percentile = about 58%, which means that I performed equal to,
or better than, that of 58% of candidates.

1. Logical rasoning and Problem solving (involves a lot of reading
comprehension - recalling and manipulation of verbal info.)= better
than approximately 46% of people

2. Understanding people (involves reading a bunch of paragraphs and
discerning the behavior or 'emotions' of people) = better than
approximately 37% of people

3. Non-verbal reasoning = better than - **__90%__** - of people

There is only one undergraduate medical school I'm interested in
(Australia), so if I don't obtain entry (highly unlikely based on my
overall percentile) it's either "better luck next year" or I look
towards graduate entry. I'm already studying medical science as an
undergraduate but my main concern here is "time". If I obtained entry
into the medical school I want to go to next yr then I could have it
all wrapped up in a bit over 5 yrs or 6 yrs if I resit the test next
yr and so on, but If I went graduate entry, the whole ordeal would
take at least 7 yrs & that's only in the event that I obtain entry
into the school I want to get into first try.

Anyway, enough of my "life story", it's not exactly the place for it!
It's funny that I've been posting here for over two years now. A lot
of things in life hardly last that long (at least for me).

Peace out!


P.S - As one can imagine, I'm intrigued to say the least about how the
Chien CWM training paradigm is being implemented soon.

[Pontus Granström]

You got any idea of it's G-load?

--
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.

[Pontus Granström]

Results:

Mean overall UMAT score at entry was 60/100 and mean GPA during university study was 6.1 (range, 1–7), with a correlation coefficient of 0.15 (P = 0.005). This relationship existed only in the first year of university study. For UMAT Section 1 score, the correlation coefficient was 0.14 (P = 0.01); for UMAT Section 2, the correlation coefficient was 0.06 (P = 0.29); and for UMAT Section 3, the correlation coefficient was 0.09 (P = 0.11). UMAT overall score for men (60.2) and women (59.8), and GPA for men (6.1) and women (6.2) were similar. However, men performed better in Section 1 (mean score 61.6 v 61; P = 0.05) and Section 3 (63.2 v 60.7; P < 0.001), whereas women performed better in Section 2 (58.5 v 55.8; P = 0.009). In multivariate analysis, only correlation between GPA and UMAT Section 1 score remained significant but was weak and lasted for 1 year of university study.

Conclusions:

Our findings suggest that UMAT has limited predictive validity for academic performance.

http://www.mja.com.au/public/issues/194_07_040411/wil11056_fm.html

Similar to the Swedish SAT, only got a 0.4 G-load, still a lot of people think it's a measure of intelligence.

[genvirO]

I can't say objectively, but it's by far the hardest test I've ever
taken, in so much as to say that you have to complete a lot of
questions in a limited amount of time (this may affect its 'g' load).
I would rate the non-verbal reasoning section alone as being more
difficult than other non-verbal reasoning tests I've taken such as the
JCTI and the RAPM, regardless of how much similarity can be found
here.

Section 1 - Logical rasoning and Problem solving = 44 questions in 65
minutes
Section 2 - Understanding people = 40 questions in 50 minutes
Section 3 - Non-verbal reasoning = 38 questions in 50 minutes

Total testing time = 2 hours and 45 minutes

NOTE: There are no breaks in between each section. There is a maximum
of only 15 seconds spent to accommodate for transitioning between each
section. The only time one is permitted to say leave for a toilet
break or for anything else is during the testing period.

If you like I can send you (or anyone else that's interested) some
practice tests (very similar to actual test) so you can try and gauge
it yourself.

[genvirO]

Thanks!

It's important to note that the study you cite is in reference to the
correlation between academic performance and UMAT performance, not
necessarily between psychometric measures of intelligence and the
UMAT, which I suspect would be much higher, regardless of what
correlation exists with the former.

"Similar to the Swedish SAT, only got a 0.4 G-load, still a lot of
people
think it's a measure of intelligence. "

I wouldn't compare the SAT with the UMAT however, because the UMAT
does focus on trying to assess fluid abilities and not crystallized
intelligence (although, if one were a veracious reader it would
probably compliment their performance in SEC 1 and 2) which is
generally the case for the SAT.

Once again, my scores were percentages. So in general, I am happy
where I placed considering that I'm more than likely, on average,
competing against people that are at least a standard deviation above
the norm when it comes to the psychometric understanding of
intelligence.

On Sep 20, 3:43 pm, Pontus Granström <lepon...@gmail.com> wrote:
> Results:
>
> Mean overall UMAT score at entry was 60/100 and mean GPA during university
> study was 6.1 (range, 1–7), with a correlation coefficient of 0.15 (P =
> 0.005). This relationship existed only in the first year of university
> study. For UMAT Section 1 score, the correlation coefficient was 0.14 (P =
> 0.01); for UMAT Section 2, the correlation coefficient was 0.06 (P = 0.29);
> and for UMAT Section 3, the correlation coefficient was 0.09 (P = 0.11).
> UMAT overall score for men (60.2) and women (59.8), and GPA for men (6.1)
> and women (6.2) were similar. However, men performed better in Section
> 1 (mean score 61.6 v 61; P = 0.05) and Section 3 (63.2 v 60.7; P < 0.001),
> whereas women performed better in Section 2 (58.5 v 55.8; P = 0.009). In
> multivariate analysis, only correlation between GPA and UMAT Section 1 score
> remained significant but was weak and lasted for 1 year of university study.
> Conclusions:
>

> *Our findings suggest that UMAT has limited predictive validity for academic
> performance.*http://www.mja.com.au/public/issues/194_07_040411/wil11056_fm.html

>
> Similar to the Swedish SAT, only got a 0.4 G-load, still a lot of people
> think it's a measure of intelligence.
>

> On Tue, Sep 20, 2011 at 7:34 AM, Pontus Granström <lepon...@gmail.com>wrote:
>
>
>
>
>
>
>
> > You got any idea of it's G-load?
>

[Jonathan Toomim]

Hmm... I guess I did say a month ago that I'd do it in a month or so,
didn't I.

[Pontus Granström]

Yes I saw that, it's a test for people with above average intelligence. Still it might have a low g-load hence your score might not be indicative of your intelligence. For example a test that
measures "how you understand" other people might be very flawed. Reading comprehension and word knowledge shows low g-loads hence unintelligent people will score high on it.

[dualnback]

Jonathan, yes you did. We're waiting on it.:)

[Pontus Granström]

From a UMAT report.

The literature indicates that such measures of reasoning are indicative of general or ‘fluid’
intelligence. There is also a close association between general reasoning ability and language
ability, possibly mediated via the working memory. The Logical Reasoning component has
much in common with Critical Thinking. Both these constructs and problem solving are
generally regarded as important attributes in the education of professionals.

[genvirO]

Thanks.

Yeah, I think you may have obtained that excerpt from here =
http://umat.acer.edu.au/files/UMAT-Executive-Summary.pdf

Looks like have some work to do! I think it's plausible to move from
47 percentile to anywhere above the 70 percentile with some hard work.
I do have 12 months to work towards it after all!!
Some things I could do to improve here are -
- verbal WM exercises, which I guess will continue to be variable-
arithmatic n-back, and hopefully in the near future, Chien CWM
training.
- Critical thinking, which I guess could involve practicing informal/
formal logic.
- Not only read a lot but attempt to attempt to analyze what I'm
reading at any opportunity.

In regards to section 2 - Understanding people - I still haven't
devised a strategy as to how I may perhaps improve here.
Some obvious things that can be practiced are the following, as to how
effective they may be though I'm not sure (better than nothing (?)).
- Practice analyzing not only others feelings but my own
- Meditation --> Specifically, compassion meditative practice (I've
read some interesting articles that have connected training on such
with increases in brain related areas)

Section 3? Well, I guess I didn't score in the 100 percentile (scored
90 percentile), so I guess I have some improvement here too, however,
it's nice to realize that I don't have to focus my attention as much
here.

:-)

"Hmm... I guess I did say a month ago that I'd do it in a month or so,
didn't I. "

No pressure! Just appreciative of the fact that your even considering
it, which is great!

On Sep 21, 6:31 pm, Pontus Granström <lepon...@gmail.com> wrote:
> From a UMAT report.
>
> The literature indicates that such measures of reasoning are indicative of
> general or ‘fluid’

> intelligence. There is also a *close association between general reasoning
> ability and language*
> ability, possibly* mediated via the working memory.* The Logical Reasoning

[Pontus Granström]

I hear a qualifying score is around the 60th percentile, your 90th percentile score is the of course very good. But of course to defend the honor of n-back you will get a 100:ed percentile score. ;)

[whoisbambam]

i certainly hope u realize that u are smart enough to go to med
school.

i worked alongside med students from loma linda university medical
center for years.

based on my experience here in the forum, in my opinion, it is within
ur grasp to be accepted to their medical school, at least.

of course, i also worked with many doing their residency and
internship from around the country, and likewise, u seem to measure up
well enough.

just my 2 cents.

[J.]

@Jonathan

I just wanted to throw it out there that I'm also looking forward to
you coding this task. Don't want to "add pressure" or make you feel
rushed. Just wanted to say thanks!

[whoisbambam]

i wonder if some of us regulars made some fresh donations if it would
make a difference in jonathan's motivation to really implement chein
complex working memory, as used in the studies, into some sort of
program? or just make the program and sell it, licensed.

i definitly have some mgT i could donate.

but i cant really perceive any different taking about 5gms a day, so
cash is probly the better deal

[J.]

I do a brief prayer that this will be coded every morning before I
check this google group... ^_^

[Alfredo G. J. O.]

Any Update Jonathan? :P

[Green]

1.) Brainturk.com has one, but it only goes up to 5 items.

2.) Someone aliased "Brandon Milhouland" (sp?) posted a link to a downloadable one of his own construction on this forum months back. However, it does not have the verbal component found in the Chein study it only covers the visual-spatial component ( google "CWM" + "curses" to find the archived thread with links).

[Green]

This is really excellent. The time-out is paced fast enough that it forces me to move faster than I feel comfortable moving, but not so fast that I am frustrated.
The multiple levels of symmetry in the spatial version is a great twist.
I played for 40 minutes today, I got up to N=9 for verbal, and N=4 for spatial. Really enjoyed it.
Really great game. Thank you!

[rgpddt]

For some reason my replies get deleted.

[☉]

I also replied (thru the "new" Google groups, by a one off exception),
and it seems to be a case of fixing what ain't broke.

Cool tasks. The verbal is easier than the spatial. I needed only 14
trials to reach level 8 in verbal, whereas I needed 48 to reach lv5
and 16 more to reach lv6 in spatial. This is clearly du to the
differential requirements in the respective recall portions.

argumzio

[Brandon Woodson]

I had fun. Leaving, my mind felt well exercised.

@argumzio: Heh, my experience was totally opposite. I got to level 5+ spatial at just over 20 rounds, but it took ~35 rounds to get to level 8 verbal. Intuition would have me believe that the opposite (easier verbal than spatial) should be true, so maybe my brain is broken.

P.S. After several failed attempts to post on Google Groups where my entire post was deleted while I was typing, when I was done typing and tried to post, etc., periodic right-click-select-all-and-copy's have become my best friend.

-Brandon

[rgpddt]

I wonder what would be the effect of making the game increasingly complex (not necessarily in terms of WM). For example, the game could randomly ask for asymmetricity, or to perform some simple mental rotations. Thus making it a sort of collection of known effective brain training games.

[foobeer]

It's pretty nice, simple to use and well done, sadly it's online. :)
I am also having troubles in the spatial task, with my usual
remembering/thinking-style I barly get one right, except it's right in
the middle or in an edge.
(Ins: I just see you meant the symmetric-task, that's allright for me so far, maybe you should adjust you position to the screen.)
I seem to be heavyly reliant on my active phonological refresh. I also noticed while playing DNB I use some up or down-pitched "hums" to indicate the position of the lit up fields.

(My email-reply didn't get to the list either, so resent.)

Green wrote:
>
> This is really something spectacular. I like the tight time restriction. It
> forces me to move at a slightly faster pace than I am comfortable with.
>
> Having trouble with the spatial symmetry task though - I just do poorly. I
> think I'm blind to one of your symmetry axises.
>
>
>
> On Saturday, October 27, 2012 7:01:26 AM UTC-5, dustcloud wrote:
> >
> > //Copy of my first post with some further clarification - the previous
> > ones seem to have been deleted automatically by the system or by a
> > moderator.
> >
> > I have made CWM game. You can find it here: http://cwmtask.comoj.com (You
> > don't have to click on the link in the main menu. However, it's a link to a
> > www.washingtonpost.com site.)
> > The differences: four axes of symmetry, random number of trials (between 1
> > and 6) and the grid
> > The level is stored in a LSO(flash cookie).

> >
> > Am Freitag, 24. Juni 2011 05:22:32 UTC+2 schrieb jttoto:
> >>
> >> As some of you may know, the literature on training reading
> >> comprehension is prolific, yet for the most part unsupported by
> >> published studies and the scientific community. A study may have
> >> stumbled upon a way to improve reading comprehension is by practicing
> >> CWM: http://www.gwern.net/docs/chein2010.pdf
> >>
> >> I do a lot of reading in my job, but my inquiry is not only helpful
> >> for me. The practical benefit in improving RC is immense, with
> >> schooling, the professional world, and standardized tests for higher
> >> education requiring that we read efficiently and accurately.
> >>
> >> A brief description of CWM is described above: "The hallmark
> >> characteristic of CWM tasks is the interweaving of
> >> storage (of the test items) and processing (the secondary decision-
> >> making
> >> task) components" I've search thoroughly throughout the web, yet the
> >> closest thing I could find is Arithmetic n-back.
> >>
> >> likeprestige has said in the past that arithmetic n-back mimics CWM
> >> perfectly, which I mostly agree with. However, due to the minute
> >> differences An-back and the regimen used by Chein, I am looking for
> >> other CWM games out there.
> >
> >
>

> --
> You received this message because you are subscribed to the Google Groups "Dual N-Back, Brain Training & Intelligence" group.

> To view this discussion on the web visit https://groups.google.com/d/msg/brain-training/-/v0r5_NTn0icJ.

[☉]

I will admit I wasn't working at my best, seeing that it was late at
night when I tried them; it is fairly typical (for me) to do worse at
visual-spatial tasks when fatigued. But I'm curious: what is your
highest n-level for DNB and when was the last time you seriously
trained? I've gotten to 100% @ D8B about a few years ago and didn't
play again save for my going up to n=6 on QNB shortly thereafter.

I would definitely like to see more variants of these kinds of tasks,
though.

argumzio

[Zaraki]

It's certainly more unforgiving than n-back since a single screw-up can send you down a level. Also a bit addicting...

[Brandon Woodson]

I couldn't tell you with much certainty where I'd max out under the best conditions. I started playing n-back about 1 1/2 years ago and only played for a few weeks before image streaming caught my attention and I stopped playing n-back. I learned that it's not typical for people to use their visuo-spatial sketch pad when performing the digit span task a few months later as my score seemed lower than I would have expected, given my supposed I.Q. (estimated based on the TRI52) (I wasn't reciting the numbers in my head auditorily, I was spelling them out as images in my head.); so I practiced using my phonological loop on the digit span task for 10-15 minutes a day for about 3-4 weeks and came back to n-back just to see if there was any carryover (There seemed to be.), and I could play at about n=9 without practice, not perfectly (100%), but well enough that I wasn't booted down a level on standard settings. That was about a year ago, and I haven't played any since. 

QNB 6-back is insane, by the way! And I would like to see hastened progress with the n-back and WM task studies, so we can see which variants (if any) are useful, but maybe it's just me being impatient as usual.

-Brandon