cognitive fun!

Learn your mind. Play it too.
Chunking is impossible!? The combined multimodal n-back task
cognitivefun | 8 years ago Reply Link me
? | 7 years ago Reply
Chunking is impossible!? The combined multimodal n-back task
One caveat in all tasks with well-defined goals is that after enough practice, the brain starts chunking together familiar batches of information, such that spare brainpower can be used for other things. This is often desired: for example, if you are learning to play a piece of music, you would practice until you think in musical phrases instead of individual notes. Then you will be able to think about other things, like varying your tempo, and the emotional expression of the piece.

But for working memory tasks, we don't want to give the brain any chance to become lazy. We can achieve this by adding more things to remember, but if the stimuli are similar to each other, it is still possible to begin to discover patterns within them, and lump them into simpler representations. In terms of n-back tasks, one could always use a high n-level and it certainly wouldn't be easy. But after dozens of trials, does your brain still think it's interesting? Tasks can be very difficult but still lack novelty (long multiplication, for example), and when there isn't novelty, the brain will naturally start wandering off.

To combat this tendency, here is an n-back task variant that aims to provide novelty (and discovery!) in every session, either by generating stimuli randomly, or, in the case of photos, by grabbing from flickr (TM). Up to three simultaneous inputs can be used among about a dozen options; if you have checked out the other tasks already, most of these sounds and images will be familiar.

If you are new to these tasks, you might want to check out the others before you use the respective options --numbers (digit span), words and consonant strings (consonant span), dots (subitizing), and music intervals (interval identification).

Aside of using a 3x3 grid, having 9 possible positions instead of 8, this test is able to reproduce the setup for all the previous n-back variants available on this site. Is there still reason to use the earlier versions? Sure. One, this test is a lot bulkier (around 10x larger) and will take longer to load (and loading sounds and images will take additional time), and more system resources to run. Two, frequently changing settings makes it hard to track progress.

If you choose to frequent this test, it will be a good idea to look around, see what setting you like most, and stick with it for a while (for example, until you can comfortably run 6-back trials).
cognitivefun | 8 years ago Reply
Auditory (Press F to respond): This is the same kind of auditory presentation as in the Jaeggi-Buschkuel variant.
Spatial (J to respond): What was the "visual" input in the J-B variant is categorized as "spatial" in this task. That is, we will be looking for n-back repeats in the location of the visual stimulus on the grid. If you don't select any visual option, the distinction is irrelevant.
Visual (SPACE to respond): If you do select a visual source, however, you will have to respond to n-back repeats according to the appearance of the image that appears in the grid. If you see a fish in the left-top, then, n-trials later, the same fish in the right-bottom, it is still a visual match -- press SPACE!

German consonants: C, G, H, K, P, Q, T, W, as in the J-B variant
/kpst/ phones: /ki/, /ku/, /pa/, /pi/, /sa/, /si/, /ti/, /tu/, as in the /kpst/ variant
Digits: English digits (including "seven"), used in the digit span and PASAT tasks
Music intervals: C-Major dyads of intervals 2 to 8. (match the n-back sounds, not the interval values)

None selected: location chosen at random, no response required
Keep visuals in center: visuals will only appear in the center block, no response required
Respond to spatial location: response required for targets that appear in the same grid location

Colored boxes: eight possible colors: dark gray, light gray, red, blue, green, yellow, purple, teal
Numbers: random 5 digit numbers
Words: up to 10 random 5 letter words from standard English dictionary
Consonant strings: random 5 letter consonant strings, as in the consonant span task
Dotted boxes: 4-7 dots as from the subitizing task
Illustration: 10 illustrations from the picture n-back task
Flickr: load random photos from flickr. If you select this option, after you finish the test session you will get a list of links to the flickr pages of the photos that were used during the test
cognitivefun | 8 years ago Reply
I agree it's a great test! Just wondering, are there any plans to allow adjusting the individual n's? For example, have Auditory on 3 back, Spatial on 4 back, and Visual on 2 back. I think that would allow an entirely new (insane) level of difficulty...
baarslag | 4 years ago Reply
It's a great new test - thanks cognitivefun! It's good to be able to experiment with all different possible tests.

As a further suggestion, how about making it possible to adjust the time delay between each sound/picture? Making the test faster would be one way of making it harder to use strategies (such as repeating the sounds to yourself), and it would be interesting to experiment with this.
TimM | 8 years ago Reply
This is a tremendous improvement over the original dual n-back. Now the brain must constantly adapt...

I can see so many parallels between cognitive training and exercise; the goal is to always force an adaptation with a new stimulus.

Is it possible that depletion of cerebral energetic resources mediates adaptation to training?

Neuropsychologia. 2007 Apr 8;45(7):1571-9.Click here to read Click here to read Links
Cognitive fatigue of executive processes: interaction between interference resolution tasks.
Persson J, Welsh KM, Jonides J, Reuter-Lorenz PA.

Department of Psychology, University of Michigan, East Hall, 530 Church Street, Ann Arbor, MI 48109-1109, USA.

A resource depletion framework motivated a novel strategy for investigating whether the central executive is unitary or separable into relatively independent subprocesses. The idea that tasks with overlapping neural representations may involve similar executive components was also critical to our approach. Of particular interest were tasks requiring resolution of interference among competing representations. Within a single experimental session intensive training reduced the ability to resolve interference on a transfer task if the training task placed high demands on interference resolution. Negative transfer was absent when interference resolution was minimally required by the task, or when the training and transfer tasks did not rely on overlapping neural representations. These results suggest a nonunitary central executive composed of separable subcomponents, at least one of which mediates interference resolution. Our results are consistent with an executive control process specialized for the selection of task-relevant representations from competitors. The results also agree with the view that higher cognitive processes are resource limited and can be temporarily depleted.

Would the body adapt by increasing metabolic supplies?

Would it be a good idea to take creatine before/after training?

Proc Biol Sci. 2003 Oct 22;270(1529):2147-50.Click here to read Click here to read Links
Oral creatine monohydrate supplementation improves brain performance: a double-blind, placebo-controlled, cross-over trial.
Rae C, Digney AL, McEwan SR, Bates TC.

Discipline of Biochemistry, School of Molecular and Microbial Biosciences G08, The University of Sydney, Sydney, NSW 2006, Australia.

Creatine supplementation is in widespread use to enhance sports-fitness performance, and has been trialled successfully in the treatment of neurological, neuromuscular and atherosclerotic disease. Creatine plays a pivotal role in brain energy homeostasis, being a temporal and spatial buffer for cytosolic and mitochondrial pools of the cellular energy currency, adenosine triphosphate and its regulator, adenosine diphosphate. In this work, we tested the hypothesis that oral creatine supplementation (5 g d(-1) for six weeks) would enhance intelligence test scores and working memory performance in 45 young adult, vegetarian subjects in a double-blind, placebo-controlled, cross-over design. Creatine supplementation had a significant positive effect (p < 0.0001) on both working memory (backward digit span) and intelligence (Raven's Advanced Progressive Matrices), both tasks that require speed of processing. These findings underline a dynamic and significant role of brain energy capacity in influencing brain performance.

Questions abound...
studentmedical | 8 years ago Reply
"how about making it possible to adjust the time delay between each sound/picture?"
This is certainly under consideration. I have delayed a speed run for the dual n-back for way too long. Perhaps by the end of the week?
cognitivefun | 8 years ago Reply
what about making the delay between presentations random? like on the flanker test...

...making casually planned rehearsal more difficult while contemporaneously preparing for rehearsal of previous rehearsals (during a longer interval time) to avoid trace decay.

or is a speed run accomplishing essentially the same thing (forcing more rapid rehearsal)?

in everyday living, the interval between salient stimuli to be updated in WM is essentially random
medicalstudent | 8 years ago Reply
My speculation is that the dynamically adjusting intervals, as in the speed runs, primarily affect the cognitive load (temporally, that is, which affects load differently from increasing the n-level).

When one is immersed in a single task, like the n-back, since the goal remains constant, there isn't much competing for attentional resources aside of the recurring stimuli. Unless there is an incentive to respond as quickly as possible (or a disincentive to respond slowly, i.e., a punishment schedule), it may not have as strong an effect as in the tests that stress fast reaction time. Shortening interstimulus delays probably mitigates rehearsal effects strongly enough.

But it could be useful, if random intervals are interspersed with other attention-demanding tasks. This is also the more useful part when comparing to everyday living as you mentioned.

(But in daily life, goals often change before we care to develop cognitive strategies to deal with them. If the phone rings while you're cooking, the goal switches from "watch food" to "answer phone," but after you hang up, unlike the n-back task, where you immediately expect a similar stimulus, the goal is resolved.)

To tackle the issue you brought up, perhaps we'd have to pre-define the goals for a set of tasks, which have different and competing cognitive requirements, and randomly space them. Sort of like a reloaded RST.
cognitivefun | 8 years ago Reply

Login to save scores

© 2008-2012 | about | widgets | blog | cognitive neuroscience for everyone