cognitive fun!

Learn your mind. Play it too.
The spatial memory updating task requires too much processing at once
cognitivefun | 9 years ago Reply Link me
The spatial memory updating task requires too much processing at once
This test doesn't seem like a good adaptation: it's ridiculously hard. It would be more sensible to update just one box at a time, instead of all of them together. But since this version is already out, perhaps "number of concurrent updates" should be left as a user-selected parameter as well...
cognitivefun | 9 years ago Reply
Very nice idea. Personal suggestions in order of preference:

1. Require the user to answer for all used boxes (one after another I suppose). Thus more precise %-right at the end.

2. Make the boxes wrap-around so for example the dot can move "left" even if it's in the left-most column (it would then wrap to the right-most column). Thus complete freedom of movement at each point. As it is now the ending point of the dot is very fixed by probability... For example if it starts in a corner or in the center after 5 iterations it will end up at one of the four mid-edges -- and if you only remember the last arrow or two you have a high chance of selecting correctly. Even if you don't remember how it started, it seems you will have much luck just selecting a mid-edge based on the last arrow movement...

3. Instead of using a 3x3 setup with unused boxes, why not just a row of boxes (as many as the user selected) -- or possibly arranged in a 1x1, 1x2, 2x2, 2x3 etc grid pattern that becomes larger. You could then make them larger.

4. Number of concurrent updates is also a nice idea -- would allow for much larger # of boxes to be used, though I like it as it is as well (strains spatial visualization heavily).

Thank you.
ilikebutts | 9 years ago Reply
"2. Make the boxes wrap-around so for example the dot can move "left" even if it's in the left-most column (it would then wrap to the right-most column). Thus complete freedom of movement at each point. As it is now the ending point of the dot is very fixed by probability... For example if it starts in a corner or in the center after 5 iterations it will end up at one of the four mid-edges -- and if you only remember the last arrow or two you have a high chance of selecting correctly. Even if you don't remember how it started, it seems you will have much luck just selecting a mid-edge based on the last arrow movement..."
I thought about this too, but the "wrapping around" is processed differently in memory, so I decided against it.

It does seem though, that this setup is causing many more problems than fulfilling its mission to stress a particular cognitive task.
cognitivefun | 9 years ago Reply
What about adding an option to slow it down a little? Some of us (ahem, me) have a hard time updating at such a rate...
bovinebrain | 9 years ago Reply
Did you take off the test?! or are you just updating?

I actually liked it and was looking forward to get to three boxes. You shouldn't have taken it off.. If it gets really tough, you can make it slower like you did for the others.. Only bad thing was the size of the initial dot

Imagine having n-back for this! It'd be a good idea for 2 boxes atleast i think.
sygenator | 9 years ago Reply
Apologies for the confusion. The test is still there, just not listed. You can still access it by going to test/23. I need to add an option to make the difficulty adjustment more gradual. 2 boxes to 3 boxes is a pretty big jump!
cognitivefun | 9 years ago Reply
and persistence for the initial dots. So many times i wasn't paying attention and the test had proceeded.
sygenator | 9 years ago Reply
I think a Click - 3 - 2 - 1 - Blocks Appear would be a better idea. If the test design isn't uniform across all trials, then one will have to wonder if one is merely testing the test rather than the participants, which is messy and unnecessary.
? | 9 years ago Reply
The thought of it being "too hard" was very immediate to me as well. Might I suggest you eliminate a row and/or column from the test design, since they are merely distractions? If not, then could you provide those additional columns/rows when the user selects a certain value of boxes to be tracked within each square (set to, say, a default of four squares for 1 box)? I would also suggest giving some colouring to the arrows so that perceptual distinguishing is easier and the test thereby becomes more directed to the task of spatial tracking; and maybe my screen's resolution is too high or my sight is very poor, but I find those tiny squares rather difficult to locate within the short span of their appearance, not to mention when an arrow of the same colour manifests my confusion. Others may disagree, naturally, but this is merely my assessment and considered suggestions.

There are many potential variations open to this initial design, as I see it.
? | 9 years ago Reply
The non-active cells do seem a bit superfluous, but the active boxes are chosen randomly from within those 9. Would it help, if the non-active boxes are simply hidden after the task begins?

You make a very good point about screen resolution, which indubitably introduces confounds into the tests. This test would be pretty pointless if you can't easily find the dots. Regarding this, I don't yet know what would be better: to zoom the test application, or zoom the test elements.

As for the colors, are you saying to use a contrasting color to accentuate the direction of it?
cognitivefun | 9 years ago Reply
I think that's a great idea to fade out the unused boxes once the test begins.

I would suggest zooming-in the test elements, because I can see things getting a little too complicated once you begin zooming in the whole test (needless bulk, primarily). I leave it to your better judgement, so long as the objects are emphasised and more prominent.

As for the colours, yes, I am referring to the colour used to indicate the arrows. They are currently black. The squares are black as well. I find myself unable really to distinguish them within such a short span of time and feel that adding colour to one or the other would be a significant benefit, because this would transfer the cognitive load of identification to a different visual area (rather than ideogrammatic-spatial, the distinguishing factor becomes colouration which is not being used by the whole of the test design (a spatial memory task)).

I hope this aids you in clarifying my points.

Might I add that I enjoy your site? Thank you for sharing it and your abilities to create it.
? | 9 years ago Reply
There is now a simple zooming button on the test page. I don't know how much this makes a difference on larger screens. Let me know if you read this.
cognitivefun | 9 years ago Reply
On my 20 in. 1680x1050 monitor I would say that is definitely an improvement. Thank you. You might even employ moving the "Boxes: N" up so the grid doesn't shift too much to the right and increase the zooming capability even further.
? | 9 years ago Reply
This is lol tough on 3 boxes. Nevertheless, I managed to get 70% by concentrating on two boxes only and guessing (about 33% correct as it turned out - subconscious at work?) if the other turned up.

This feels almost as WM-demanding as the dual n-back and so could puff up the ol' gF (Klingberg reports Ravens PM gains using an adaptive WM task superficially quite different to the Jaeggi one, non?).
cevapcici | 9 years ago Reply
60% on 3 boxes with following all movement. If you change it, could you keep the current one as a speed version pls?
cevapcici | 9 years ago Reply
I would just like to request that the spatial speed test be modified to allow additional time on the initial presentation of blocks so the user can have a little more time to take in the block positions.

The speed of presentation becomes an issue for me at the 4-block level. I cannot practice at the 4-block level because I always seem to have trouble catching the location of one of the blocks.
godel2 | 8 years ago Reply
I agree. It may be a good idea, if this idea will be implemented, that an auto-adjustment on the initial observation time be given for every n items, say, 1.5 sec per each item.
? | 8 years ago Reply
whoops. Just to be clear: by "spatial speed test" I meant "spatial updating test" sorry.
godel2 | 8 years ago Reply

Login to save scores

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