cognitive fun!

Learn your mind. Play it too.
More on the dual n-back differences
cognitivefun | 10 years ago Reply Link me
More on the dual n-back differences
A user has written in about a different online dual n-back test at soakyourhead.com. Apparently this is explicitly modeled after the original paper's protocol. I have not seen what it looks like because my computer doesn't run SilverLight, but it struck me that a difference in how the stimuli are presented could influence the test performance. I apologize for this oversight.

In the test on this site, the visual stimuli are presented as pictures that differ in appearance. While the differentiating factor is the position of the white block within the gray block, it is presented as a whole.

It is possible that removing the background block, and presenting the position of the visual stimuli only would result in a different encoding method; in particular, perhaps it would become abstracted into a spatial location. This deserves a follow up test, as much as another multitude of variations.

Again, if anyone feels like pointing out that the dual n-back test on this site is different from that of the paper, they will be right, and I am well aware of this nor do I think it is a problem, and will not hesitate to discuss the matter, in public or private.
cognitivefun | 10 years ago Reply
I noticed that test informs (flashes red) on a bad choice, but keeps quiet when a choice is missed. The test on this site informs in both cases.

Which is more consistent with the "official" test, and more importantly, does it matter?
rectotron | 10 years ago Reply
I am not familiar with the SYH implementation. I also don't know to what extent feedback mechanisms were used in the "official" test, but from the paper, it seems sparse.

I would think feedback mechanisms do make a difference, so I would say yes to your question. As to what method is better, I can't say for sure, but this can be said safely: the feedback needs to be in close proximity with the stimulus and response (maybe around the order of 0.5 sec, which is close to the observed optimum time for stimulus-response behavioral training, but I don't know how similar it is for reinforcement or punishment feedback).

Tests here would provide feedback on positives, false negatives, and false positives. That true negatives do not have feedback is simply to reduce the number of visual items on screen. I actually don't know to what extent is learning promoted by this paradigm. This is a good topic!
cognitivefun | 10 years ago Reply

Login to save scores

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