Very interesting post. It's also heartwarming to see that chess isn't only about bashing moves, but also about thinking and planning :-)
@pohlheimer said in #12:
> it is 4, isn't it?
It should be #6 - There are three sequences of shape change across the columns, for example: (1) Outer shapes change from circle to square to diamond (wrapping around the right edge back to the left), (2) There is a horizontal contour that changes from a flat horizontal line to a steep horizontal wave, then to a shallow horizontal wave (again, wrapping around the right edge), (3): There is a vertical contour that changes from a flat vertical line to a vertical wave to a diagonal line (again, wrapping around the right edge).
Together, this means that we need an outer square shape, a flat horizontal line, and a vertical wave to satisfy each transformation across the bottom row.
I need to check against the original scale to be 100% sure, but I'm pretty confident in that answer.
> it is 4, isn't it?
It should be #6 - There are three sequences of shape change across the columns, for example: (1) Outer shapes change from circle to square to diamond (wrapping around the right edge back to the left), (2) There is a horizontal contour that changes from a flat horizontal line to a steep horizontal wave, then to a shallow horizontal wave (again, wrapping around the right edge), (3): There is a vertical contour that changes from a flat vertical line to a vertical wave to a diagonal line (again, wrapping around the right edge).
Together, this means that we need an outer square shape, a flat horizontal line, and a vertical wave to satisfy each transformation across the bottom row.
I need to check against the original scale to be 100% sure, but I'm pretty confident in that answer.
Do you have equivalent topic references that do not require a subscription. Has cognitive sciences not gone through more publicly available outlets, like physics, bioinformatics, machine learning, and what not. I mean those titles are appealing.
None of them accessible in full. Not even on PubMed. I can still read your extraction from them, that is fine.
None of them accessible in full. Not even on PubMed. I can still read your extraction from them, that is fine.
The TOL experiments look a lot like turn by turn calculation. The pure logic of it. How much room was there in the solutions that would be distinct and yet solve the puzzle. I do not think that chess pattern recognition is about that type of context, it is more about goal setting than method to get there. And assessment of desirability while doing problem solving and often going back while calculating to a previous calculation. The width of the possible choice or the room, might not be tested here.
Yet I understand chess makes better calculator of us, or we were better calculator and all went to play chess.. Self selction or trauing was not the question... I just wonder about excluding or having wanted pattern recognition in a corridor about ordering and timing, nothing really actually spatial about it on the perceptio side. although vision required, the information did not seem to have much spatial signature. or mayb I am stuck with chess.
Can you explain what is essentially spatial information. I see more ordering. and reordering. the fact that it might have color ID in one case, and width in the other does not seem to be part of the problem to solve. although we may have a fast linguistic to task understanding about bigger and smaller, that would be the extent of pattern recognition being fast, and that would not change because of chess being about 2D information pattern.
I guess the study is not saying much about pattern. it is comparing sequential move planning with logical rules of ordering given a possibly unique solution to find? I may have missed something. I welcome having had that wrong though. please respond.
Yet I understand chess makes better calculator of us, or we were better calculator and all went to play chess.. Self selction or trauing was not the question... I just wonder about excluding or having wanted pattern recognition in a corridor about ordering and timing, nothing really actually spatial about it on the perceptio side. although vision required, the information did not seem to have much spatial signature. or mayb I am stuck with chess.
Can you explain what is essentially spatial information. I see more ordering. and reordering. the fact that it might have color ID in one case, and width in the other does not seem to be part of the problem to solve. although we may have a fast linguistic to task understanding about bigger and smaller, that would be the extent of pattern recognition being fast, and that would not change because of chess being about 2D information pattern.
I guess the study is not saying much about pattern. it is comparing sequential move planning with logical rules of ordering given a possibly unique solution to find? I may have missed something. I welcome having had that wrong though. please respond.
can you decompose fluid intelligence without using the word intelligence.
maybe reasoning types.
deduction, induction, analogy, generalization, what else.
reframing (s that a thing, scoping the problem, slicing the problem projecting the problem, embedding the problem).
dissecting the problem. running as fast as possible to produce an answer, or find all the methods one could arrive at the solution but in different ordering of scanning of the patterns of variation X pattern of equivalent (maximal similarity over one aspect).
maybe reasoning types.
deduction, induction, analogy, generalization, what else.
reframing (s that a thing, scoping the problem, slicing the problem projecting the problem, embedding the problem).
dissecting the problem. running as fast as possible to produce an answer, or find all the methods one could arrive at the solution but in different ordering of scanning of the patterns of variation X pattern of equivalent (maximal similarity over one aspect).
Just use sci-hub ru
Far too long ago that I did basic statistics but I was wondering if 25 individuals would be sufficient to conclude with statistical evidence?
@schruv said in #18:
> Far too long ago that I did basic statistics but I was wondering if 25 individuals would be sufficient to conclude with statistical evidence?
Good question. These days that sample size would probably strike people as a little light, though it does depend on the size of the effect you're trying to detect. If I were planning this study from scratch today (and didn't know about this result), I would probably shoot for more like 50-60 people per group.
That said, their effect sizes look to be pretty healthy, so I'm not especially worried about the differences they report in ToL accuracy. What I wonder more about is whether or not there may be small effects we can't see with this sample size in the other assessments like the matrices, VWM, etc. Those null effects might be victims of low statistical power.
As always, replication would be awesome.
> Far too long ago that I did basic statistics but I was wondering if 25 individuals would be sufficient to conclude with statistical evidence?
Good question. These days that sample size would probably strike people as a little light, though it does depend on the size of the effect you're trying to detect. If I were planning this study from scratch today (and didn't know about this result), I would probably shoot for more like 50-60 people per group.
That said, their effect sizes look to be pretty healthy, so I'm not especially worried about the differences they report in ToL accuracy. What I wonder more about is whether or not there may be small effects we can't see with this sample size in the other assessments like the matrices, VWM, etc. Those null effects might be victims of low statistical power.
As always, replication would be awesome.
Intruiging topic. Very interesting too about that second data point and its interpretation.
I think in chess you need pattern recognition as well as calculation. I can see how calculation skills might help with other non-chess problems which also involve 'calculation of variations'. For example in the tower of London exercise you may apply calculation skills by comparing different 'moves' and different sequences of moves and then you may also have some process equivalent to evaluating a position, to judge whether you've made some progress.
I think in chess you need pattern recognition as well as calculation. I can see how calculation skills might help with other non-chess problems which also involve 'calculation of variations'. For example in the tower of London exercise you may apply calculation skills by comparing different 'moves' and different sequences of moves and then you may also have some process equivalent to evaluating a position, to judge whether you've made some progress.