Comments on: What a dead fish can teach you about neuroscience and statistics http://boingboing.net/2012/10/02/what-a-dead-fish-can-teach-you.html Brain candy for Happy Mutants Fri, 24 May 2013 09:54:00 +0000 hourly 1 http://wordpress.org/?v=3.4.1 By: imag http://boingboing.net/2012/10/02/what-a-dead-fish-can-teach-you.html#comment-1549096 imag Thu, 04 Oct 2012 20:15:00 +0000 http://boingboing.net/?p=184176#comment-1549096 Great interview.  Thank you. Great interview.  Thank you.

]]> By: gypsyspacemuffin http://boingboing.net/2012/10/02/what-a-dead-fish-can-teach-you.html#comment-1548041 gypsyspacemuffin Wed, 03 Oct 2012 17:26:00 +0000 http://boingboing.net/?p=184176#comment-1548041 Immediately made me think of this. http://www.youtube.com/watch?v=721Kcrx0ZCE Immediately made me think of this.
http://www.youtube.com/watch?v=721Kcrx0ZCE

]]> By: Tom McCarthy http://boingboing.net/2012/10/02/what-a-dead-fish-can-teach-you.html#comment-1547113 Tom McCarthy Tue, 02 Oct 2012 21:41:00 +0000 http://boingboing.net/?p=184176#comment-1547113 I'll be honest. I only clicked on the link to see if you had a picture of a fish inside the MRI machine. I’ll be honest. I only clicked on the link to see if you had a picture of a fish inside the MRI machine.

]]> By: Mark Dow http://boingboing.net/2012/10/02/what-a-dead-fish-can-teach-you.html#comment-1546997 Mark Dow Tue, 02 Oct 2012 20:32:00 +0000 http://boingboing.net/?p=184176#comment-1546997 I'd add that fMRI responses can be robust in individual subjects, in many cases. While most studies use group averages to make inferences about a population, some stimuli/experiments  (e.g. sensory responses like vision) are robust enough to make inferences about location and magnitude of neural activity in individuals. Of course the statistics still need to account for the large number of tests -- about one test per voxel. Also, "...interested in data at the .5 level." should read ".05 level", "p < .05", or 5% chance that the statistic would happen in randomized data. I’d add that fMRI responses can be robust in individual subjects, in many cases. While most studies use group averages to make inferences about a population, some stimuli/experiments  (e.g. sensory responses like vision) are robust enough to make inferences about location and magnitude of neural activity in individuals. Of course the statistics still need to account for the large number of tests — about one test per voxel.

Also, “…interested in data at the .5 level.” should read “.05 level”, “p < .05", or 5% chance that the statistic would happen in randomized data.

]]> By: Bucket http://boingboing.net/2012/10/02/what-a-dead-fish-can-teach-you.html#comment-1546944 Bucket Tue, 02 Oct 2012 20:01:00 +0000 http://boingboing.net/?p=184176#comment-1546944 Mmm. I think the "sushi" section of my brain is lit up like Christmas right now.  Mmm. I think the “sushi” section of my brain is lit up like Christmas right now. 

]]>