Comments on: Can a Dead Fish Prove that Modern Brain Studies Are Bunk?

By: Frank Powers

Frank Powers — Sun, 07 Feb 2010 10:26:26 +0000

Come on now, Brett – change that bloody title already. It’s still misleading, regardless of the article’s informational value.

By: Vasco H

Vasco H — Sun, 07 Feb 2010 00:56:36 +0000

Already said but: you really cannot shoot shock titles like this that have nothing to do with the article. Is this Ananova or what?

By: Kyle

Kyle — Sun, 07 Feb 2010 00:22:49 +0000

so you respond to craig but don’t change the title?

come on, brett. The title of the article is either intentionally misleading to garner readers or simply ignorant… either way you’d be better off changing it.

interesting enough finding but intellectual dishonesty is a no go. change the title.

By: Ivan

Ivan — Sat, 06 Feb 2010 00:21:12 +0000

I will never feel the same about eating lox again. I thought that that fish was giving me the evil eye, now I guess I was correct. Salmon have been underestimated for years. Is this proof of life after death?

By: Can a Dead Fish Prove that Modern Brain Studies Are Bunk? « Wisdom Beat

Can a Dead Fish Prove that Modern Brain Studies Are Bunk? « Wisdom Beat — Mon, 28 Sep 2009 17:38:36 +0000

[...] Read the article. [...]

By: jp

jp — Tue, 22 Sep 2009 03:48:37 +0000

Interesting article. In market research there it is now de rigueur to use fMRI to test advertising effectiveness… the flaws inherent in that methodology go beyond the false positives, but Mr. Bennett does bring up a good point.

As for the misleading title… well, let’s be honest… its hard not to read an article whose title refers to debunking studies with a dead fish, so I guess Brett accomplished his goal and we informed ourselves in the process.

By: Brett Israel

Brett Israel — Mon, 21 Sep 2009 22:00:44 +0000

Hi Craig – Thanks for stopping by to comment and we’d love to open up the conversation between you and our readers!

By: Craig Bennett

Craig Bennett — Mon, 21 Sep 2009 21:33:33 +0000

The title and content of this article are both somewhat misleading. The point of our poster/paper is not that modern brain studies are bunk, but rather that we have to be very careful regarding how we compute and report our results. Also, for the record, we believe that fMRI is an enormously useful for investigating brain function. We use it every day as part of our research and think that it holds enormous promise in helping us understand human cognition.

The issue at hand is one of controlling for false positives (a result that isn’t really a result) while still having enough ability to detect real signals. It is a kid of zero sum game. You can set your thresholds so high that you get no false positives, but then you would not observe any real results either. on the flip side you can set your thresholds so low that you detect all of the real results, but then you will have an overabundance of false positives as well. Where do you draw the line? In our commentary we argue that we need to set a stricter threshold than many neuroimaging papers might typically use.

I will keep an eye on the comments if anyone has questions. One thing I do have to say about this article is that it has the best photo associated with it of any weblog post yet!