Friday, January 12

I would normally wait to see this covered somewhere other than a self-interested press release, but then I would normally forget about it and it would fall through the cracks, so I'm going ahead and posting it as is. Supposedly it's a Tufts University initiative and will be published in "Psychonomics Bulletin and Review," so you know it's gotta be kosher.

Anyhoo, the gist is that pigeons are smarter than you think. In one study, the birds have now proven capable of remembering "up to 1,200 images such as landscapes, cars or people simultaneously" - that is, they can match a given image with the right food lever. Kind of a complex game of "Concentration," which I remember as challenging enough even without 1200 options.

The second study is eve more intriguing:

    In the second study, the pigeons first viewed a page of 16 small pictures that were either all the same or all different. For example, if the first page showed 16 small identical mugs, the birds were then able to successfully choose a page showing 16 identical hourglasses over a page showing a variety of pictures. The pigeons also were able to identify pages of all different pictures. The results of this study demonstrated the pigeon's ability to compare two relationships, sameness or difference, rather than just two pictures.
Wow. Again, I'd like to see all this in the words of the Tufts researchers. But if true this is pretty impressive, going beyond simple counting or recalling to a) being able to tell how one group of objects is or is not like another group, and b) being able to tell what the researchers want you to show that you can do. They always seem to leave this part out - I wonder how many animal cognition experiments have failed because the animals just didn't catch onto the point of the experiment, or else weren't that interested in playing along.

Tuesday, January 9


That's the unspoken message in two science stories I just ran across on the front page of Google News, though neither, of course, spells it out this way. And in one case, the evidence is only circumstantial: In nearly one in five nutrition studies done on milk, there is a conflict of interest on the part of the author(s) that calls into question the credibility of the entire study. While it's possible the "author conflict" turns out to be irrelevant to a given study's conclusions, it's worth noting that this percentage of occurrence - 18% - is greater than that found in juice studies (8%) and soft drink studies (6%) combined.

Unfortunately, as far as I can tell the authors of this study don't even mention that fact, only their sun-rises-in-the-east conclusion that beverage studies' conclusions may be biased in favor of the studies' funders. But the data is there in "Table 1" on page 4 of this document (the "author conflict" data). Dunno why the study's authors didn't see fit to point out this anomaly. Who's funding them, anyway?

The other item is more straightforward: Milk cancels health benefit of drinking tea, much as milk cancels health benefit of eating chocolate. Of course if one food is health-positive and adding another renders it health-neutral, it's a pretty good bet, isn't it, that the second food is health-negative, that is, bad for you in some particular way. But I guess it's asking too much for any of the news reports I've read to even bring up that possibility. It's all just some bizarre phenomenon that only affects milk when in combination with black tea. Right.