Quickies

Skepchick Quickies, 1.5

Jen

Jen

Jen is a writer and web designer/developer in Columbus, Ohio. She spends too much time on Twitter at @antiheroine.

Previous post

AI: You've Lost That Lovin' Feelin'

Next post

Wakefield: from disgraced MD to GDMFL

26 Comments

  1. January 5, 2011 at 9:24 am —

    I will use my supernatural powers of psychic predictions to predict that Prof. Daryl Bem will be found to be full of crap. If I am right, that is concrete proof that I can see into the future and that paranormal gifts are true. If I am wrong, Prof. Bem is right, that proves that paranormal gifts are true. Take that science!

  2. January 5, 2011 at 9:26 am —

    SGU had a discussion about Bem’s research (starts at about 27:50).

    http://theskepticsguide.org/archive/podcastinfo.aspx?mid=1&pid=281

    In the show notes, they link to a wired article from November, which has two links to two failed attempts to replicate Bem’s work.

  3. January 5, 2011 at 10:40 am —

    …is that a Juggalo education poster?

  4. January 5, 2011 at 11:32 am —

    If I am a researcher trying to replicate Dr. Bem’s results the first thing I would ask for is a copy of the software used to pick the “random” pictures. I would test that software against other program(s) to determine if the effect shows up to a greater or lesser extent depending on the program used.

    I am not a software expert person thingy, nor am I a psychologist, but it seems that if there was a flaw in the program (or the computer) it may allow a subtle pattern to form; so subtle that the subjects looking for “regular” pictures didn’t notice it but when given the proper motivation (PORN!) they were able to pick up on the pattern to a small extent and used it to get to the gooey porn center. Also if the same software (computer) was used in the other experiment it could explain the results there too with the extra concentration needed to complete the task standing in for the porn as a motivator.

    What was the results size in the second experiment anyway? The article reads like is was also 53.1%; if so curiouser and curiouser indeed.

    That hypothesis could explain the results without rewriting physics plus Occum’s razor and all that.

  5. January 5, 2011 at 12:54 pm —

    Am I missing something? The obesity article says that 17.2 canadians are obese, and that 58.6 men and 43.5 women are so. I don’t understand that. The numbers add up to over 100, so it’s not the breakdown of the obese, but they are both over 17.2, so they can’t feed into it, since there isn’t another option. Unless, they are counting children, which would mean that canada would have to be experiencing a baby boom of exponential proportions.

    Or, am I overlooking something?

  6. January 5, 2011 at 1:21 pm —

    @Infinitemonkey I’m afraid you have missed something 58.6 of men and 43.5 of women are overweight, obese is the next level up from overwieght.

    If only that ESP artical was true just think about it the next time your doing a test you could wait for the answers to trickel from the future, nope not happened to me either.

  7. January 5, 2011 at 1:42 pm —

    Great quickies today!

    I so enjoy the measles poster. Is it too much to pretend it’s a draft of a future Health and Human Services campaign?

    I think the NPR article on Bem’s work is better than the average reporting of this type of research. It at least mentions that the results need to be replicate and the failures are already beginning. But seriously, when are we going to end this file-drawer-fueled cycle of do a bunch of crappy experiments, some will inevitably show tiny effects, report these like they’re an amazing discovery, harness the enthusiasm generated by these reports to get support for a bunch more crappy studies? I know, I know. Never, that’s when.

  8. January 5, 2011 at 1:51 pm —

    @B Hitt: Agreed. The article is better than most (and even includes a link to a CSI article on the study). The writer, Robert Krulwich, is one of the hosts of Radiolab which is a wonderful, imaginative and above all else, curious radio show and podcast about science. I highly recommend it.

  9. January 5, 2011 at 2:34 pm —

    53.1%? That’s just a number. What significance test did they do? What’s the margin of error there? Did they not even take Stats 101? Useless!

  10. January 5, 2011 at 3:04 pm —

    My thoughts exactly, @Advocatus Diaboli. Bem appears to use large leaps of logic to reach his conclusions. This reporting is also why I have a hard time listening to Robert Krulwich on RadioLab when he goes off on spiritual tangents on a science show.

  11. January 5, 2011 at 3:19 pm —

    @davo_301: Ok, thanks for clearing that up.

  12. January 5, 2011 at 3:29 pm —

    @Advocatus Diaboli: That’s what I was thinking. I don’t know much about stats (I just took a 101 course and that’s it, hah) so I wasn’t sure if I was right that they need a margin of error for that to prove significance…good to see someone else pointing that out more confidently than I could :)

  13. January 5, 2011 at 3:52 pm —

    Anyone else just a little creeped out by the fact he used porn as a motivator? Why not a bowl of chocolate, or a picture of a cute kitten?

  14. January 5, 2011 at 4:26 pm —

    @Holytape: Believe it or not, using porn is one of the tried-and-true techniques of research in psychology and cognitive neuroscience. I’ve seen it used in the protocols of tons of studies. It’s seems a little unsettling but when you need an emotionally salient and evocative stimulus, you can’t go wrong with sexually explicit or violent/gory images. We’re evolutionarily programmed to have strong responses to them.

    @mrwilson41: Yeah, I have a mixed reaction to Radiolab. On one hand, I can’t complain about a show that is so skillful in packaging science in an entertaining and inspiring way for the general public.

    On the other hand, because science is what I do every day for a living, the hyper-critical, over-sensitive super-nerdy-pedant side of me cringes a lot when I listen. Particularly when a scientist is explaining her/his work and one of the hosts says “So basically what you’re saying is . . .[statement that is 10X loftier, poetic, and awe-inspiring but 100X less scientifically accurate]. They tend to do this most of all with neuroscience, which is my field so it’s worse.

    There are plenty of journalists and commentators out there that are talking about science in a more “sciencey” way but also an entertaining and inspiring way (SGU crew) (I need new adjectives).

  15. January 5, 2011 at 4:44 pm —

    @B Hitt: Thanks for your insight. I enjoy most of RadioLab’s podcasts and think of Jad Abumrad almost like Ira Glass. I just have to skip some episodes that go into non-science topics.

    Sublime, Eloquent, Passionate, Poignant, Lucid …

  16. January 5, 2011 at 5:14 pm —

    @B Hitt

    (I need new adjectives)

    Here’s a noun that will help with your adjectives – Thesaurus. j/k

  17. January 5, 2011 at 5:28 pm —

    I heard about the girl finding the supernova yesterday – that story makes me happy :)

  18. January 5, 2011 at 5:46 pm —

    Re the ESP article,the good thing was that reading the comments,I noticed that most of the higher rated comments were very sensible ,skeptical,and coherent.Contrast that with most other popular internet blog sites.(shudders!!!)

  19. January 5, 2011 at 6:43 pm —

    About RadioLab…I also notice that Robert Krulwich seems to be the softer on science of the two. But it is a great show to get people who “hate” science to think about it once and awhile. It has worked with my friends….which is anecdotal evidence but….

  20. January 5, 2011 at 8:00 pm —

    I’d like to see *INDEPENDENT* verification of the ESP claims — by other researchers using different software.

    I, too, would like to inspect his software. Java, for example, is very very popular now, and it’s known to have a fairly weak “random number generator.”

    Getting real random numbers from a computer is difficult, because computers are designed to be deterministic. So the best we can hope for is unpredictable, not really random — unless one puts some real effort into it. (And when partially informed coders put ill-informed effort into it, they usually get *worse* results, not better! It’s true!)

    [Yes, I’m a computer expert. 30 years of industry experience.]

  21. January 6, 2011 at 9:53 am —

    Very good article in today’s NYT on Bem’s ESP research. Quite critical and they even bring up the peer review process, bias and statistics:

    http://www.nytimes.com/2011/01/06/science/06esp.html?_r=1

  22. January 6, 2011 at 10:53 am —

    If this ESP work is so accurate why don’t more people win the lottery?

  23. January 6, 2011 at 10:57 am —

    Specifically, why can’t I win the goddamn lottery?

  24. January 6, 2011 at 12:38 pm —

    Mark Hall,it’s because the effect size is so small.I would recommend to increase your chances significantly,you should buy 20 million or so tickets.Good luck!

  25. January 6, 2011 at 4:12 pm —

    @tmac57: Fuck. We’re supposed to buy tickets?

  26. January 7, 2011 at 12:42 am —

    The awesome thing about Kathryn Gray finding that supernova is that she was actually LOOKING for one. That is so effing cool. And her middle name is Aurora! And the person who inspired her was a 14-year-old girl. These kids rock.

Leave a reply