Thursday, February 19, 2009

Encephalon #64

The latest issue of Encephalon (a neuroscience and psychology blog carnival) is out at Neurocritic, and it's really a very fine piece of work. Check it out.

Monday, February 16, 2009

Muppet chickens and Kubrick

Someone somewhere (I didn't bother to check) made a version of the chickens from The Muppets doing the Blue Danube Waltz. Someone else figured it would be a good idea to lay this over the famous sequence using that music from 2001: A Space Odyssey. Who am I to argue?

Thursday, February 12, 2009

Dead or Alive? Knowledge about a Sibling’s Death Varies by Genetic Relatedness in a Modern Society

ResearchBlogging.orgHappy Birthday, Charles Darwin! This piece is a modest contribution to a large scale blog love-in that I mentioned in an earlier posting. It's about a recent study by Thomas Pollet and Daniel Nettle driven by a simple hypothesis about genetic relatedness, and how this might make a difference to what people know about other people. Specifically, it was hypothesised that individuals would be more likely to know whether a related sibling was alive or no. Here's the abstract:
Using a large sample of non-institutionalized individuals from the Netherlands
(n = 7610), we examined the influence of relatedness on an individual’s knowledge about whether their sibling is alive or not. Respondents were generally less likely to know whether their sibling was alive if they were not fully related. The effects were stronger for differences between paternal half-siblings and full siblings than for differences between maternal half-siblings and full siblings.
Hamilton (1964) argued that selection should favour investment in closely related kin, and this hypothesis has been largely confirmed in a wide variety of studies. A reasonable precondition for investing in someone is knowing whether they are alive or dead, and so whether someone has this knowledge is a plausible proxy for openness to investment. More specifically the authors say "Our hypothesis is that respondents will be more likely to know whether a sibling is alive when they are fully related than when they are not."

This study is part of a larger, ongoing, study of kinship in the Netherlands (the Netherlands Kinship Panel Study (NKPS)). For the present research "respondents [were] asked to list all their siblings (adopted, step, half- and full siblings) and to indicate whether they know whether this sibling is alive or not".

As the authors note, an alternative view is that levels of affiliation and dispostion to invest might follow development, in particular who was around when an individual grows up. "This proximate theory would suggest that relatedness in most families is confounded with being raised together." They proposed to track this confound by distinguishing both paternal half-siblings (who are generally raised apart) with full siblings and maternal half-siblings (who are generally raised together) with full siblings. The study excluded adopted siblings because there were so few in the study. (It would be interesting to know more here, but a specific study would probably be needed to ensure enough adopted individuals for useful analysis.)

Data analysis considered up to the first six siblings, and controlled for "education (9 categories: from incomplete/primary to university/postgraduate; treated as interval), age, age differences between the siblings, gender of the respondent, and number of family transitions before age of 16" where a "family transition" was "an alteration in the respondent’s family living situation since birth, for instance going to live with grandparents or with another family member". The analysis constructed multinomial logistic regression models to calculate the "independent effect of sibling type on knowledge about the sibling’s death". I was surprised to hear that over 1200 of the subjects (total number 7610) even had as many as six siblings.

Figure 1 below shows the raw proportion of respondents who did not know whether a sibling was alive or dead as a function of sibling type and number.

The analysis using the regression models showed that across "all six models, sibling type proves to be a highly significant predictor of knowledge about a sibling’s death [...]. The other variables did not prove consistent predictors." And the predictive value of sibling type is dramatic: "respondents are 25.75 times more likely to not know whether their sibling 1 is alive when" that sibling is a step-sibling than when he/she is a full sibling (emphasis added).

The authors are careful to note that their hypothesis concerned ultimate outcomes, rather than proximate mechanisms. It's an interesting question by what means relatedness is tracked, and to what extent it depends on epigenetic factors. Surprisingly the authors don't mention a study published in 2007 in Nature regarding sibling detection. That paper is well worth a look - it doesn't settle the questions here, partly because the proposed system is partly dependent on observed interactions with mothers, and duration of coresidence with candidate 'siblings', but it should have been discussed. Also, the conclusion of the Nature paper is worth quoting in this celebratory article:

"These results contribute to a growing body of findings showing that humans are not immune to the evolutionary forces that have shaped other species, and that Darwinism has a central role in discovering the neural and psychological architecture of our species."

References:

Hamilton, W.D. (1964). The genetical evolution of social behaviour I, II. Journal of Theoretical Biology, 7, 1-52.

Lieberman, D., Tooby, J. & Cosmides, L. (2007) The architecture of human kin detection, Nature 445, 727-731.

Thomas V. Pollet, Daniel Nettle (2009). Dead or Alive? Knowledge about a Sibling’s Death Varies by Genetic
Relatedness in a Modern Society Evolutionary Psychology, 7 (1), 57-65 DOI: http://www.epjournal.net/filestore/ep075765.pdf

Monday, February 9, 2009

Comparative Mind Database

There's a possibly cool project in the offing to create a Comparative Mind Database. My mate Dave's mate Wayne has written some useful thoughts on the challenges facing such an endeavour over at his blog. Wayne's right - there are some tricky questions. I think his list is pretty good, but would add a concern that isn't specifically related to design, but to the economics of the affair.

If it takes work (over and above what can be automated) to get research reflected in the CMD, which it will, then it has, as far as possible, to be in the interests of the people doing the science to do the work. That pretty much has to mean that the payoff in terms of impact for work already done and quality of ongoing work should be greater than the cost, over some acceptable time scale.

Bill Gates TED talk

Here's Bill Gates, talking about mosquitoes and the problem of creating great teachers. Both big issues.

Thursday, February 5, 2009

Mini blag roundup

I've been a bit slow blagging for a week or two (not counting today). I'm not heading for another prolonged layoff, and I haven't been incapacitated with digrizfizzes. I've just been pre-occupied with various affairs in meatspace. Also I've been semi-working on a few pieces that are taking longer than expected.

To help fill the dead air, here's a roundup of a few things I've read that are well worth looking at.

First, over at Not Exactly Rocket Science (a fine blog if you don't know it) you'll find the following:

(1) A good write up of an interesting and careful study showing how self-reported race classification was sensitive to changing socio-economic status.

(2) A good, if ultimately rather depressing, piece on how 'absolutists' are less capable of rational compromise, although more responsive to symbolic incentives.

(3) A very interesting piece on a genetic variation that is associated with different levels of aggression in response to perceived provocation. [This Achewood on Morally Challenging Hot Sauce is more relevant that you might think until you read the aggression article.]

Then at the British Psychological Society Research Digest (also a good one to keep an eye on) are these:

(1) An account of a study of the rhetoric of the (far right) British National Party, and how it's like a conspiracy theory.

(2) A study confirming that you really do spend more when using credit than you do with cash, and when you're sad. Cash/Credit study (sad)

(3) A very interesting study finding that training in face discrimination reduces magnitude of the effect on the race Implicit Association Test (IAT):

And finally over at Epiphenom, a report of a study showing that spiritual guidance actually reduces success in substance abuse treatment.

Construal Level and Procrastination

Here's a study that's got a bit of attention, including at The Economist, at intellectual vanities and at Psychology Today. The paper found that the degree to which a task was construed abstractly made a difference to the level of procrastination: more abstract meant more procrastination.

You can currently get the paper at website the first author (Sean M. McCrea) here.

This is the full abstract:
According to construal-level theory, events that are distant in time tend to be represented more abstractly than are events that are close in time. This mental association between level of abstractness and temporal distance is proposed to be a bidirectional relationship, such that level of representation of an event should also have effects on the time when the activity is performed. In the present studies, participants were asked to respond to a questionnaire via e-mail within 3 weeks. The questionnaire was designed to induce either an abstract or a concrete construal. Using a variety of manipulations of construal level, the studies supported the predictions of construal-level theory. Individuals were less likely to procrastinate performing the task when the questionnaire induced a more concrete construal. Furthermore, this effect did not depend on the attractiveness, importance, or perceived difficulty of the task.
Based on a body of previous research suggesting among other things that events in the distant future are more likely to be represented abstractly (and vice-versa), and that concrete construals lead to better task performance, the authors hypothesised that "forming a concrete representation of a task will reduce procrastination, independently of any effects of planning or understanding of the task."

There were three complementary experiments, all using subjects who were students at the University of Konstanz. In all cases students had a questionnaire (about various tasks) to complete and return by email. In all cases they rated the importance, difficulty, pleasantness, convenience and interest of the tasks In all cases the number of hours taken to return the questionnaire (there was a deadline) was recorded (there was also a coding system for 'early', 'on time', 'late' and 'not returned' for an additional analysis). Subjects were randomly assigned to the abstract or concrete condition in each experiment.

There were also some differences between the three experiments.

In experiment 1 the questionnaire itself included either abstract or concrete demands relating to the same generic tasks ("open a bank account"). The abstract condition required participants to describe "what characteristics are implied" by the activity, in the concrete condition the demand was to say something about how to go about the task.

Experiment 2 was much the same, except that the abstract task required providing categories ("A bird is a _______" while the concrete task required providing examples ("A bird is an example of a _______").

Experiment 3 was designed to address a risk that, despite the inclusion of difficulty and other ratings, any difference in return times in experiments 1 or 2 was related to something other than the abstract/concrete construals. Here all subjects completed the very same questionnaire, but half were primed concretely and half abstractly. The prime was a colour print of Seurat's 'La Parade' (detail from it here), with subjects either told something about the technique of pointillism (concrete) or the desired effect (abstract).

One way of looking at the bottom line of the results (there are several analyses in the paper) is to look at the response times for the three experiments:
Experiment 1:
Concrete: 175.78 hours
Abstract: 503.85 hours

Experiment 2:
Concrete: 301.76 hours
Abstract: 532.20 hours

Experiment 3:
Concrete: 338.75 hours
Abstract: 491.29 hours
As the authors note this is a significant addition to our understanding of procrastination. It's been known for some time that factors such as delay discounting are important for explaining procrastination, but this shows in addition that "the way the task is represented influences when individuals complete it."

This is important, and possibly useful. Maybe some of my mate Dave's more tardy grad students would find it helpful if they thought in more concrete terms about their overdue work ("write 500 decent words a day") instead of being distracted by hopeless tasks that happen to have vivid concrete descriptions ("read the whole internet"). Are you reading this, punks?

Finally, all discussion of procrastination should include a look at John Perry's fabulous essay on structured procrastination.

ResearchBlogging.org
Sean M. McCrea, Nira Liberman, Yaacov Trope, Steven J. Sherman (2008). Construal Level and Procrastination Psychological Science, 19 (12), 1308-1314 DOI: 10.1111/j.1467-9280.2008.02240.x

The Intrepid Aardvark

There's a new group blog for science living woo-loathing bloggers in South Africa. Or, more precisely, currently in South Africa, but with tentative dreams of going pan-African somewhere down the track. The group blog is called The Intrepid Aardvark, and following two self-introductions by members (full disclosure: I'm one of them) it's now got its first 'proper' article, about dodgy reporting of a study in acupuncture in the national media.

Wednesday, February 4, 2009