Friday, November 30, 2007

Article #6 - Blanchard et al. and Measuring Pedophilia

Blanchard, R., Kuban, M., Blak, T., Cantor, J., Klassen, P., & Dickey, R. (2006). Phallometric Comparison of Pedophilic Interest in Nonadmitting Sexual Offenders Against Stepdaughters, Biological Daughters, Other Biologically Related Girls, and Unrelated Girls Sexual Abuse: A Journal of Research and Treatment, 18 (1), 1-14 DOI: 10.1007/s11194-006-9000-9

Blogging on Peer-Reviewed Research So now we're delving into the mind of the pedophile. It's a disturbing, uncomfortable place to be, but it can also be quite fascinating and revealing of human nature in general. To me, at least, it's much more interesting to study the fixated, or "classical," pedophile, versus the opportunistic molester. They are generally much more deliberate and provide much more elaborate reasons for their behavior, which can be traced to the fact that they have a genuine attraction to children. While the opportunistic offender generally operates under a mixture of drugs, rage, and hormones, the fixated offender often truly appears to believe that they are doing something that should be okay, if only society loosened up.

Of course, there are differences among fixated pedophiles, as well. The paper for this week provides a good look at not only what these differences are, but also the techniques used to study this kind of behavior. The researchers got their subject pool from community and legal referrals for "illegal or disturbing sexual behavior" (except for the controls, who were referred by physicians), and they were grouped by their relation to the victim: biological fathers, stepfathers, nonpaternal incest, unrelated molesters, offenders against women, and the control group, who had no illegal activities but reported sexual complaints.

In order to gauge the degree of pedophilic orientation in each of these groups, the researchers employed a phallometric device - basically, a pressure gauge that fits over the penis. The subjects then heard descriptions of sexual acts with various subjects, male and female, ranging from prepubescents to adults, accompanied by slides. To get the "pedophilia factor approximation score" the results for the adult descriptions were subtracted from those of the prepubescent and pubescent descriptions.

As you can see in Fig. 1, there was a clear trend between relation to victim and pedophilic score. As the authors broke it down, the less of a social relation a perpetrator had to his victim, the more his pedophilic orientation. It appears that the more effort a person exerts to get to a victim, the more pedophilic he is likely to be. Interestingly, the biological fathers and stepfathers scored equivalently, while nonpaternal but related perpetrators scored on par with unrelated molesters. However, if you'll remember from class, stepfathers are more likely to offend than biological fathers (i.e. the percentage of stepfathers who are molesters is higher than the percentage of biological fathers who are molesters), which Dr. Geiselman attributed to some planning or manipulations by stepfathers who only marry for access to children. This would imply that stepfathers should display more of a pedophilic orientation, driving them to go through this effort to get what they want. But the results show that the ones who are more pedophilic are already related to the victim by chance!

There appears to be a complicated interaction between pre-existing motivation and access to victims that determines whether a person will offend or not. So are stepfathers scheming to get their hands on little kids, or are they slightly abnormal guys who just happen to end up in situations where their weaknesses are laid bare? Is there something special about raising an unrelated child that makes a stepfather more likely to offend? Many mammals rely on pheromone signaling to differentiate related and unrelated potential mates, i.e. those who are safe and those who might lead to genetic defects. It could be that similar systems operate in humans, and when a stepfather who would otherwise not display and pedophilic tendencies comes into regular, close contact with a young, unrelated child, this pheromonal system doesn't send the inhibitory signals that it would if the child were related.

Sunday, November 18, 2007

Article #5 - Zimbardo's Stanford Prison Experiment

Haney, C., & Zimbardo, P. (1998). The Past and Future of U.S. Prison Policy: Twenty-Five Years After the Stanford Prison Experiment American Psychologist, 53 (7), 709-727

Blogging on Peer-Reviewed Research As Neal Osherow mentioned in class, all the really groundbreaking research in social psychology was done 30-40 years ago, before people started realizing all the ethical implications of what they were doing. Thus, we ended up with things like the Stanford Prison Experiment (SPE). If you're not familiar with it, then please check out Philip Zimbardo's slide show and website devoted to it. By setting up what at first glance would appear to be a simple study where undergrads were randomly assigned to be either "prisoners" or "guards," everyone involved, including the experimenters, ultimately got sucked into the situation and modified their behavior to fit stereotyped roles and beliefs of who deserved punishment and who was worthy to deliver it. The SPE has been criticized on many grounds (the Wikipedia page does a good job of summing up most of them) but it's still a clear example of just how situations can determine the boundaries and appropriateness of human behavior.

In their 1998 look back on the SPE, Haney and Zimbardo describe how information about imprisonment learned from psychology and the practice of institutionalization in the United States have diverged sharply since the 1970's. Many of these statistics and graphs should be familiar, as Dr. Geiselman has used them in his lectures, so you should know the basic story already: in the late 1970's the positivist school of criminology held sway and there was a push for rehabilitation and treatment of prisoners; however with the 80's came a shift to more conservative policies and a crime control emphasis - gang activity and crack grabbed media headlines and threw society into a panic. From then to our current situation the focus has been on locking more and more people up, to the point where now the majority of inmates in federal prisons are there on drug offenses for which they'll never receive treatment adequate to prevent recidivism.

Zimbardo and Haney point out that the setup of modern prisons is everything it shouldn't be: antagonistic guards lump all prisoners together, regardless of psychological, social, or cultural differences; judges, society, and prison administrators are intent on imposing "prison pain;" when/if a prisoner is released, they have little to no preparation for re-integration into society, and the longer they've been incarcerated the less successful they're likely to be on the outside. They relate these factors back to similar situations that developed in the brief days of the SPE, arguing that, even though inmate advocates have made some slight progress towards more humane treatment (with some disastrous setbacks), ultimately any change will have to come from the top down. Unfortunately, it seems that finding a person who is both far enough removed from actual prison environments to be able to avoid falling into a stereotyped role and in a position to effect actual change requires going all the way to the Supreme Court. The Reagan-era crime-&-punishment conservatism is currently still in ascendance, however, so no progress on prison reform is likely until the makeup of the court shifts to be more sympathetic.

Stepping away from prisons, it can be seen how the implications of the SPE would carry over to other situations. In an interrogation a person is detained by authority figures, often left in a state of uncertainty as to his fate or even why he's there, and may be treated alternately with sympathy or scorn. Interrogations are much, much shorter than the six days of the SPE, and the methods derived by some of the SPE guards would never be allowed in most precincts, but with a heavy confluence of stress factors similar results of compliance will be likely.

Sunday, November 11, 2007

Article #4: Jury Deliberation

Gastil, J., Burkhalter, S., & Black, L. (2007). Do Juries Deliberate? A Study of Deliberation, Individual Difference, and Group Member Satisfaction at a Municipal Courthouse Small Group Research, 38 (3), 337-359 DOI: 10.1177/1046496407301967

Blogging on Peer-Reviewed ResearchI've been slacking a little on getting these posts out by Friday, but I'm getting back on track. This one is directly relevant to lecture, as it covers aspects that directly affect jury verdicts and the quality of those verdicts. In the taxonomy of jury studies, this is one that uses actual jury members, so there are some specific advantages and disadvantages on that basis alone. While the results can be assumed to be very valid and relatable to the real world, the experimenters couldn't control for specific variables or conditions. Since this was a survey-style study, their goal was to identify response trends that could be attributed to specific characteristics of the jurors. Of course, there are also strengths and weaknesses inherent in surveys (respondent truthfulness, self-selection, etc.) but we won't go into all of those right now.

The authors put forward several hypotheses to test.
Specifically, they propose that jury members will be more likely to deliberate when they are motivated (1d) and have a positive view of the American jury system (1a), and when the jury is made up of more similar people (1b) with more political and educational experience (1c). Further, they're interested in whether it's better to have more juries be more homogeneous with regards to these variables, even if that means everyone's lower on them, and how deliberation affects satisfaction. This goes back to how Dr. Geiselman mentioned how judges generally try to be accommodating to jurors to encourage them to come back (though this doesn't exactly square with the exasperated and condescending behavior described in the book during trials; hopefully these are the exceptions).

One thing you might notice about the study: these are small juries. The mean was 6 jurors, but it says there were some with 5 or 7. I, for one, wasn't previously aware that jury size could deviate like this. But keep in mind that this sample doesn't use the standard 12-member jury. Do you think this might affect the ability to generalize the results? We already know that smaller juries are more likely to reach a verdict; certainly the quality of deliberation has some effect on that. Another thing to keep in mind while reading the results is the type of trials these juries tried:
Nearly a third (32.2%) of those participating in this study sat on juries hearing lowlevel assault charges, another 16.1% heard drunk-driving cases, and the rest heard a range of minor offenses from sexual indiscretion to reckless driving. The median juror spent 2 days in the courtroom, with 91% spending 3 or fewer days there. The median juror deliberated for no more than 1 hour, with 84% deliberating for 2 hours or less.
So these cases were relatively short, with most deliberations lasting less than 2 hours. This might mean that the results won't hold for longer deliberations.

But on to the results. Tables 2 & 3 clearly lay out the correlations between the variables, and the findings are quite striking. [Nota bene: in psychology, and social psychology in particular, a correlation of .25 is actually pretty big. It's notoriously difficult to isolate the specific factors that lead to these correlations, so finding these numbers in a survey study pretty strongly supports the conclusions. And when interpreting the findings, keep in mind that a correlation of .30 explains about 10% of the total variance. Still not very big, but significant.] Both political confidence and interest in the trial led to greater ratings of deliberations about the facts of the case and the freedom of expression in deliberation. Greater trust in the jury system was linked with jurors that treated each other more respectfully, and greater political partiality was correlated with more debate over the instructions. Absolute levels of political knowledge and education weren't correlated with anything, but the differences between jurors definitely had an effect: across the board, differences in political knowledge (and in political confidence) were associated with reduced deliberation; however, greater differences in education levels led to jurors rating respectful listening higher. Finally, when there were bigger differences in interest in the trial, jurors discussed the facts of the case more.

So what implications do these findings have? Well, it's clear that, in this sample, juror characteristics affected the quality of the deliberation, which in turn affected the satisfaction of the jurors. But from Table 1, it's clear that the ratings for all the aspects of deliberation were generally high across the board. So did these factors really affect any of the verdicts? And, as I mentioned before, can these results generalize to larger juries or more complex cases? Also, how well do these findings represent and relate to verdict juries - those who start the deliberation with a preliminary vote? The book suggests that there are definite differences between verdict juries and juries who wait until after deliberation to vote. Clearly, there's a lot of complicated issues arising from jury studies. Anyone inspired to get a subscription to Jurimetrics?

Monday, November 5, 2007

Article #3 - Testosterone and Recognizing Social Cues

van Honk, J., & Schutter, D. (2007). Testosterone Reduces Conscious Detection of Signals Serving Social Correction: Implications for Antisocial Behavior Psychological Science, 18 (8), 663-667 DOI: 10.1111/j.1467-9280.2007.01955.x

Blogging on Peer-Reviewed ResearchHope everyone did well on the midterm and had a nice, relaxing weekend to recuperate. Here's a quick little article to ease you back into normality.

The authors of this study wanted to look at the effects of testosterone, so, in a clever bit of experimental control, they used an entirely female sample. Men's normal testosterone levels vary too much to extract any reliable information about their effects, but women have low enough levels to be able to control them entirely by injection. By doing just this and having the women rate faces displaying varying levels of emotion, they found that testosterone reduces our sensitivity to the emotions of fear and anger. This would have direct ramifications on a person's ability to use social learning: if a disapproving look is harder to discern, then it might be harder to learn not to kick your friends. The possibility, then, is that this determines a lifelong disadvantage, where a person continually falls further and further behind of his peers in social and moral development, possibly resulting in criminal behavior.

To tie these findings in to some developmental literature, researchers who study bullying in children and adolescents find that one common factor among many bullies is that they have trouble interpreting emotions. They'll see someone smiling and assume they're being mocked, for instance. Obviously, these two findings raise an obvious hypothesis: bullies are more likely to have high testosterone levels. The testosterone doesn't make them more aggressive, per se, but it does increase their exposure to situations where aggression seems to them to be a reasonable response.

In fact, there's very little to link aggression and testosterone by a direct causal mechanism. So it might then be asked (to bring it back to legal applications) what do these findings mean for defenses based on rage? You might recall that people have tried to get off on charges by claiming they were not responsible for their actions due to temporary chemical imbalances, such as the infamous "twinkie defense." A sudden surge in testosterone might impair a person's ability to reason and function socially, but is not likely to directly trigger aggression. How much control do people have over their own behavior, and how liable should they be over how they react to situations?

One more dilemma that might be added to the four in the textbook could be that between Biological Determinism and Free Will: psychology assumes that there are many chemicals, genes, developmental patterns, etc. that determine how we behave; in the legal system, however, it is assumed that people have the ability to determine and direct their behavior except for very rare circumstances (i.e. <1% of cases result in a successful NGRI verdict). Has the line between these two positions been drawn in the right place?