discussion 4- social perception I
Jenny Perella
to trustworthiness. They predicted that the amygdala, orbiotfrontal
cortex, fusiform gyrus, and the superior temporal sulcus would be
active in trustworthiness assessments. I wasn’t a big fan of this
article – I felt it left out a lot of relevant information (but I may
have just done a poor job reading it). First of all, I really didn’t
like that their methods section was at the very end of their paper.
This seems to be a fairly common tendency in these types of articles,
and for some articles it works. Not for this one. I spent a good deal
of time while reading the article trying to understand what they were
doing and why… I guess I should have skipped ahead and read the
methods section first. However, I didn’t think that the methods
section was very thorough, and I was still left with questions about
the study. Specifically, the authors briefly inform the reader that
trustworthiness assessments ‘can be’ implicit (no theory explaining
this), and then later seem to be comparing implicit to explicit
assessments of trustworthiness… I was unclear how the implicit/
explicit distinction was measured. What tasks did they use? I also
thought they could have gotten better results/more variability if they
had assessed degree of trustworthiness (on a scale) rather than in yes/
no format.
In their results, Winston et al. reported that happy faces were rated
significantly more trustworthy than neutral faces. While the
difference may be statistically significant, I wasn’t convinced that
it was meaningful – happy faces had a mean trustworthiness rating of
4.0 and neutral faces had a mean rating of 3.9. What exactly does this
difference really mean? Standard deviations were not reported.
Moreover, to me only the angry faces graph showed an obvious
correlation. I know the line of best fit was statistically calculated,
but I’m not sure I’m convinced there is a true relationship in the
other emotions.
I thought the Anderson et al (2003) paper was much better than the
Winston paper. This study examined the role of the amygdala in
processing fearful, disgusted, and neutral faces as a function of
attentiveness. One thing in particular I found interesting was the
authors’ statement that automatic amygdala processing “may be confined
to more coarse affective properties of faces, such as their valence or
arousal/intensity” (p. 5631). It would be interesting to use a picture
set with standardized valence and arousal scores, like the IAPS, to
see if this theory holds out.
Adams et al (2003) used fMRI to see if amygdala activation to anger
and fear varied as a function of gaze direction, arguing that the
source of threat matters. The authors predicted and later found that
the magnitude of amygdala activation was greater in response to
ambiguous (averted anger gaze, direct fear gaze) rather than clear
(direct anger gaze, averted fear gaze) threat. I wasn’t completely
clear on the reasons for this prediction; the authors seemed to assume
that this hypothesis makes sense implicitly, and to me it doesn’t.
What’s more, the authors do not address disparity of results between
left and right amygdala -- they found these results in the left but
not right amygdala. Was this expected or not, and why/why not?
I thought it would be interesting to use facial morphs, morphing from
fear to anger and vice versa, to see how brain activation gradually
changes while shifting from one mental process to another. That is,
what areas of the brain are implicated in the processing of the
progression of fear to anger?
Coccaro et al (2008) examine brain activation associated with IED
patients as compared to healthy controls. Here, the authors did find
effects in response to direct anger gazes. Adams et al (2003) argued
that previous literature could not often find effects because it was
more important to measure the source of the threat, but that doesn’t
seem to be the case in this particular study. Specifically, this study
examined amygdala-OFC activation in response to emotional
informational processing. What about activation during the experience
of emotion? What about quickness/readiness to experience an emotion,
specifically anger? IED are presumably quicker to anger, and
experience anger at a greater magnitude, but again, the process would
be interesting to investigate.
Stuart Daman
this week. As a sidenote, I dread this task a bit, because I feel that
the Methods sections of neuroscience reports are not written as well
as those that I normally read (I am more sure of this now that I have
read more of these studies). Not only to do authors make use of many
acronyms (sometimes without telling us what they are for), but it's
full of much techno babble. I understand why this babble is necessary,
but it's tough to dredge through it. On a complementary note, the
authors do a gReaT job with their techno babble, but then often do
what I feel is a bad job explaining what the participants actually do
(often the "procedure" subsection). In addition, I still hate how
neuroscience journals put Methods sections at the end of reports. I've
also seen more made-up words this week than ever, although perhaps
they're simply jargon with which I am not familiar (e.g.
"conspecifics", "convolve", ) Onto the article commentary!
The Winston et al. (2002) paper showed differential activation in the
amygdala for perceiving trustworthy and untrustworthy faces, also in
comparison to no change in activation for age judgments. I am confused
by many studies when they say that they are looking for what mediates
this and that, but then never conduct what I know to be a mediational
analysis. For example, this study looked at how labeled emotional
expressions interacted with trustworthiness. This is NOT mediation,
this is moderation, we all know that. Interactions show us how
differences in factor X differ across levels of another factor Y; this
is moderation, not mediation. In their discussion, they suggest that
the right insula may act as a pre-processor of information, providing
us with "gut feelings". I'm sure all of us have had gut feelings
before (not to mention the notorious policeman trust in his gut), and
it's really neat to see that there is suggestion of a possible
biological underpinning for this phenomenon. Their "Debriefing"
subsection confused me greatly, because I don't understand what part
of it was a debrief at all. Furthermore, the "Data Analysis" section
seemed like it was solely intended to confuse, I could not take much
away from it other than jibberish. Before I can comment on it in
class, I'll probably have to reread it while in an fMRI machine and
alter, I mean model, the data by convolving a delta function at each
event onset witha cononical hemodynamic response function and its
temporal derivative to create regressors of interest; whatever the
hell that means.
The Anderson et al. (2003) paper looked specifically at the
automaticity of processing threat stimuli. I found it interesting how
they touched on information provided by Gray (1987), a theory which I
would love to greatly familiarize myself some day, regarding stress
and action. They basically found that threatening stimuli were
responded to in the same way, whether or not they were attended to,
suggesting that they are processed automatically. The other thing they
showed was that disgusting stimuli were processed differently.
However, from the way they say things, it took me a while to take this
away from it. On more than three occasions, I got confused because I
thought they said something contradictory to something that they said
earlier. I was not always able to resolve the apparent discrepancy
either, as in the case of these two sentences:
"Automaticity appears unique to amygdala processing of social signals
of fear." & "Thus, amygdala automatic processing is not specific to
fearful faces..." (p. 5631).
These two sentences are contradictory, as I understand them. The
latter part of the second sentence may alleviate this contradiction,
but writing it as they do is no better than sloppy.
Their comments on attentional load are also interesting. Much research
has examined how mental processes differ when under a cognitive load,
and it's funny to see it germinate into neuroscience as well. It makes
me curious whether there is any sort of similarity in patterns of how
changes in occur when under cognitive load (vs. no load or more/less
load). Is it possible that there is some sort of pattern for how
cortex activation differs depending on cognitive load? Or are mental
processes too different for such a commonality to exist? Perhaps some
more general processing region is activated differently, or perhaps
there is a multitasking area of the frontal cortex somewhere.
The Coccaro et al. (2006) paper irritated me by often providing
results and not elaborating on them. On the other hand, their results
suggested that individuals with aggression disorders (at least non-
depressive IED) are related to specific changes in mPFC activation.
This would tie-in rather well with things we discussed a week or two
ago about the self, and by extension empathy or imagining things from
another's perspective. When a person thinks about intimate or close
others, activation occurs, and less when imagining less close others.
The suggestion here seems to be that aggressive individuals also have
lesser activation here. This connection between lack of empathy and
aggression in terms of brain activation is not well laid out in this
research, but perhaps I'm making gross and not yet grounded
speculations. I was also confused at they they used diagnostic
criteria that go above and beyond that of the DSM, I wish the author
had commented in more detail on why simple DSM criteria was apparently
not sufficient (although it may be important to note that this extra
criteria was all published by the first author of this article). I
also got flustered when, at the very end of the "Tasks and Materials"
section, the author referred to their procedure as an experiment--they
did not experimentally manipulate anything. They also mentioned
something "to rule out structural abnormalities and anatomical
localization", why? This sort of information could be crucial to
what's going on! For example, I know that there is research that
persons with schizophrenia have smaller brain ventricles, which is
basically a structural abnormality. Unless we know for sure that
structural abnormalities are not the basis for a disorder, they should
not be ignored. The discussion section was good in this paper, but
lacked some of the insight and extra thought I feel we've seen in many
other papers already, although they did include limitations, where
many others have not. The mentioned TMS study also sounds cool!
The Coccaro et al. (2006) and Carter & Pelphrey (2008) studies both
used goggles to present stimuli in an fMRI machine. At first thought,
I felt this was a great improvement over some of the other
presentation methods. On the other hand, once I thought about it
further, I quickly became concerned with how sure researchers are that
these goggles do not interfere with fMRI recording. I suppose that
this is what is meant by "MRI-compatible" (mentioned in Cocarro et
al., but not Carter & Pelphrey). Carter & Pelphrey (2008) also
employed the simplest paradigm, because participants did not respond
to stimuli behaviorally, they simply viewed them. This certainly
removes some of the various sources of error introduced by the other
methods. On the other hand, they also had the smallest sample(s),
which is discouraging (largest n = 10). They also specifically used
males in their stimuli. We all know that there are differences in how,
how often, and when males and females use/experience different
emotions; so wouldn't it be useful to see if or how people (males and
females) respond differently to males and females making angry or
happy faces at us? This would certainly be a great avenue for future
research. This was also the only study that provided information with
which I was sure to say that they committed the nonindependence error.
Other points of their analysis posed red flags to me as well, such as
not using uncorrected tests. I also was rather interested in their
employment of eye-tracking, another arguably neuroscientific method.
It amazing to see what we can learn simply by where people look at
specific stimuli.
Monicarodr
beka strock
So first, while I was procrastinating this week, I took a look at the newly posted docs, and loved these quotes from the webslides, which I thought addressed some of our recent discussions well:
Louvrain GLM 3.3, slide 28 (RE: explanation of independence error, hahaha!): “Egregious example: Identify Area X with contrast of A > B. Do post hoc stats showing that A is statistically higher than B. Act surprised.” AWESOME. LOVE it. :)
Louvrain 1A FMRI, slide 16 (RE: BOLD signal/physiological/theoretical explanations): “We sort of understand this (e.g., psychophysics, neurophysiology) [stimulus/neuronal activity]…We’re *&^%$#@ clueless here! [neurovascular coupling/hemodynamic response]... We sort of understand this (MR Physics) [detection by MRI scanner/BOLD response]” also loved this summary.
First, I really liked that they looked at clinical samples with specified brain areas, I think that gives this whole area of study much more support than the “look it lights up when people do this” method.*
*In the Cocarro et al, they discuss the methodological corrections in general terms which do not really explain how they made some of the adjustments (i.e., correct for head motion, smoothing, mapping, etc. p170-1, p174) making it sound like their estimates are conservative, yet on p174 they state that they used a lower than usual uncorrected threshold of p<.0005. Also, it seems that they do not address the independence error because they were using ROIs—but ultimately they also did a whole-brain analysis as well, so not sure how that helps their case. Aside: I don’t know if anyone watches the TV show Dexter, but he got an MRI done on the show (season 2) which referenced the brain differences in psychopaths, so even the public media has picked up on this stuff—though it wasn’t a detailed reference and certainly not to this specific study. Just thought that was interesting. ;)
*The Anderson et al methods section was much more thorough, yet again the “corrections” are covered in about 1 sentence on p5628 that very generally states the usual items accounted for. However, as a pro, these authors actually stuck to just ROIs as hypothesized instead of publishing the usual page-long summary activation chart, so kudos to them.
** Even correlating ratings with activation of positive or negative evaluations using the specified process is also superior to the “lights up” method in my opinion, so that is a strength of this analysis. Despite the fact that the Winston et al did not examine only the ROI (therefore committing independence error while evaluation overall activations in the brain), they did specify a different correction factor and p-level for each type. An unusual extra step was added of evaluating a different model and reevaluating facial expressions with a different subject pool to ensure independence of facial expression from trustworthiness. I have mixed feelings on both steps. While it is good to consider other models, but just stating “all statistical conclusions drawn from original model” doesn’t fix the fact that you ran the analysis and would have reported it if it was better in all likelihood. I like the addition of independently evaluating facial expressions to lend credibility of the expression ratings, but I would think the original subjects ratings are important for other reasons, i.e. their activations associated with those judgments and I’m not sure what claim they are trying to support with this evidence.
Carter & Pelphrey incurred many of the standard criticisms of the “light up” experimental designs, exemplified by the phrasing, the statistical contrast, “revealed a network of brain regions that responded more strongly to the threatening as compared to the friendly condition” (p 156). That’s a NICE way of putting it. Only about half of the listed brain regions that were “revealed” were on the original ROI from what I can tell. In addition, there are later 4 ROIs listed whereas only 2 are defended theoretically (STS & amygdala). While they did not address their methods for the whole brain scan in great detail (disapprove!) leading me to believe the independence error was in full effect, at least the authors didn’t flaunt the usual “chart of activations” (approve!), and they did back up the theoretical basis for the additional areas they found after the ROIs were confirmed.
Oh, and Adams et al…what methods?
Lindsay Morton
As a whole, this week’s articles, while in conceptual terms seemed interesting, were somewhat uninspiring in terms of research. My thanks go to Jenny for the recommendation to read the methods of Winston, Strange, Doherty, and Dolan (2002) before the results. Though this made the article more understandable in that it helped clarify the distinction between explicit (i.e., stating the face is trustworthy) vs. implicit (i.e., making a judgment about facial features while automatically, unconsciously evaluating trustworthiness), there were two main factors that undermined the legitimacy of the findings. First and foremost is the good ol’ non-independence error. Second, we have the fact that the raters first made categorizations (i.e., trust and student) while being scanned, and then they rated all stimuli on trustworthiness. Does anyone else think that previous exposure and recent judgments might have biased the results? It really is too bad that they just didn’t pilot the faces with another group of 16 to use in their analysis.
One of the things that I had the hardest time swallowing this week was the claims of Carter and Pelphrey (2008). In my opinion, presenting animated displays of happy vs. angry faces does not provide any information about dominance/affiliation perception. How can any statements about dominance and affiliation be made if you don’t ask participants to explicitly indicate the intentions of the animated man? True, the authors cite Knutson (1996), who apparently has done work on facial expressions and trait inferences, but without covering what was found in that research and how it was performed, I remain an unconvinced critic.
The sad part is that as I read the introduction to this article, I was excited to see what insight Carter and Pelphrey (2008) might provide into the field of social perception. Then I hit the methods. First I think we can all agree on the fact that this research is one of the many that suffers from non-independence. Even putting that aside, as social psychologists, the need to recognize the issue of gender stands out to me. The participants (all 10 of them) were half female, being exposed to a happy vs. angry male figure. It seems plausible that a study in which perceptions of emotion and intention are entailed would take into account the possibility that mixed-gender interactions could play a role in the participants’ reactions. Add to this the unknown factor of how animated vs. actual interpersonal interactions may differ in significant ways and there’s a new list of questions and concerns that arise.
On the other hand, the work by Coccaro, McCloskey, Fitzgerald, and Phan (2006) struck me (in a better way) on several points. First, I think that this type of neurological work could really be useful in the case of classifying disorders. The DSM (the most popular diagnostic manual) is literally always under revision. Criteria for the disorders are not static and definitely not mutually exclusive. Through the use of brain imaging, we may be able to find the correlates of specific disorders to distinguish between them and to determine the best form of treatment. (On a side note, I have to point out the happy fact that the authors clearly highlighted that brain differences found in this study do not imply disorder causality). Over the past couple of classes, you may have also heard me lament that fMRI cannot be used to look at communication across brain regions. For this reason, I also thought the methodology of relating brain activation across different regions was a fascinating way of testing possible regional connections with fMRI data (though I admit that the findings may have to be taken with a grain of salt). Last, I don’t remember who was assigned anatomy this week, but I thought the amygdala and OFC bidirectional connection (discussed well on pp. 174) might be one that plays a role in some of our research (e.g., Jenny and Stuart’s work on stimuli evaluation) and might be cool to discuss more in class.
David Dinwiddie
in presenting their article. As Stuart mentioned they say that
"Automaticity appears unique to amygdala processing of social signals
be that the automaticity is unique to any intense face. Therefore we
don't really know any more about the amygdala than we did before and
so I have a hard time seeing what this article does to add to the
field on its own. However when couple with the Whitney, Carter and
Adams articles it seems clear that there is greater activation in the
amygdala when processing anger displays or signs of untrustworthyness.
In the Whitney article they showed that people trusted people less who
had angry faces and so maybe the amygdala activation is strongest just
when one feels vulnerable and unable to trust. This seems to be
supported but I still have a bit of a problem in extending such
artificial lab results to a real life setting. Seeing an angry face on
a screen is much different than seeing a face of someone right infront
of you who you think might hit you.
The Coccaro article was probably my favorite one of these articles. I
agree with lindsay in saying that brain imaging could be a very
valuable tool in the future in diagnosing and potentially helping
people who suffer from psychological disorders. Knowing how different
parts of the brain react to stimuli isn't really useful unless we use
that knowledge to help people. And yes Beka I do watch Dexter (best
show ever and actually very realistic) and had the same thought when I
read this and I know they actually do that in real life as well
although I think with psychopathy it's just the size of the different
regions which are diffferent rather than the blood flow. Maybe knowing
more about blood flow in psychopaths would help us to find cures for
psychopathy which have never been found before.
> > On Wed, Feb 17, 2010 at 7:38 PM, Monicarodr <rodr....@gmail.com> wrote:
>
> >> I posted some material that may help with methods.
>
> >> On Feb 17, 2:21 pm, Stuart Daman <daman.stu...@gmail.com> wrote:
> >> > Here's my personal reminder that I'm going to emphasize the Methods
> >> > this week. As a sidenote, I dread this task a bit, because I feel that
> >> > the Methods sections of neuroscience reports are not written as well
> >> > as those that I normally read (I am more sure of this now that I have
> >> > read more of these studies). Not only to do authors make use of many
> >> > acronyms (sometimes without telling us what they are for), but it's
> >> > full of much techno babble. I understand why this babble is necessary,
> >> > but it's tough to dredge through it. On a complementary note, the
> >> > authors do a gReaT job with their techno babble, but then often do
> >> > what I feel is a bad job explaining what the participants actually do
> >> > (often the
>
> ...
>
> read more »- Hide quoted text -
>
> - Show quoted text -
Camille Barnes
One issue I had with the Coccaro article was that each person in the
experimental group had at least 1 axis II personality disorder. So, I
don’t understand how they can determine that the brain patterns they
saw were characteristic of IED patients and not of axis II personality
disorders; this seems to be an important confound. However it does
seem consistent with the symptoms of the disorder, that individuals
with IED would have greater amygdala activation to anger and have
reduced OFC activation which aids in emotional regulation. This brain
activation pattern seems consistent with the angry outbursts of IED
individuals.
In the Anderson article, I did not like that they selected the FFA as
the area that was activated when looking at the faces, and the PPA as
the area that was activated when looking at houses. The Carter
article looked at similar areas, but chose them according to the brain
coordinates. Also, since there have been studies indicating that
people process houses and faces differently by parts versus whole.
Because of this research, I think it would have been interested to
determine what parts of the stimuli individuals were attending to.
Also, I did not understand why for some areas, such as the contiguous
bilateral ventral striatal region, they only reported results for
disgust and not for fear. It was interesting how there are different
brain activation responses for the same stimuli depending on whether
processing was intentional or not. It is interesting that the
Amygdala seems to process the intentional fear faces and the
unintentional fear faces to the same degree.
In the Carter, & Pelphrey article, I liked how the experimenters used
both FMRI and eye tracking measurements. Even though they found no
effect of emotion on eye gaze I think it was still useful and maybe
would provide more information if actual people were used instead of
animated characters. Though I thought using the animated characters
was a very good way to provide control. It is interesting that our
brains respond in a similar manner to characters that we consciously
know are not real, as they would to actual people.
The Adams article only found Amygdala activation in the left portion
of the Amygdala. Other articles, did not find the fear/anger reaction
to be lateralized. This article not only included indicators that a
threat is present through facial expressions but also an indication of
where the threat was located through direction of gaze. It added an
additional piece to the puzzle that the other articles had not yet
explored.
I had a little trouble following the point of the Winston article. I
understand that judgments of trustworthiness aid survival. But I
didn’t understand what aspects of a face determine if it is
trustworthy or not. Trustworthiness is not like expressions fear or
anger which have been shown to be cross culturally similar. Telling
people to rate faces on trustworthiness does not mean they are rating
them on the same thing, or that similar things are happening in the
brain. I just did not think there was adequate theory explanation, to
make this study viable.
> ...
>
> read more »
Jennifer Vosilla
I focused on the different areas of
the brain discussed in the articles so I thought I’d give an overview of where
the areas are and their believed functions. Also I've attached a word document with some pretty pictures pointing out each area, which I will try to remember to bring copies of for tmrw.
Amygdala: found on pg. 123 of our coloring book
The amygdala is almond-shaped and part of the limbic system. It is located in the medial temporal lobe, anterior (toward the nose) to the hippocampus. According to the articles we read this week, the amygdala is involved in judgments of trustworthiness, processing threat-related ambiguity, and automatic fear processing. Also, patients with amygdala lesions displayed impulsively aggressive behavior. The article looking at Intermittent Explosive Disorder (IED) opened my interests in the amygdala’s relationship in other psychological disorders. There is significantly greater left amygdala activation in patients with Borderline personality disorder (Donegan et al., 2003), significantly increased response in the amygdala in patients with severe social phobia, and significantly larger right amygdala than left amygdala in patients with Schizophrenia (Tanskanen, 2005).
Orbitofrontal Cortex (OFC): pg. 195 of the coloring book
The OFC is part of the prefrontal cortex that is said to be involved in decision-making and sensory integration. Patients with lesions in the OFC have been noted to have limited control over emotions, poor decision-making, & limited awareness of moral implications of actions (Coccaro et al. 2007).
Superior Temporal Sulcus (STS): p. 115
The STS is located in the temporal lobe and it separates the superior temporal gyrus from the middle temporal gyrus. It is involved in the perception of the direction of someone’s gaze and biological motion (visual perception of moving, animate objects).
Anterior Insula: Sorry, I couldn’t find this in the coloring book.
The anterior insula is part of the insular cortex deep within the lateral fissure between the temporal lobe and the frontal lobe. It is thought to be involved in the processing of sense of disgust to smells and to the sight of contamination and mutation, norm violations, emotional processing, empathy & orgasms.