The article Vul et el did was on studies of functional MRI (fMRI)
scans which use voxels and apply some sort of gaussian smoothing to
determine which ones to use. Same exact method Firth et al used.


http://www.edvul.com/pdf/VulHarrisWinkielmanPashler-PPS-2009.pdf

These basic steps common to most fMRI data analyses yield
matrices consisting of tens or hundreds of thousands of numbers
indicating activation levels in different voxels. These can be
(and indeed generally are) displayed as images. However, to
obtain quantitative summaries of these results and do further
statistics on them (such as correlating them with behavioral
measures—the topic of the present article), an investigator must
somehow select a subset of voxels and aggregate measurements
across them. This can be done in various ways. A subset of voxels
in the whole brain image may be selected based on purely anatomical
constraints (e.g., all voxels in a region generally agreed to
represent the amygdala, or all voxels within a certain radius of
some a priori specified brain coordinates). Alternatively,
regions can be selected based on ‘‘functional constraints,’’
meaning that voxels are selected based on their activity
pattern in functional scans. For example, one could select
all the voxels for a particular subject that responded more to
reading than to nonlinguistic stimuli. Finally, voxels could be
chosen based on some combination of anatomy and functional
response.

In the articles we are focusing on here, the final result, as we
have seen, was always a correlation value—a correlation between each
person’s score on some behavioral measure and some summary statistic
of their brain activation. The latter summary
statistic reflects the activation or activation contrast within a
certain set of voxels. In either case, the critical question is,
‘‘How was this set of voxels selected?’’ As we have seen, voxels
may be selected based on anatomical criteria, functional criteria, or
both. Within these broad options, there are a number of additional
more fine-grained choices. It is hardly surprising,
then, that brief method sections rarely suffice to describe how
the analyses were done in adequate detail to really understand
what choices were being made.

We display the raw results from our survey as the proportion of
studies that investigators described with a particular answer to
each question (see Fig. 2). As some questions only applied to a
subset of participants, we display only the proportion of the
relevant subset of studies.
The raw answers to our survey do not by themselves explain
how respondents arrived at the (implausibly high, or so we have
argued) correlations. The key, we believe, lies in the 53% of
respondents who said that ‘‘regression across subjects’’ was the
functional constraint used to select voxels, indicating that
voxels were selected because they correlated highly with the
behavioral measure of interest.

Figure 3 shows very concretely the sequence of steps that
these respondents reported following when analyzing their data.
A separate correlation across subjects was performed for each
voxel within a specified brain region. Each correlation relates
some measure of brain activity in that voxel (which might be a
difference between responses in two tasks or in two conditions)
with the behavioral measure for that individual. Thus, thenumber of
correlations computed was equal to the number
of voxels, meaning that thousands of correlations were
computed in many cases. At the next stage, researchers selected
the set of voxels for which this correlation exceeded a
certain threshold, and reported the correlation within this set of
voxels.

What are the implications of selecting voxels in this fashion?
Such an analysis will inflate observed across-subject correlations and
can even produce significant measures out of pure noise. The problem is
illustrated in the simple simulation displayed in Figure 4. First, the
investigator computes a separate correlation of the behavioral measure
of interest with each of the
voxels (Fig. 4a). Then, he or she selects those voxels that exhibited
a sufficiently high correlation (by passing a statistical threshold;
Fig. 4b). Finally, an ostensible measure of the ‘‘true’’
correlation is aggregated from the voxels that showed high
correlations (e.g., by taking the mean of the voxels over the
threshold). With enough voxels, such a biased analysis is
guaranteed to produce high correlations even if none are truly
present (Fig. 4). Moreover, this analysis will produce visually
pleasing scattergrams (e.g., Fig. 4c) that will provide (quite
meaningless) reassurance to the viewer that s/he is looking at a
result that is solid, is ‘‘not driven by outliers,’’ and so on.


.

On Sun, Feb 19, 2012 at 3:09 PM, Larry C. Lyons <[email protected]> wrote:
>
> I've been stying out of this discussion mainly because I've been reading
> the article that Sam provided a link to.  I'm still going through the
> article, but first and foremost it has nothing to do with the topic at
> hand. Vui et al were looking at the unexpectedly high correlations between
> personality traits and vartious systems within the brain. Last I checked
> the psychological constructs known as personality traits are not political
> orientation. Even someone who has taken an intro to psychology course at
> the high school level should have picked up that one. Did ytou actually
> read the article Sam?
>
> Second the main thrust of the criticism was how these personality traits
> were correlated with problematic measures of brain activation. The study I
> mentioned looked at the differences between specific groups. In  other
> words a very different form of statistical analysis was used. This
> meta-analysis you cited does not address that.
>
> There are other criticisms I could make about the methodological approach
> the authors take. However everyone else would find it completely boring. I
> also think that this study is important for entirely different reasons than
> Sam gave and those reasons are not at all relevant to this discussion,
> mainly involving recent research in the neuophysiology of personality.
>
> In a nutshell Sam you need to actually read and attempt to understand the
> material you present in cases like these. That way you do not come across
> with so much egg on your face.
>
>

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~|
Order the Adobe Coldfusion Anthology now!
http://www.amazon.com/Adobe-Coldfusion-Anthology/dp/1430272155/?tag=houseoffusion
Archive: 
http://www.houseoffusion.com/groups/cf-community/message.cfm/messageid:347121
Subscription: http://www.houseoffusion.com/groups/cf-community/subscribe.cfm
Unsubscribe: http://www.houseoffusion.com/groups/cf-community/unsubscribe.cfm

Reply via email to