Jimmy Cagney in White Heat (1949): "Made it, Ma! Top of the world!"
Jordan Ellenberg has a great article up on Slate on “The Mathematics of Narcissism.” There are two related themes running through it.
One theme surrounds the overhaul of the current Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition Revised (DSM IV-R), the “Bible” for psychiatric professionals. The DSM is the way in which psychiatrists and psychologists classify mental disorders into clumps so that we can use a common language in talking about patients and diseases. The other is about the ranking of graduate programs. Finding the commonality between these two themes is a great reason to read the article, which I highly recommend.
The overhaul of the DSM is an infrequent and important event in psychology and psychiatry. In the immortal words of Vice President Joe Biden, “It’s a big fucking deal!” Basically, the revisions in the DSM codify and standardize the current models of how the mind works. In order to standardize the classification and treatment of psychiatric disorders, we need a book that describes to the best of our current ability what is wrong and what is right.
For example, homosexual behavior was removed from the DSM-IIR in 1973. This reflects not so much a change in how homosexual behavior is processed by the brain (at least, we would certainly hope not) but rather a change in how science and scientists perceive the workings of the human brain. Are there politics in such decisions? You betcha, and we’ve discussed some of these in previous 538 Refugees posts. Still, one would hope that, over time, we develop models of how the human mind works that do a better and better job of explaining experimental findings.
As we’ve discussed before, science may progress by testing hypotheses (fine-grained, detailed), but it also works by coming up with explanatory models (coarse-grained, universal). As our tools and observations improve, so do our explanatory models.
The news that narcissism is to be removed from the next edition (DSM-V) and the rankings of graduate programs in Ellenberg’s view (and in mine) have similar problems. What we, as humans want, is a simple ranking. AFI has the Top 100 Movies of All Time. (White Heat doesn’t make the cut, but “Made it, Ma! Top of the world!” is the 18th best quote in the movies and Cody Jarrett is the 26th most villianous villian.) Which is the “best” graduate program? Which is the “best” measure of narcissism? (“He’s a 98% narcissist, but I’m only an 83% narcissist.”)
In political terms, who is the “best” Republican Presidential candidate in 2012? What is the “best” solution to reduce the deficit?
In reality, that sort of analysis doesn’t do a very good job of what we need to do. We need to say, “these are characteristics most often shared by people who _____________.” To wit:
Diagnostic criteria for 301.81 Narcissistic Personality Disorder
A pervasive pattern of grandiosity (in fantasy or behavior), need for admiration, and lack of empathy, beginning by early adulthood and present in a variety of contexts, as indicated by five (or more) of the following:
- has a grandiose sense of self-importance (e.g., exaggerates achievements and talents, expects to be recognized as superior without commensurate achievements)
- is preoccupied with fantasies of unlimited success, power, brilliance, beauty, or ideal love
- believes that he or she is “special” and unique and can only be understood by, or should associate with, other special or high-status people (or institutions)
- requires excessive admiration
- has a sense of entitlement, i.e., unreasonable expectations of especially favorable treatment or automatic compliance with his or her expectations
- is interpersonally exploitative, i.e., takes advantage of others to achieve his or her own ends
- lacks empathy: is unwilling to recognize or identify with the feelings and needs of others
- is often envious of others or believes that others are envious of him or her
- shows arrogant, haughty behaviors or attitudes
But people don’t do very well with such classification schemes. They crave black-and-white answers in cardinal order (11th best quote: “What we’ve got here is failure to communicate.”). Unfortunately, in the real world of politics, as in the real world of science, there are no such answers—only probabilities.
I believe that we need to work towards a world where we say, “these are the features we’re looking for in (health care reform/tax policies/deficit reduction schemes)” and then pick the plan that comes closest to the ideal. Others may see it another way, which is why we have a discussion section.
Related Articles
- Is It Time to Redefine Narcissism? (thedailybeast.com)
- Narcissism in High-Functioning Individuals – Big Ego or Severe Disorder? (brainblogger.com)
- Narcissism: The Malady of Me (nytimes.com)
- Bullish: How to Tell if You’re a Narcissist (and How It Might Be Holding You Back) (thegloss.com)
- The common statistical thread between psychiatric diagnosis and grad school rankings. (slate.com)
- Personality Disorders Shakeup in DSM-5 (psychcentral.com)
- It’s All About Me: But Is Narcissism A Disorder? (npr.org)
- Mind: A Fate That Narcissists Will Hate: Being Ignored (nytimes.com)
538 Refugees 
Interesting how this coincides with the presidency of Barack Obama. Can’t have a President be found clinically to suffer from a mental illness, now can we?
;^)
[For those who have anger issues, the winky smiley face means this was a joke]
Interesting that the first post on a thread about narcissism is by Bart.
“has a grandiose sense of self-importance (e.g., exaggerates achievements and talents, expects to be recognized as superior without commensurate achievements)”
BTW, Bart, I find your attempt at humor in extremely poor taste.
Back on thread topic — Thanks for pointing this out, Monotreme. I’ve used clustering methods a few times in my professional career, with some success, and found the link very interesting.
Here’s the thing that’s left out of the article, which is very important. The classification or diagnosis of either a graduate school for mathematics or a mental condition depends on the use to which the result will be put. If the result is to be used as a method for allocating funds, then a dimensional, continuous ranking makes sense. If the result is going to be used to decide where you are going to go, say for treatment (or for graduate school), then clustering makes more sense.
In other words, if you are going to have to choose, then cluster. If you are going to adjust, or balance, then dimensionalize. (I made up that word.)
Perhaps an inability to decide on the use of the classifications was the stumbling block.
Of course, given the also-well-known correlation between the study of mathematics and mental illness, perhaps a person should choose graduate schools and treatment regimens simultaneously, to save time.
(speaking as a Graduate of one of them 10 elite schools — in mathematics).
Shortchain,
That’s an interesting point. I’ll have to give that some thought, but I think you’re right. Clustering methods should work best when making a unitary choice.
The problem with cardinal methods (I made that up, too) is that when you draw a line, there isn’t much difference between the 23rd best grant (for example) in a stack of grants to be funded and the 24th best grant. Still, it can’t be helped.
For grant funding, I always thought that it should be a sliding scale, so the best grant(s) get fully funded; the next batch get 90% funding; and so on. That would diminish the “cliff” effect of being the 24th best when only 1-23 can be funded.
I don’t know if such an approach would work for the Federal budget in general, however.
I think all successful politicians, at all levels, would score pretty high on the narcissism scale.
Monotreme,
I think clustering is also good if you have to select a few items from a stack. It breaks down when you have to select more items than you have in a cluster.
We might have a theorem there about when to use cardinality and when to cluster.
I use clustering in assigning grades, BTW.
Don’t forget that these are not mutually exclusive approaches. In fact when clustering complex data one almost always reduces the dimensions first. I am most familiar with these methods from analyzing complex ecological data.
Anyway I have my doubts about how this can be useful in socio-economical-political decision-making where outcomes of different actions cannot be predicted with much certainty. Yes, you could probably rank different approaches to an issue by how much they cost or cluster them in meaningful ways but is either likely to tell you how to achieve full employment while reducing dependance on foreign energy sources?
I have over the years worked with the DSM-4 and run up to same more than I like to think about. While not a perfect tool by any measure it was after considerable use a fairly understood process and in the hands of the right people produced results (not sure if that is the best wy to say it) that most involved could understand and to some degree accept. In the hands of the wrong people it could and often was manipulated to exclude individuals on the one hand to save money and on the other for a while to include individuals that really didn’t belong.
I am concerned that the DSM-5 or what ever it will be called has been held up as long as it has been. We keep hearing that it will be out soon but soon never seems to get here. As I have 2 grand children that fall into or next to the Autism spectrum (and the first sorry SOB that even hints that Autism isn’t real isn’t going to like me much) this has now become a much more personal exercise. The cluster of people that will fall into the Autism spectrum has spooked the people that will hve to pay to get these individuals through life. I see behind the scenes efforts tolimit that cluster. I hope I’m wrong but at this moment I doubt it.
From workng with young people for over 30 years I find removing narcissism a little troubling. From experience I can tell you that certain manifestations of personality disorders are difficult to identify early on but a strong trait of narcissism can show up and began to be identified prety early on. If this is broken up and accounted for in other clusters ok……maybe but if it is simply being ditched for unknown reasoning I do have concerns. One of the problems with clustering that I see is that if the parameters are not clearly defined and understood by the practioners then the out comes can be pretty hard to define.
Interesting that the first post on a thread about narcissism is by Bart.
Indeed, Bartles has an obsessive (((need))) to be 1st in every 538 thread! 😛 and yes his ad nauseam attempted Obama/Dem sarcasm is embarrassing at best.
Which begs the question ;^) should limbo, beck, billo, hannity, coulter, buchanan, and yes 538’s very own bart donate their brains to science ~ hopefully sooner rather than later …
;^) ;^) ;^)
drfunguy,
The problem you find, which I agree is quite real, isn’t in the analysis methodology, but a lack of understanding of the causality relationship between choices we make and the results we get from making them.
These are independent problems. If I want to pick a graduate school, I can use the clustering method to apply only to those top-ranked schools. Then, after making my decision, I might spend four years working toward a thesis under a particular professor, who then leaves for another school…
The problem is similar with diagnosis of mental illness. The problem with putting someone into the “narcissist” cluster is that there’s no treatment for that condition, so far as I know, other than, perhaps, to elect them to public office, as Monotreme suggests, and shipping them to Washington — which just passes the buck.
“dimension reduction” is a fancy way of saying, “well, I’ve got round holes, and triangular holes, and this here cylinder is going into that round hole.” It may produce useful insights by reducing a complex situation to one which we can handle — or it may throw out information that we shouldn’t have ignored.
For example, there are medications which can control paranoia. I’m speaking as someone who only has slight information on this, so if anyone can correct or elucidate this better, I would appreciate any corrections. It might make sense, therefore, to use a clustering approach in diagnosis which identifies whether a patient is paranoid — in spite of the fact that, as per the new guidelines, that will no longer be an official diagnosis.
I’ve seen a few examples of attempts at reducing the dimensions of a problem where the result was useless. (Management loves them their single-number-metrics, though.)
I don’t know if I’ll be able to express this well, so I ask your forbearance with my likely clumsiness in saying what I’m trying to say…
If I understand what the new DSM guidelines are suggesting, it’s not so much that they’ll be removing the label of “narcissism” but that they’ll be identifying the specific traits rather than lumping that set of traits into a specific condition. For conditions that are mostly behavioral, I can see how this would be useful.
Sometimes, I think labeling, especially of psychiatric conditions, can be counter-productive. Most of the labels for these conditions are really just grouping of similar symptoms into a classification for convenience in describing/treating/etc. It’s not like chicken pox, where if you have this virus or bacteria, then you get schizophrenia, but if you have that virus, then you get bi-polar disorder. It’s diagnosis by symptom rather than “disease” anyway. Therefore, by identifying and treating the classes of specific symptoms rather than labeling it as a broader disorder, that could be more helpful in guiding the treatments.
Furthermore, I see how sometimes once something is labeled as a particular disease, that there is stigma and surrender to the label rather than remaining pro-active and positive through treatment. It can be a relief to finally have someone say, “Johnny has ADD,” instead of always hearing what a bad kid he is, but then one must be careful not to make ADD the excuse for everything. It doesn’t really matter if you call it ADD, ADHD, hyperactivity disorder, or whatever. What matters is that there are coping mechanisms (and medications) that can help Johnny overcome his concentration and focus issues.
Similar with other hi-profile mental conditions such as Autism. Does it matter whether it’s “low spectrum Autism” or “Asperger’s Syndrome” or “Anti-social disorder” or whatever? Autism, like many mental disorders, falls on a continuum rather than having discrete delineations between each level of the disorder. Sometimes there is a lot of fighting over the label rather than just getting down to treating the kid’s withdrawal and non-communication.
I do note that sometimes certain labels are crucial for determining benefits (from medical insurance or government disability or whatnot). But perhaps that’s the problem with those labels. It shouldn’t matter whether Johnny falls just this side or that side of the fuzzy line between this disorder or that one. What matters is the treatment of the conditions and symptoms so that he may eventually become a functioning, productive member of our society.
mclever,
I’m with you on this, and I think you said that very well. I don’t know that it’s that important, anymore, that a person becoming “productive”. I’d weaken that to “capable of participating in society”.
Thanks, shortchain.
“…capable of participating in society.” I like it. Much better than the way I said it.
🙂
Which begs the question ;^) should limbo, beck, billo, hannity, coulter, buchanan, and yes 538′s very own bart donate their brains to science ~ hopefully sooner rather than later …
Amen to that. The sooner, the better.
I haven’t seen the tool that provides interactive rankings of the math schools, but to me that seems very appropriate, especially if the intention is to help students with the choice of which program to attend.
Yes, there’s a cluster of elite schools (10?), but the precise order within those elite schools may not be so clear, because different criteria will matter more to some people. Smaller class size? A particular specialty? Placement of graduates?
When I chose what music school to attend, I chose what was then listed as #4 or #5 rather than #1, because the brass program at the lower-ranked school was considered superior, while the composite rankings put more emphasis on voice and strings. The trumpet instructor at the lower-ranked school had taught most of the teachers at the other top schools, so why settle for second best at some ostensibly higher-ranked institution? Plus, the director of bands at that school was legendary! There were other tangible benefits, including the ability to also major in math at a top-tier university, but if I were choosing just on the music criteria alone, I would have made the same choice. Yet, if I’d been a vocalist instead of a brass player, then my choice might have been foolish!
From a clustering vs. discrete perspective, this makes sense. The list of top music (or math or econ or engineering) schools almost always includes the same 5 or 6 names in varying order, with some fuzziness in rounding out the top ten. I saw a list recently with Harvard as the top music school, and my immediate reaction was WTF?? until I realized they were looking at doctoral programs, and the Harvard music business degree scored very highly. But I’ll bet that any vocalist would prefer Julliard, and violinists flock to Eastman… Most undergrad musicians would go to Indiana or Michigan, long before even thinking about Harvard.
Barted ~ For those who have anger issues …
Recent bart ~ Asswipes all
Indeed!
I’m not narcissistic, I’m just more important than you.
Which begs the question ;^) should limbo, beck, billo, hannity, coulter, buchanan, and yes 538′s very own bart donate their brains to science ~ hopefully sooner rather than later …
Amen to that. The sooner, the better.
Yes, even small contributions add up over time. Those than cannot give a lot, should give what they can.
Those who think they know everything really, really annoy the hell out of those of us who know we know everything …
Thought I had made an error, but was mistaken.
I prefer: I’m not narcissistic, you’re just less important than me
Where’s the national championship thread ?!?
Rooting for the Ducks, but, but, but watched Auburn come back from (24) pts. down against Alabama in the Iron Bowl ~ you all remember Alabama ~ they were national champions last year …
In other happy news: Tom Delay sentenced to (3) years in prison! :::happy dance::: lol
Go Ducks! 🙂
There is no “i” in “team”, but there are five “i”s in “narcissistic personality disorder”.
Yeah, Go D U C K s
I think a friend of mine that is an Auburn alum may have started a pregame party a little early. Just had one of the most incoherent calls ever. That crowd sounds as though they may be lucky if they are able to see the whole first half. Too funny. I’m sure there are no Duck partys.
HaHaHaHaHaHaHaHHHHHAAAAA!!!!!!
Congrats to Auburn! ~ Oregon had the ball on the (2) yard line, 1st and goal and couldn’t score ~ ‘nuf said!
Great game! Congrats to champions, Auburn.