Dashew on CA

[Jamie]

I work in Corp. Finance, but I have a liberal arts background and some in the dismal science.

I've found it very helpful. I can't tell you how many times I get served up a business case or a set of Financials where the technical experts have built a beautiful and intricate model based on some weak assumptions. Or where my accountants will come back to me with a set of numbers that tie out, but don't make sense.

They all sort of frown at me suspiciously, and go back and crunch some more and then come back to me and ask me how I knew.

Because I come from a non-traditional background, I'm more open to people's educational history; achievement is more important than the degree. If you are smart and motivated, the technical aspects of my field are easily taught.

The most important class in B-school was ironically the one people slept through - Managing People at Work, which was really about leadership and organizational design.

Political Science: It's a bit misnamed, but I wish people paid more attention to some of the basic principles - like that you need a strong middle class to have politics in the first place.

[Orestes Munn]

My final bitch, and then I'll go back to sorting out racing scores, is the plethora of pseudo-sciences. The list is endless and endlessly entertaining:

Political Science - Exactly how to does one set up a controlled scientific experiment in the political relm? The Germans tried it during the Nazi era, not so popular.
Sociology - 'nuf said
Psychology - more than 'nuf said, as we watch Freudian refuted over and over and over again.
BV

I don't really know what political science is, other than the art of opinion polling, but the social, cognitive, and clinical domains domains of what used to be called "psychology" have adopted all the rigor of the "natural" sciences and now use, perhaps excessively, tools such as functional MRI and, increasingly, genetic analysis. Social psychology has revolutionized economics by dumping the theory that individuals always optimize utility and actually observing the irrational and costly things which human brains force their owners to do—what an idea!

Incidentally, I think the most serious challenges facing humanity are, ultimately, problems of human behavior. We desperately need scientific advances in this area.

[LarryHoward]

Interesting drift.

Dealing with a challenge with my 2 younger kids right now. Both are taking "meaningful" courses. Daughter in Biology and son in Mech E. i've encouraged them to take electives in project management and in basic business finance as my experience is that new grads generally lack the ability to plan and execute even a minor project and generally lack the understanding of business finance to logically consider alternatives. Neither course is taught as a stand alone at a major university. You have to be a finance major for the business finance class or a management major for elementary project management.

At the University of Maryland, one course is offered in adult financial management and reviews budgets, investment fundamentals, basic estate fundamentals, etc. One section with limited seats offered in a school with 32 thousand students. The wait list is over 200. No consideration for adding additional offerings. Plenty of courses in the history of rock and roll, etc. Any wonder that I have to mentor young engineers on paying their bills and not living paycheck to paycheck.

[kimbottles]

Basic personal finance should be a required class in high school. I am amazed how many people have no clue about this rather important subject.

[B.J. Porter]

Political Science - Exactly how to does one set up a controlled scientific experiment in the political relm? The Germans tried it during the Nazi era, not so popular.
Sociology - 'nuf said
Psychology - more than 'nuf said, as we watch Freudian refuted over and over and over again.


BV

One of the more entertaining things I did in college was take two courses entitled "Animal Behavior" at the same time. One was offered by the Psych department, one by the Biology department.

Completely different, antithetical views on how to explain the same behavior. Though the Psych prof was much more dismissive of the Biologist's approach, the Biology people largely ignored what the Psych people had to say.

Psych focused on Skinner boxes, operant conditioning, hard wiring of behavior, etc. Bio class focused on population biology, genetics, selective and evolutionary pressures, etc.

It made test taking interesting though, when I knew I could induce paroxysms of cynicism and derision by answering a Psych question with a Bio answer...

[Orestes Munn]

Political Science - Exactly how to does one set up a controlled scientific experiment in the political relm? The Germans tried it during the Nazi era, not so popular.
Sociology - 'nuf said
Psychology - more than 'nuf said, as we watch Freudian refuted over and over and over again.


BV

One of the more entertaining things I did in college was take two courses entitled "Animal Behavior" at the same time. One was offered by the Psych department, one by the Biology department.

Completely different, antithetical views on how to explain the same behavior. Though the Psych prof was much more dismissive of the Biologist's approach, the Biology people largely ignored what the Psych people had to say.

Psych focused on Skinner boxes, operant conditioning, hard wiring of behavior, etc. Bio class focused on population biology, genetics, selective and evolutionary pressures, etc.

It made test taking interesting though, when I knew I could induce paroxysms of cynicism and derision by answering a Psych question with a Bio answer...

Much of that dichotomy has been resolved by the recent explosion in behavioral genetics/epigenetics.

[BeauV]

Kim - In addition to personal finance, folks really should be taught some simple basic stuff like fixing a leaking toilet, replacing failed electrical sockets, etc... it would save them a fortune. While I still have wonderfully fond memories of my poetry professor, and reading poetry remains one of my many addictions, my shop teachers taught me more useful stuff in high school than anyone else if it's measured by frequency of use.

OM - I agree that recent invasions of the soft sciences by what have traditionally be thought to be hard sciences (such as chemical or mechanical explanations for things that folks thought were psychological problems with the spirit or soul) has improved things - in my view. Doesn't that make my point seem more reasonable. I am genuinely thrilled that fields from Psycology to Economics are actually trying hard to put structure and rigor into their fields. I remain unable to believe these are fields one can call science until a researcher can actually run repeatable controlled experiments to sort out some pretty subtle difference between correlation and causality. This lack of ability to control experiments for quite reasonable and proper reasons is a big problem with medical research in general. How long to does one give a placebo pill to a clinical trials patient when it has been reasonably well proven that the placebo effect isn't working but the drug is? I do like to see trials halted because the results a simply too positive to morally continue to deny the drug or procedure to all patients, but it does disrupt the science a bit.

I actually think I'm a bit extreme on this particular issue because my world view doesn't include believe in the existence of things for which there is no physical evidence. Thus, because some folks feel there is a soul, that is not evidence of its existence. My personal view is that a mind is a wonderful manifestation of an astoundingly complex biological device. This, obviously, eliminates the existence of a soul as anything other than as a side effect of this complex device. However, the device can get screwed up in various ways (most of them chemical in nature) and not function in a what is socially defined as the correct way. Finding a way to fix this is a good thing, as a malfunctioning mind can lead to incarceration or suicide. A good sailing friend just lost his 48 year old daughter to suicide which he attributes to the effects of a serious bipolar disorder. The drugs helped tremendously, but could never cure her. On the other hand one of my daughter's first serious boyfriends had sever bipolar problems and drugs seem to have completely cured him. No amount of couch therapy ever had any effect whatsoever on either of these people.

BV

[Orestes Munn]

Clinical trials are an extreme case and probably the most difficult and complex experiments in all of science to do right and ethically and much of what is published in that area will necessarily be untrue. Don't confuse them, however, with well designed experiments looking at which brain areas are active during social or other behaviors or behavioral experiments which allow us to distinguish between theories, i.e., mathematical models, of how humans behave. These paradigms have their weaknesses too, but reproducibility, controls, and rigorous statistical methods are threshold requirements as much for behavioral research as for any other mainstream scientific field, (with the variations in quality seen universally in experimental science). Some examples of current research which you feel fails these criteria might help.

[kdh]

I like that you took Latin, Eric. I've never regretted it. Thanks, Miss Vancavage.

[kdh]

Beau, about psychology I used to think the way you do until I read, "Thinking, Fast and Slow," by Daniel Kahneman. His life's work, basically. Hugely insightful.

[kdh]

Beau. about Cambridge and Doctors of Philosophy, keep in mind that thinkers were Renaissance men not long ago. Philosophers pontificated about everything, and most of it was bullshit. The scientific method is a pretty recent phenomenon.

[Olaf Hart]

Beau, about psychology I used to think the way you do until I read, "Thinking, Fast and Slow," by Daniel Kahneman. His life's work, basically. Hugely insightful.

My wife is reading that book on her Kindle at the moment.
We were discussing topics from it over breakfast this morning!

[kimbottles]

My SWMBO really enjoyed that book, I found it interesting, might be accurate too......

[BeauV]

Keith - OK, I'll read "Thinking, Fast and Slow" I'm always up for a good read.

OM - You've made good points about higher levels of discipline moving into fields that had been a bit squishy; which is a very good thing. I did read a lot of stuff from a Dr. Toga at UCLA a few years back that was quite interesting as he observed brain activity when people were doing certain things (like reading or singing). But, when I asked him if that part of the brain had "activity" roughly equal when the person wasn't doing those things, it got pretty squishy.

Here's another example from my own experience with my previous doctor. I had a very high PSA test result during a physical. My previous doctor immediately pushed me to visit a specialist who examined my prostate, ugh! He didn't find anything with a manual exam. They did the blood test again, it was still quite high. I went to another specialist who couldn't find anything with a "normal" exam and did a scan. Still nothing, but the PSA reading kept going up. Finally, a bike riding friend, who is now my doctor, asked: "Did you tell them you do about 10-20 miles a day on a bike?" Well, no, I hadn't mentioned that and no one had asked. My new doctors said something along the lines of: "Have your blood tested after you've been off your bike for a week and see what the PSA number is." I did that, after a two week break from the bike for vacation, and the PSA number was not only within normal readings it was actually quite low. Lesson learned: diagnostic tests can have multiple causes for a given reading. I do realize that this is entirely anecdotal, and that it is not an example of poor logic or experimental structure. But, sadly, it is also typical of many of the interactions I've had with the medical community. Correlation and causality are frequently confused, along with a number of other errors of logic, which can result in a great deal of frustration and in my case a significant amount of additional unnecessary testing and worry.

The PSA example is a pretty simple one, now image the difficulty in dealing with variable that appear to correlate with various psychological problems, where things seem less well understood and defined, where the patient is attempting to report on how they feel using a language with all its subtleties and complexities. I really don't mean to be disparaging well meaning folks, and every one of the folks in medicine I've ever dealt with are well meaning. It's just that this is an extremely difficult problem and at times folks tend to express a level of certainty about the "cause" of a problem that isn't really justified by the evidence.

[Jamie]

People tend to beat up on the dismal science, mostly because at the the heart of it, it is trying to model the behavior of people which is hard to try boil down to a bunch of formulas, even at the aggregate level. We're a long way from Asimov's Psychohistory.

In my past life I worked in the Medical Device and Pharma world, so maybe I'm not so impressed with clinical trials. Now a days, the cost of developing a new compound ensures that Marketing and Finance get involved pretty early in the process. Marketing decides what markets to pursue, scientists decide what pathway to follow and Finance models each one out over the product lifetime. One big cost is that a lot of countries reimbursement system require not only that you prove your compound or device is better than a placebo, but better than the existing standard of care. This means more studies. Not all countries have standardized data requirements - this is largely a protectionist or cost saving mechanism - this means more studies too. Product safety - trying to tease out all of the safety issues is another big cost driver. But, none of the analytical tools used is much more complicated than what is taught in high school. The rest, is all soft, flabby judgment.

We made a number of anti-psychotic drugs - some were very long acting - like 3 months - 4 injections a year. If you can catch it early enough, and do not allow the patient to relapse, they can live an almost normal life.

[Orestes Munn]

Keith - OK, I'll read "Thinking, Fast and Slow" I'm always up for a good read.

OM - You've made good points about higher levels of discipline moving into fields that had been a bit squishy; which is a very good thing. I did read a lot of stuff from a Dr. Toga at UCLA a few years back that was quite interesting as he observed brain activity when people were doing certain things (like reading or singing). But, when I asked him if that part of the brain had "activity" roughly equal when the person wasn't doing those things, it got pretty squishy.

Here's another example from my own experience with my previous doctor. I had a very high PSA test result during a physical. My previous doctor immediately pushed me to visit a specialist who examined my prostate, ugh! He didn't find anything with a manual exam. They did the blood test again, it was still quite high. I went to another specialist who couldn't find anything with a "normal" exam and did a scan. Still nothing, but the PSA reading kept going up. Finally, a bike riding friend, who is now my doctor, asked: "Did you tell them you do about 10-20 miles a day on a bike?" Well, no, I hadn't mentioned that and no one had asked. My new doctors said something along the lines of: "Have your blood tested after you've been off your bike for a week and see what the PSA number is." I did that, after a two week break from the bike for vacation, and the PSA number was not only within normal readings it was actually quite low. Lesson learned: diagnostic tests can have multiple causes for a given reading. I do realize that this is entirely anecdotal, and that it is not an example of poor logic or experimental structure. But, sadly, it is also typical of many of the interactions I've had with the medical community. Correlation and causality are frequently confused, along with a number of other errors of logic, which can result in a great deal of frustration and in my case a significant amount of additional unnecessary testing and worry.

The PSA example is a pretty simple one, now image the difficulty in dealing with variable that appear to correlate with various psychological problems, where things seem less well understood and defined, where the patient is attempting to report on how they feel using a language with all its subtleties and complexities. I really don't mean to be disparaging well meaning folks, and every one of the folks in medicine I've ever dealt with are well meaning. It's just that this is an extremely difficult problem and at times folks tend to express a level of certainty about the "cause" of a problem that isn't really justified by the evidence.

Toga is a smart guy and a leader in functional neuroimaging. I have met him and know his boss, Mazziotta, a little. I think their greatest contribution has been to try to assemble big libraries of brain mapping data across institutions and modalities. This could have big payoffs.

Clinical medicine is a mess and full of the kind of cognitive errors that Kahneman made famous (and got a Nobel for). Medical schools do not not teach critical thinking and it is all too easy to get very bad care from smart, well trained, and well-meaning people, with the entire, double-edged, clinical armamentarium at their disposal. As my grandfather, a cynical old German doc, used to say, "There is no condition so benign or insignificant that, in the hands of a competent physician, cannot prove fatal". I have seen this borne out more than once.

[kimbottles]

Keith - OK, I'll read "Thinking, Fast and Slow" I'm always up for a good read.

OM - You've made good points about higher levels of discipline moving into fields that had been a bit squishy; which is a very good thing. I did read a lot of stuff from a Dr. Toga at UCLA a few years back that was quite interesting as he observed brain activity when people were doing certain things (like reading or singing). But, when I asked him if that part of the brain had "activity" roughly equal when the person wasn't doing those things, it got pretty squishy.

Here's another example from my own experience with my previous doctor. I had a very high PSA test result during a physical. My previous doctor immediately pushed me to visit a specialist who examined my prostate, ugh! He didn't find anything with a manual exam. They did the blood test again, it was still quite high. I went to another specialist who couldn't find anything with a "normal" exam and did a scan. Still nothing, but the PSA reading kept going up. Finally, a bike riding friend, who is now my doctor, asked: "Did you tell them you do about 10-20 miles a day on a bike?" Well, no, I hadn't mentioned that and no one had asked. My new doctors said something along the lines of: "Have your blood tested after you've been off your bike for a week and see what the PSA number is." I did that, after a two week break from the bike for vacation, and the PSA number was not only within normal readings it was actually quite low. Lesson learned: diagnostic tests can have multiple causes for a given reading. I do realize that this is entirely anecdotal, and that it is not an example of poor logic or experimental structure. But, sadly, it is also typical of many of the interactions I've had with the medical community. Correlation and causality are frequently confused, along with a number of other errors of logic, which can result in a great deal of frustration and in my case a significant amount of additional unnecessary testing and worry.

The PSA example is a pretty simple one, now image the difficulty in dealing with variable that appear to correlate with various psychological problems, where things seem less well understood and defined, where the patient is attempting to report on how they feel using a language with all its subtleties and complexities. I really don't mean to be disparaging well meaning folks, and every one of the folks in medicine I've ever dealt with are well meaning. It's just that this is an extremely difficult problem and at times folks tend to express a level of certainty about the "cause" of a problem that isn't really justified by the evidence.

That's why my doctor made me get a split seat...........now normal PSA all the time.

[bob perry]

I'm going back to Jack London. He uses small words.

[Jamie]

Lesson learned: diagnostic tests can have multiple causes for a given reading.

And that's before you actually get to dealing with the equipment and lab technicians actually running the testing. All sorts of things can happen with the calibration of the equipment, interpretation of the results and peculiarities of the reagents used to test. False results happen a lot more than you might think.

Most labs are seen as cost centers and run by metrics like test per hour and turn around time. If I get a results that doesn't make sense, I get re-tested.

[BeauV]

"There is no condition so benign or insignificant that, in the hands of a competent physician, cannot prove fatal"

I ADORE THAT!!

[Olaf Hart]

I have been involved in two public health clusters, one with SIDS and the other Legionnaires Disease.

We managed to stop both before we had enough deaths or cases to publish meaningful correlations, but we still had 20 cases of SIDS and over 100 of Legionnaires Disease.

We certainly knew what the problems were, because we were able to stop the outbreaks, but if we had let them go on longer we could have published!

We were able to pass our findings on to other researchers, so eventually the correlations were proven. But traditional clinical research models were useless.

In my field the most useful data comes from longitudinal studies, or meta analysis of multiple studies. An old Professor of mine used to say if it takes a thousand people to prove something, it is probably worth knowing. If it takes ten thousand it is definitely not worth knowing.

And I know that I and many of my colleagues give zero credibility to any research funded by drug companies.

[cap10ed]

"As my grandfather, a cynical old German doc, used to say, "There is no condition so benign or insignificant that, in the hands of a competent physician, cannot prove fatal". I have seen this borne out more than once."
I like your grand dad. That truism has the ability to transcend into other professions.

[LarryHoward]

"As my grandfather, a cynical old German doc, used to say, "There is no condition so benign or insignificant that, in the hands of a competent physician, cannot prove fatal". I have seen this borne out more than once."
I like your grand dad. That truism has the ability to transcend into other professions.

Very much so. I can relate that to young design engineers who can make a simple bracket so complicated, it can't be manufactured without a 5 axis mill and a 90% scrap rate. Of course, the customer just wanted to mount a new display. I keep being reminded of Colin Chapman who extolled designers to "Simplify and add lightness" and Kelly Johnson's Skunk Works efforts to keep things simple.