I got a server error when I tried to leave a comment on Tyler Cowen’s blog. Here is the comment:
Suppose that the credibility of a scientist is a function of two variables: (1) the evidence he offers; and (2) the strength of his beliefs. I am willing to allow the partial derivative of (2) to be positive and to stipulate Tyler’s argument that faking data is a signal of strength of beliefs. But the partial derivative of (1) is much stronger, and fake data enters with a negative sign.
READER COMMENTS
pj
Dec 2 2009 at 1:52pm
More to the point, the old lawyer’s adage: When both the facts and the law are against you, pound the table. A SHOW of strong belief — pounding the table — is often a sign that the evidence is against the believer. Well-supported beliefs can be advanced quietly. So Tyler probably gets the sign of (2) wrong also.
joecanuck
Dec 2 2009 at 2:02pm
who’s Tyrone?
Radford Neal
Dec 2 2009 at 2:58pm
But a scientist who strongly believes a hypothesis will probably also think that data supporting it is likely to be found soon. So there’s no reason to fake data – and every reason not to, given the risks of being caught (both personal, and for the public credibility of the hypothesis). So the partial derivative with respect to (2) should also be negative.
SydB
Dec 2 2009 at 3:30pm
“A SHOW of strong belief — pounding the table — is often a sign that the evidence is against the believer. Well-supported beliefs can be advanced quietly.”
This is just so wrong. A quite or loud discussion is a sign of nothing. Kling–and Mankiw as another who tries to sell this bill of goods–argue that there is some sort of polite world of tea-sipping intellectuals who are agreeably debating topics in a comfortable smoking room off the side of Plato’s celestial realm of pure ideas.
Nope. Doesn’t work this way.
Arnold Kling
Dec 2 2009 at 3:36pm
Joecanuck, Tyler Cowen occasionally writes posts as his “evil twin, Tyrone”
TimS
Dec 2 2009 at 4:25pm
I too had trouble posting on MR, so here’s my two cents on this topic:
P(S) = my prior probability for the problem being serious
P(D) = my probability for the scientists behaving dishonorably
P(D|S) = my probability for the scientists behaving dishonorably GIVEN that the problem is serious
P(S|D) = my probability for the problem being serious GIVEN that the scientists behaved dishonorably
Bayes rule tells me that
P(S|D) = P(D|S)P(S)/P(D)
so that after learning that the scientists behaved dishonorably, my probability for the problem being serious is scaled by the ratio P(D|S)/P(D). So, if P(D|S) is less than P(D), then I believe it is less likely that the problem is serious (Tyler’s response #1). If, on the other hand, P(D|S) is greater than P(D), then I will believe it is more likely that the problem is serious (Tyler’s response #2). Isn’t this all that Tyler said?
It turns out that if I believe the scientists are more likely to behave dishonorably if the problem is not serious than they are if it is serious, then I get the condition for response #1. If I believe the scientists are more likely to behave dishonorably if the problem is serious than they are if it is not serious, then I get the condition for response #2. Again, isn’t this all that Tyler said?
Ryan Vann
Dec 3 2009 at 12:36pm
There is entirely too much emphasis on the scientists and their motives and not enough consideration of the underlying data going on.
I don’t see how fudging numbers could do anything other than decrease one’s confidence in how compelling the data is, which means AGW confidence (or at least confidence in the magnitude) should come down, and uncertainty should go up.
CJ Smith
Dec 3 2009 at 1:40pm
@SydB:
Different people have different styles of addressing issues, Syd. Not everyone pounds the table, and not everyone discusses things calmly.
I can tell you from the perspective of an attorney (I don’t put Esq. after my name in this blog, even though I am a licensed and practicing attorney), PJ’s comment has a lot of merit, The adage as I have heard it is, “When you have the facts, pound the facts; when you have the law, pound the law; when you have neither, pound the table.”. That’s why attorneys with the facts and the law like non-jury trials and hearings, while attorneys with only sympathetic clients like jury trials. Most judges focus on the facts and the law, not the pounding on the table, while juries can be swayed to a bad decision by a fervent emotional appeal or a sympathetic client. That’s why attorneys trot out the family to swear the criminal defendant was physically and mentally abused for all his life so shouldn’t go to jail for abducting and raping the victim; or plaintiffs’ attorneys show the immediately post-op photos of surgeries that are no longer visible. Many multimillion dollar jury verdicts are reversed or reduced on appeal for exactly that reason – the facts and the law don’t support the verdict or the damages award.
Myself, I prefer to discuss things calmly – if you want to rant and rave, you’re demonstrating that your belief is so strong that nothing I can say will change your mind or your position, regardless of my factual or logical support, so there’s no point in arguing with you, other than to exercise your lungs. Conversely, if you support a position in a thoughtful and well supported manner, I’ll at least consider it, and may change my opinion; but trying to shove it down my throat means I’ll just ignore it (“You’re an idiot, but I want you to think like I do.”). For an example of a calm exchange between two people with strongly divergent views, see my discussion with Tom Dougherty in “Bernanke Mea Culpa Redux.” He made thoughful theoretical and practical arguments supported with data, and did so in a polite and professional manner. He did change my thinking and position on the matter, although I initially had strong beliefs to the contrary. I think most of the people on this blog are the same way – we have strong positions and disagreements, but generally don’t resort to ad hominem attacks or flame wars, and can be swayed by facts, not how many times a person with a different opinion claims we are ignorant, deluded or can’t spell.
Comments are closed.