MaxedOutMama
Friday, May 13, 2005
The Dunning Effect
I once read a psychological study that concluded that those who were terrible at a task were so bad that they didn't realize they were incompetent. See this brief description of the study, and here's a more detailed explanation. Dr. Dunning of Cornell has been studying this human characteristic for years:
This study came to mind after I read Jim Lampley's explanation on Huff 'n Puff regarding how he can be sure that Bush stole the 2004 election. To him it is obvious that the segment of the exit polls that showed Bush losing must be right. He ignores an initial slice that favored Bush and the final adjusted result that favored Bush. He's just absolutely confident that the ones in the middle favoring Kerry were correct. It's a statistical certainty, not that he needs to rebut those who claim otherwise with facts, because, as he tells us:
Jim is logically challenged, let's put it that way. So logically challenged that he thinks he's cleverer than the rest of us. Only a Dr. Dunning can really understand good ol' Jim. From the Mitofsky evaluation report:
Consider with me Lampley's statement "I could spend some time dissecting what I believe is an obvious whitewash, a delicate sidestep away from the potential public relations disaster of being tied forever to the most notorious election theft in history." How does it favor Mitofsky's image to produce a report showing that some of the errors were due to factors such as recruiting and training poll interviewers? If they wanted to support vote-fixing, why wouldn't the firm have mis-reported the data in the first place? Jim doesn't seem to understand that checks of precincts that produce manual ballots that can be counted don't support his thesis of vote fixing. Maybe he considers those comparisons to be irrelevant, because he knows there were "various other machinations aimed at suppressing, destroying or losing Kerry votes."
There are some people who are immune to reason and objective proof. The rest of us just have to learn to live with the Lampleys of this world. Dr. Dunning claims that remedial logic training did show some indications of lowering individual scores on the Dunning Scale. But as any therapist can tell you, first you've got to want to change. Jim doesn't. I leave you with this slightly hopeful observation:
Researchers gave participants tests in logic, English grammar and humour, then asked them how well they thought they had done. Those who scored the lowest rated their abilities very highly; those who scored at the high end believed they weren't very skilled. When the high scorers were given the low scorers' tests to grade, they revised their own abilities in their own favour, but the same didn't happen for low scorers. Some actually thought they were even more skilled after having seen the tests of the high scorers. The researchers have concluded that the skills required for competence are the same skills needed to recognize competence. In other words, the incompetent have no idea how incompetent they really are.This explains a lot about life - I will call this aspect of human nature the "Dunning Effect". Let me point out that the study discussed above used a species of exit polling. Both the competent and the incompetent misrated their scores, but the distinctive feature of the incompetents is that they were unable to correct their own self-assessment after they were confronted with the actual data. If you are really high on the Dunning Scale (10 = incompetent certainty, 1 = competent uncertainty), you lack the ability to assess objective data at all. Thus nothing tends to disturb your complacent certainty of your own competence.
This study came to mind after I read Jim Lampley's explanation on Huff 'n Puff regarding how he can be sure that Bush stole the 2004 election. To him it is obvious that the segment of the exit polls that showed Bush losing must be right. He ignores an initial slice that favored Bush and the final adjusted result that favored Bush. He's just absolutely confident that the ones in the middle favoring Kerry were correct. It's a statistical certainty, not that he needs to rebut those who claim otherwise with facts, because, as he tells us:
But none of that is necessary, because the entire Edison/Mitofsky report is irrelevant to the argument, given that it is based on the assumption the final official vote tally is accurate. Make no mistake: my argument is that the final official vote tally is anything but accurate, that it is the product of massive vote fraud carried out through the programing of Diebold voting machines and various other machinations aimed at suppressing, destroying or losing Kerry votes.The funniest part about poor Jim's certainty is that all exit polls are always adjusted by precinct returns. Jim blithely ignores Mitofsky's own report ("I could spend some time dissecting what I believe is an obvious whitewash, a delicate sidestep away from the potential public relations disaster of being tied forever to the most notorious election theft in history."), ignores other polling leading up to the election that was pretty accurate, ignores Pew's work before and after the election (confirming the election results), ignores other very recent exit polling problems, ignores the futures markets, ignores the Weekly Reader poll ( the most historically accurate of all), ignores the fact that manual counts of precincts with who weren't electronically voting seems to debunk the vast Diebold conspiracy theory, ignores the obvious overcount of women in the bad polling, and, finally and triumphantly, ignores the most basic rule of statistics - statistical analyses based on flawed data produce flawed results. Repeat it after me, Jim, "Good math + bad data = bad results."
Jim is logically challenged, let's put it that way. So logically challenged that he thinks he's cleverer than the rest of us. Only a Dr. Dunning can really understand good ol' Jim. From the Mitofsky evaluation report:
On November 2, 2004, the Election System created by Edison Media Research and Mitofsky International for the National Election Pool (NEP) produced election estimates and exit poll data for analysis in 120 races in all 50 states and the District of Columbia. In addition, between January and March 2004, Edison and Mitofsky conducted exit polls for 23 Democratic Primaries and Caucuses. For every election, the system delivered on its main goals: there were no incorrect NEP winner projections, and the exit poll data produced on election day were used on-air and in print by the six members of the NEP (AP, ABC, CBS, CNN, FOX and NBC) as well as several dozen media organizations who subscribed to that data.Over and over again the report tells us that only the final numbers count, and that the final numbers did have Bush winning. But Jim hasn't read this report. He doesn't need to - the fallacy is obvious! And even if he did, he would be able to ignore this statement:
Exit polls do not support the allegations of fraud due to rigging of voting equipment. Our analysis of the difference between the vote count and the exit poll at each polling location in our sample has found no systematic differences for precincts using touch screen and optical scan voting equipment. We say this because these differences are similar to the differences for punch card voting equipment, and less than the difference for mechanical voting equipmentFor those lower on the Dunning Scale, Jim's dogged certainty is stunningly funny. If Jim's belief is true, the organization that produced the data that proves Bush lost is also the organization that is engaged in a massive cover-up of that fact. That's quite a reach. Jim doesn't seem to understand that checks of precincts that produce manual ballots that can be counted don't support his thesis of vote fixing.
Consider with me Lampley's statement "I could spend some time dissecting what I believe is an obvious whitewash, a delicate sidestep away from the potential public relations disaster of being tied forever to the most notorious election theft in history." How does it favor Mitofsky's image to produce a report showing that some of the errors were due to factors such as recruiting and training poll interviewers? If they wanted to support vote-fixing, why wouldn't the firm have mis-reported the data in the first place? Jim doesn't seem to understand that checks of precincts that produce manual ballots that can be counted don't support his thesis of vote fixing. Maybe he considers those comparisons to be irrelevant, because he knows there were "various other machinations aimed at suppressing, destroying or losing Kerry votes."
There are some people who are immune to reason and objective proof. The rest of us just have to learn to live with the Lampleys of this world. Dr. Dunning claims that remedial logic training did show some indications of lowering individual scores on the Dunning Scale. But as any therapist can tell you, first you've got to want to change. Jim doesn't. I leave you with this slightly hopeful observation:
When high-scoring subjects were asked to ``grade'' the grammar tests of their peers, however, they quickly revised their evaluations of their own performance. In contrast, the self-assessments of those who scored badly themselves were unaffected by the experience of grading others; some subjects even further inflated their estimates of their own abilities.Maybe we could reduce the incidence of moonbattery (a willful assault upon reasoning) by mandating courses in logic in elementary schools. Maybe.
``Incompetent individuals were less able to recognize competence in others,'' the researchers concluded.
In a final experiment, Dunning and Kruger set out to discover if training would help modify the exaggerated self-perceptions of incapable subjects. In fact, a short training session in logical reasoning did improve the ability of low-scoring subjects to assess their performance realistically, they found.
