Friday, December 28, 2007

How do We Extract Signal from Noise?

. Friday, December 28, 2007

As we head toward the New Year, I offer some fundamental skepticism. In a very cheery examination of the current financial situation, Ambrose Evans-Pritchard seems almost to relish the collapse of the international financial system. He seems uncannily able to find bankers willing to make extreme statements, such as: "Things are very unstable and can move incredibly fast. I don't think the central banks are going to make a major policy error, but if they do, this could make 1929 look like a walk in the park." How's that for a little New Year cheer?

I feature Evans-Pritchard's recent work because it nicely highlights one of two questions I have been pondering for the last two weeks (I will post on the second question tomorrow). How do we humans extract meaningful signal from all the noise? We seem strongly attracted to worse-case scenarios; we compare the current situation to 1929 rather than to other financial episodes that did not produce a global cataclysm (1987, for example, or the bursting of the dot com bubble). The media feeds us quotes with no evidence about where the people quoted fall in the distribution of opinion and analysis. Are the quoted bankers representative of the median view, or is the writer drawing from the gloom-and-doom end of the spectrum? Do the people quoted have incentive (monetary or other) to portray events in a particular manner rather than present "objective analysis?" In short, how does one figure out the "truth" in a world in which the critical question is "what is going to happen?", when the information we have at hand is of uncertain quality and quite possibly biased, but the degree of bias is unknown (and unknowable)?

This seems to be a very hard problem, and yet one we face on a regular basis. For those not interested in the current financial crisis, then ask the same question of global warming. If you are unwilling to fall down the global warming warren hole (and having disappeared into that hole for a full week, I can hardly blame you if you are not), then ask the question of genetically modified organisms. If GMOs do not interest you, then what about economic development? All of these issues pose the same dilemma: we are asked to make decisions today to achieve some future consequence relying on knowledge first produced and then reported by groups of people who may or may not be motivated by the objective quest for truth and that therefore may or may not be biased in ways we cannot accurately measure.

In short, why do we believe what we believe, and should we? How do we extract the meaningful signal from all of the noise, and how do we know that we have extracted the meaningful signal? I don't have the answer, but we seem to believe that the political process is good at selecting the "correct signal." Why do we believe this? Should we?


Anonymous said...

Deep thoughts, Dr. Oatley.

Anonymous said...

I believe Mill's marketplace of ideas theory is an attempt to addresses this question, sort of, maybe.

How do We Extract Signal from Noise?




Add to Technorati Favorites