Tuesday, March 29, 2005

I put my hand up to "confirmation bias"...


I admit it. I do it...


I came across the term in, of all places, an advertisement for financial services. But, on reading it, the immediate wider application was also apparent.


Definition

In statistical inference, confirmation bias is a type of cognitive bias towards confirmation of the hypothesis under study. See also bias (statistics). To compensate for this observed human tendency, the scientific method is constructed so that we must try to disprove our hypotheses. See falsifiability.


In psychology, confirmation bias is a phenomenon whereby, in a variety of settings, decision makers have been shown to notice more, assign more weight to, and actively seek out evidence that confirms their claims, and tend to ignore and not seek out evidence which might discount their claims. As such, it can be thought of as a form of selection bias in collecting evidence



And again...
Behavioral Finance: A Confirmation Bias in Action

Iain Fraser commenting on Royal Bank:


On Feb. 18, 2005 we wrote: "Royal Bank ($66.25) is at a new 12 month high. Going through $64.50 was an upside breakout and the technical expectation is for a move to the $70-$72 range. Also on our analysis it still maintains its" No.1 rating."

One week later the stock closed at $73.10! Whoopee!


Much of the time when a stock goes to a new high it keeps on going.


Yes, much of the time when a stock goes to a new high it keeps on going. Except for those times when it doesn`t. Then it falls and we lose money. If investing were that easy, we could all buy stocks only when they make new 52-week highs and retire rich. Unfortunately it`s not that easy.


This is what behavioural finance calls a confirmation bias:


In psychology, confirmation bias is a phenomenon whereby, in a variety of settings, decision makers have been shown to notice more, assign more weight to, and actively seek out evidence that confirms their claims, and tend to ignore and not seek out evidence which might discount their claims. As such, it can be thought of as a form of selection bias in collecting evidence.


When investors notice a particular event a few times, such as a stock ripping through its 52-week high, they tend to make notice of other stocks that obey this rule and tend to ignore those stocks which do not follow the trend. This results in clouded judgment which can impair the investment process. As investors, it is important to remain as objective as possible and to guard against cognitive biases which give us a false perception of reality, such as the notion that "much of the time when a stock goes to a new high it keeps on going."



I will not use the Schiavo case to illustrate the point. Anyone who wishes to google "schiavo" through the blogworld can sit down and enumerate the number of times that those supporting the husband have quoted news article and source after source giving that side of the story. No, don't look here; as I said I am as guilty as any...


Similarly, if you look to sites that support the Schindlers you will find again "loving parents", Biblical quotation on quotation, news article and source after source.


One could use the start of the Iraq War, the Iraq elections, the UN, any number of the "hot topics" that vitalise the people (including me) who comment upon the state of their world at anything above the "heart on the sleeve" entirely personal level.


Back in June last year, I wrote on the different emphases on the same story between The Guardian - a fairly stolid right of centre "traditional" British paper, and Fox News. I came to the following conclusions;


Well, there it is…somewhere in there is the truth.


The differences between Guardian and Fox are perhaps slight, but very subtle. In both cases there are key elements of the story attributed to anonymous but key figures. In both cases there are direct quotes from named sources.


Perhaps the biggest difference though is the amount of space that Fox gives to Chalabi himself, and the Iranian government. I can hear the right arguing the truth comes from those involved. I can hear that and my head says “vested interest” for both parties. Chalabi, because of what he has already lost, how much more he can lose, and what he thinks he can gain. The Iran government because they have been caught with honey on their paw.


...

Most significant difference in my view?


The Guardian starts by stating that the Iranian government passed false intelligence through Chalabi and INC to Washington…


In the Fox article, the Iranian government denies that Chalabi passed intelligence to them.


Nice, subtle, and revealing…


I followed that with a number of ideas on propaganda, but in terms of the opening gambit on this post the comparison between Guardian and Fox is closest to the point. For once (I recognise) I have managed to use a specific example to present a fundamental idea. What a pity that I had not heard of "confirmation bias" at that time, but perhaps just as well. I can carry too few ideas at once as it is...


I have battled at times with the Artist over his insistence that I present "evidence" supporting opinions I express. Oft times this of itself has raised the ire level somewhat for no other reason than (as he has pointed out quite rightly) my opinion will always be better than his; in exactly the same way that he believes his opinion will always be better than mine (for example). So, the effect of this is to immediately send me into a frenzy of googling trying (and I meekly admit sometimes without success) to find authoritative "sources" to support my contention. I have pointed out the futility of this in the past. No one "learns" anything; and far more importantly no one examines the sources being used critically to judge their validity to the point being made. So, in place of informed debate we end up with what amounts to a mud pie fight; with the mud pies being "suitable and appropriate quotations of important sources" and whether taken in or out of context is really not material.


The difficulty with suppressing the expression of opinion - not supported by "evidence" - is that it neatly prevents such analysis unless of course you can find some other "more authoritative commentator" who has picked exactly the same matter using the same logic that one is wishing to present. Then, can it be considered an "original" thought. Obviously not... But I must move on.


What has become apparent to me is this - obvious as it might sound;


Any site, any blog, any commentator, that consistently presents only one side of an argument; that consistently presents the same sources as "evidence"; that persists in discounting any sources or evidence to the contrary without examination; obviously suffers from a very bad case of confirmation bias.


I have much to work on...

Postscript -
An after-thought after reading the financial example I gave at the start of this post...

To what extent was the dot-com sharemarket boom driven by confirmation bias?
Anyone prepared to bare their conscience?

1 comment:

Dave Justus said...

It is always hard to spot biases that match are preconcieved beliefs. Probably that is one reason that Conservatives scream that the media is Liberal and Liberals scream that it is conservative.

I try, although I don't always succeed, to seek out the strongest challenges to my beliefs rather than the weakest ones.