The Morning Eclipse

Sunday, February 01, 2009

We do business with Confirmation Bias

In an experiment by P. C. Wason, the results show us that it is inherent in people to continuously discover data that will confirm one's decision. On the flip side, it also shows that once we've made a decision, we may unconsciously turn a blind eye to data that will prove wrong our decision.

Here's a very simple example.

In the book of Paul Caroll, he tells of a simple game. The objective of the game is to guess the rule that is behind a numerical sequence. In other words, one must discover what points will determine a 3 number sequence.

Example: 2, 4 and 6. This is a sequence of three numbers that can be produced from a rule that states: "Three digits of even numbers".

In the beginning of the game, the participants are encouraged to guess the hidden rule. The participants of the game can ask the moderator as many times as they want if whether a particular sample sequence fits the hidden rule.

Example 1: Participant A can ask the Moderator if the sequence " 12, 14, 16" fits the to-be-guessed rule.

At this point, the moderator can only confirm or deny the correctness of the sequence.

Example 2: Participant B asks whether "32, 34, 36" fits the to-be-guessed rule. Moderator answers "Yes".

Example 3: Participant B asks whether "208, 210, 212" fits the to-be-guessed rule. Moderator answers "Yes".

Example 4: Participant B asks whether "64, 62, 60" fits the to-be-guessed rule. Moderator answers "No".

Example 5: Participant B then guesses that the rule is "Ascending even numbers" but Moderator answers "No".

On the other hand

Example 6: Participant C asks whether "59, 60, 61" fits the to-be-guessed rule and Moderator answers "Yes".

Example 7: Participant C asks whether "71, 75, 452" fits the to-be-guessed rule and Moderator answers "Yes".

When everyone's done guessing, the Moderator gives the answer.

One may wonder why Participant B is wrong and Example 6 and 7 are a yes. Well, this is because the to-be-guessed rule is "Any three ascending numbers". Examples 1 2 and 3 fit the rule but Example 4 contradicts it by counting backwards. In fact, the sequence neither has to be even numbers nor increment by two.

In Wason's experiment, it turns out that only a minority or in some cases none get the answer. This simply shows that as one keeps getting confirming answers, a person gets the idea that he is correct and keeps on finding scenarios/ information that will further strengthen his belief. It is seldom that this person will think of the opposite and dare ask a paradigm shattering question like Example 6 which throws the rule of "Ascending even numbers" out of the window.

In business, this is common and plentiful. We must be careful to avoid Confirmation Bias. We must be always ready to take time and try to look from a different perspective.

Success is 1% inspiration and 99% perspiration. Never tire.



Post a Comment

Subscribe to Post Comments [Atom]

Links to this post:

Create a Link

<< Home