Two regional surveys disagreed this month. How do you avoid trading the one that agrees with your view?

This week New York's factory index jumped 18 points to 25.5 while Philadelphia's fell from 37.8 to 20.9 (Empire report, Philadelphia report). Both surveys are real and both are regional, so each can be read as supporting a different view of the US economy.

Confirmation bias is the habit of seeing the data that agrees with what you already think. It is strongest when the news is mixed, because either side can be cited, and it is most costly when the position is large.

We would like to hear how you guard against it:

• When data disagree, do you write down both readings before deciding which to act on?
• Do you have a rule for the minimum evidence you need before changing your view?
• Have you traded the survey that confirmed your bias and lost? What changed afterwards?
• Do you deliberately look for the case against your trade before you enter?

How to build a trading plan covers writing these rules down.

The best defence is to write both readings down before you look at prices, and to note which one you were hoping for. If the two readings had been reversed this month, would you have traded differently? Did that habit change any of your trades in the last month?

Please keep replies to your own experience. Posts that promise a direction or sell signals will be removed.