You decide to take a long trade on the euro and then hear news which could impact it negatively. What are you likely to do in that situation?

A recent study shows people are twice as likely to seek information that confirms their beliefs than they are to consider evidence that contradicts them. Psychologists call this mental gremlin the “confirmation bias” and when applied to trading, it’s something that can cause us to take unnecessary losses.

In short, our minds tend to act as compulsive yes-men who echo whatever we want to believe. What’s worse, a recent analysis of psychological studies with nearly 8,000 participants concluded that people are twice as likely to seek information that confirms what they already believe as they are to consider evidence that would challenge them.

Now, while it’s all well and good to recognize this phenomenon, what’s important to know is why it’s so hard to change a mind that’s already made-up.

First, people tend to be mentally lazy because it’s easier to focus our attention on data that supports our thinking rather than to seek out evidence that might disprove it. Being proven wrong is painful and people naturally avoid it.

Second, it’s easier for people to rationalize than to be rational, which means having to take a cold, hard look at things in an unbiased way. We can always find after-the-fact explanations of why our predictions didn’t work, and we reinterpret our failures as near-misses. You’ll know you’re doing this if you find yourself saying things like “the euro would have gone up if only X had happened,” or “99 times out of 100 I would have been right if not for this freak event.”

The old saying is that a little bit of knowledge can be a dangerous thing. It’s true because the more you learn the more certain you become of your decisions. Gathering more data is a good thing but it won’t make your predictions more accurate if each new fact only resembles the original one because the information diversity is reduced, thereby decreasing its value.

So how can you fight confirmation bias?

Psychologists recommend doing an exercise; imagine that you have looked into a crystal ball and have seen that your trade has gone against you. Next, come up with the most compelling explanations you can find for the failure. Doing this can help you realize that your beliefs might not be as solid as you thought.

Next, try estimating the odds that your technical or fundamental analysis is wrong. Say there’s a 20% chance your brilliant idea is incorrect; that is like saying you will be proven wrong one in every five times. This way, if your trade does go bad, you will be less likely to dig in your analytical heels and desperately try to prove that you are still right. Psychologists suggest that doing this provides the mental “cover” for admitting that you’re wrong.

Before you trade in the first place, write down a statement of what would cause you to change your view of the investment. If any of those events come to pass, the written record will make it harder for you to pretend nothing has changed or that you don’t have to do anything in response.

In my trade room this week, I went long the A$ soon after the market opened on Sunday. My reasons were because I believed that the up-trend would remain intact and that traders might “price in” a better-than-expected number for U.S. retail sales (which would be supportive of stocks and therefore negative for the dollar).

But I also evaluated what would get me out of this trade before I even got in it. A bad read on retail sales was one, and Fed Chairman Bernanke saying anything that might be interpreted as market-negative was another.

As it turned out, Bernanke said in his prepared statement that the Fed is “attentive” to changes in the dollar’s value and “will help ensure that the dollar is strong” and that was all I needed to hear to get out of my short dollar trade.

Now, as it turned out, after a brief decline the A$ started rising again, and I got stopped out of my trade at break even. That’s OK though, because what’s important is that I followed my plan and 99 times out of 100 that approach will turn out to be the correct one.