4 Tips When Using Cognitive Biases

January 25, 2019      Kiki Koutmeridou, Chief Behavioral Scientist, DonorVoice

In the previous posts, we busted some myths and explained the risks of having a simplified view of biases. Taking both into account, I’ll close the week with 4 short recommendations.

1. Avoid “all or nothing” thinking. Don’t believe in the universal application of biases – don’t assume that an effect observed in one context will automatically apply in another. At the same time, don’t be quick to think a bias doesn’t exist – failure to replicate an effect doesn’t automatically invalidate it.

2. Replication is necessary. Findings from just one study, though interesting, aren’t conclusive. Avoid applying a brand-new insight before replicating it. There needs to be repeated testing in similar and different contexts before we can draw valid conclusions for a specific bias. If an observed effect is real, failed replications help provide clarity around its effectiveness and applications.

3. Be aware of the context. When designing your test or intervention, consider the context carefully as it might influence the outcome. Failure to replicate the effect could be down to that. Even if you think the circumstances are exactly the same, remember that no study can control for all confounding variables.

4. Hypothesize and test. When it comes to biases, the right approach is testing. Even when you’re dealing with biases believed to have a wide application, you still should come up with a hypothesis and test it. Skipping this step could lead to ineffective or, worse, damaging applications.

Kiki