top of page

Halloween special: Diversification bias

One famous Halloween study found that people tend to want to diversify, and as a result they end up choosing variety even when they don’t particularly like it.


The study was set up as follows: Two households collaborated to allow trick-or-treaters to choose a pair of candy bars from two large stacks of Milky Ways and Three Musketeers. However, the manner in which the choice was offered was different across two groups:

  • In the 1st group, kids made one choice at one home, and then another choice at the next home.

  • In the 2nd group, kids were asked to choose any two candy bars at the same time.

Interestingly, how candy was offered affected choice. The researchers found that when kids chose their two candy bars one at a time, they tended to choose the same candy bar — presumably the one they liked. But when they chose two at the same time, they diversified.

🦇 They named this diversification bias. 🦇


What do trick-or-treaters tell us about human psychology?


In non-Halloween contexts and with adults as the subjects, other researchers found similar effects in shopping behaviors. For example, one study found that when people plan their lunches for the upcoming week, they tend to buy a diverse assortment of foods. But when they bought lunch one day at a time, they often ended up eating the same turkey sandwich and strawberry yogurt every day.


Here are three significant implications:

🎃 In finance🎃

Investors tend to naively diversify when they plan their portfolios. Naive diversification happens when investors blindly spread their assets across all the options available to them, without regard to what those options are. For example, if you believe in Bitcoin and decide to invest in it, alongside a dozen other cryptocurrencies you have never heard of before, just for the sake of diversifying.

🎃 In recruitment 🎃

Making hiring decisions individually, as is usually the case, leads to a lot more homogeneity in the working force, compared to making selection decisions in sets. For example, if you hire eight people at the same time, it will likely be a much more diverse group than if you hire eight people in sequence. Zooming out of an individual decision turns recruitment decisions into a much more active, conscious process. This gives you the chance to be more deliberative about creating diversity within the workforce.

🎃 In general 🎃

You might have noticed how diversification bias in the first case (Halloween candy) is not useful, whereas in the second case (diverse nutrition) leads to a healthier diet. That's because biases are neither good or bad. They are features of how our brain works. Our brains evolved this way over tenths of centuries to help us interpret the world and survive. In today’s world, biases are sometimes useful, other times not. It’s up to us to decide how to decide, so as to capitalize on these brain features.

bottom of page