How do we estimate frequency in the world? A quick strategy, known as the availability heuristic, is to think about how easy it is to generate examples of the event in question. For instance, you can quickly evaluate that it is more common for restaurants to offer coffee than Brussels sprouts by recalling that you nearly always see coffee on the menu, but can only think of a handful of menus that include Brussels sprouts. The ease of generating coffee examples compared with Brussels sprouts examples helps you conclude that coffee is a more frequent menu item than Brussels sprouts.
Sometimes, however, commonality may be less clear. For instance, how popular is commuting by bicycle in your city? (Really – make a guess!). In order to estimate the popularity of biking, you might run through a list of scenarios and think about whether biking examples come to mind. Do you or your friends bike as a main form of transportation? How often do you see bikers in the street? Can you think of bike racks in your city, and if so, are they used?
Our ease in generating examples to estimate frequency can vary greatly according to our perspective and experiences. For example, if you yourself are a biker, you may find it very easy to think of friends who bike, other bikers along your commute, or bicycle enthusiasts in your neighborhood. Your heightened awareness to other bikers on the road may lead you to overestimate the overall number of bike commuters in your city. In contrast, if you yourself do not bike, do not know people who bike, and rarely notice anyone using the bike lane, this difficulty generating examples of bikers will lead you to underestimate the number of bike commuters in your city.
The availability heuristic can be a useful indicator of frequency in many situations (e.g. you can probably judge how popular a song is by how easily you can think of it played), but it can lead us to overestimate things that are easy to bring to mind (e.g. shark attack fatalities, which get a lot of press but are actually rare) and underestimate events in which it’s harder for us to generate examples (e.g. jellyfish sting fatalities, which are much more common yet rarely make the news!).