Last week, I asked about differences in perception between a 39-yard, 40-yard and 41-yard field goal. The primary goal was exploring how people perceive various distances and differences between those distances. In other words, we might have some rough understanding of how hard a “40ish-yard field goal” is and thus lump in 39 yard and 41 yard field goals into one big perception of “40ish-yard field goals” as opposed to truly recognizing differences in difficulty between 39, 40, and 41 yards. Furthermore, it showed the psychology behind round numbers (such as 40) and their potential to change behavior. This week’s survey explored a similar consumer psychology idea but in a different way:
Imagine you overhear two people talking about a high school basketball player.
Person A: “How is his shooting?”
Person B: “Good.”
What do you think the player’s shooting percentage (shots made/shots attempted) is?
The shooting was either described as “good” or “bad” which was similar to the survey from a few weeks ago. There was one key difference though: the slider (where you provided your answer) was set up to either show two decimals places (ex. 39.47) or round numbers (ex. 40). In other words, half the people could only answer with round numbers (39, 40, etc.) and the other half were shown numbers down to two decimal places (39.47, 39.75, 40.23, etc.). As an example:
Who cares (other than maybe the parents of the kid or the actual kid) whether a high school basketball player shoots 20.18% versus 20%? Thus, asking for that amount of precision is a bit odd and might influence responses.
Much of this is inspired by writings from Paul Grice, a famous linguist and philosopher (a Berkeley professor too – Go Bears!), who is known for his views on conversation norms. Essentially, these are general principles that seem to hold true in many conversations. One of them is the category of quantity which makes two arguments:
1. Make your contribution as informative as is required (for the current purposes of the exchange).
2. Do not make your contribution more informative than is required.
As mentioned, why would someone ask you to guess a high school basketball players field goal percentage down to two decimal places!? Why say 20.18% when you can say 20%? Nobody cares about the .18 difference. Thus, it (arguably) violates Grice’s norms and has the potential to influence behavior.
Although the point of my #1QFriday series is to discuss ideas and loosely test them through a survey, it would still be great to get a few more responses. So, if you haven’t already signed up (or know someone who might be interested) you can sign up for the email list here and I’ll send it out the survey link on Friday. That being said…
What do you notice?
There appears to be subtle differences based on the answer choice options (round numbers vs. two decimal places). For example, the people who saw the shooting referred to as “bad” estimated 30% when they could only respond with whole numbers but 26% when they were able to be even more precise (ex. 25.96% vs. 26%).
In addition, when looking at response times, the people who were forced to answer with round numbers took about 29 seconds while the people who were forced to answer with decimals took about 44 seconds. Keep in mind, the actual answer slider looked exactly the same (as shown above) in either condition. One argument is that in the precise condition (two decimal points) people *thought* about it longer because the answer required a certain level of precision.
What might play a role?
Gricean norms. As discussed, being more precise than the situation calls for (ex. shooting percentage down to two decimal places) potentially violates norms of conversation. Thus, it might have influenced responses and led to subtle differences (like you see in the bad shooting scenario) between people who answered with round numbers and those who did not.
Processing style and time. As the difference in response time indicates, people in the precise condition took much longer to respond. One argument is they are processing the information differently due to the answer format being in two decimals places. We don’t necessarily have a solid understand of shooting percentage down to two decimals places, or think of shooting percentage with that level of precision, and thus we might think about the question longer or differently.
So what is the takeaway?
Differences in precision matter. This study explored that idea in a slightly different way but you can also think about it in different ways such as a project timeline, as Zhang and Schwarz illustrated in their study. If I say it will take 364 days versus 1 year….in which case do I sound like I know what I’m talking about? “Your project will be done in 364 days” versus “Your project will be done in 1 year.” Mathematically and numerically they are equivalent but explaining the project in days is much more precise. As such, I’m probably trusting the guy/gal who says 364 days because they seem more confident in their estimate.
If it were me, I would look at every situation in my business where I present numbers. How precise are you? Should you be more precise? Less precise? It could be projects (365 days vs. 1 year), pricing ($3.57 vs. $4), discounts (17% off versus 20%), delivery times (1 week vs. 7 days), etc. It makes a difference. Any other examples/scenariors you can think of where you might be making tradeoffs between being specific and being general?
#1QFriday is a blog series that includes my random thoughts and musings. I do not claim to be the first person to think of these things or deny that others have done research on them. I am simply discussing interesting marketing, psychology and linguistic topics that come to mind. Furthermore, I am avoiding statistical analysis as I want the emphasis to be on the ideas. I recently read (Amazon affiliate link) Paul Grice: Philosopher and Linguist and was inspired by a less statistical and more philosophical approach. Thus, the results here could happen by chance. I want to focus on thinking about, discussing and debating the ideas at a broad level rather than whether or not the differences are statistically meaningful. I typically send out a quick survey on Friday and post a follow up analysis/discussion on Monday.