9

I have been spending some time sitting in a shopping center, and I have noticed a weird pattern. There is always one cash register open. When one person arrives at the cash register, many others arrive at the same time, creating a cluttered line. A few days ago, I interviewed the workers about this, and they told me that this is, in fact, a common phenomenon occurring in grocery stores. I wonder what the biological or human behavior patterns are that lead people to decide to go shopping/at the cash register in such wave-like patterns. This is not an illusion. Please, if someone can tell me, is there a name for this phenomenon? I didn't know where I could ask this, but I thought here would be okay, if it is not okay, then just delete my question.

Julius Hamilton
  • 1,559
  • 4
  • 29
Oneprime
  • 101
  • 4
  • 1
    Sounds like a case of "Might is right". There was a study with ants that had a choice of regions to live, which were indistinguishable from each other, and the bulk of ants would group together as a matter of popularity. – DanielFBest Mar 29 '24 at 16:52
  • 4
    Poisson distribution? – Scott Rowe Mar 29 '24 at 17:30
  • 5
    Vote for closure: better suited for https://psychology.stackexchange.com/ . Consumer purchasing behaviors of any form are probably well researched among psychologists and sociologists and social psychologists. :D There is a strong element of conformity by people in social situations. See Solomon Asch. – J D Mar 29 '24 at 18:00
  • 4
    How can such a question--maybe interesting but absolutely not related to philosophy--have more than 7 upvotes? – Johan Mar 30 '24 at 15:56
  • 1
    @johan HTQ effect. This should be moved, either to psychology, or to cross validated. – James K Mar 30 '24 at 16:13
  • 1
    @JD Why not migrate to https://economics.stackexchange.com/questions/tagged/behavioral-economics ? – user1147844 Mar 30 '24 at 21:20

2 Answers2

9

If we make the slightly controversial (read: probably not entirely true) assumption that people arrive at cash registers independently of each other, with no statistical correlation or synchronization (in the short term), then we should already expect them to appear unevenly. This is known as Poisson clumping, and it has nothing to do with human behavior, it is a purely statistical phenomenon, arising out of the fact that random numbers tend to exhibit more streaks and less even spacing than most people expect. To avoid clumping, shoppers would need to consciously avoid going to checkout during periods of high activity, which they (evidently) don't do (enough).

Of course, it is plausible that this assumption is false and that there is also a behavioral component to this phenomenon. To determine whether that is the case, you would need to take a number of measurements of when people arrive at cash registers, and use a statistical significance test to compare those arrival times to a Poisson point process.

Kevin
  • 2,028
  • 10
  • 18
  • @ScottRowe: That is why I said "(in the short term)" - you and other shoppers may have a preferred time of day for going shopping, but it is much less likely that you all choose precisely the same minute to appear at checkout as a result of some systemic bias. – Kevin Mar 29 '24 at 17:32
  • "I was agreeing with you." – Scott Rowe Mar 29 '24 at 17:33
  • @TKoL "As I was going up the stair, I met a man who wasn't there. He wasn't there again today, I wish, I wish he'd stay away" – Scott Rowe Mar 30 '24 at 02:05
  • +1, but people going shopping together, checking out around the same time, and paying separately is often the result of a common systematic bias known as friendship. I'd upgrade from probably to definitely not entirely true. – g s Mar 30 '24 at 04:33
  • Thank you for the answers this is very interesting. I will be spending more time there and calculating the amount people who arrive to make more out of this. – Oneprime Mar 30 '24 at 09:23
  • Interesting claims. But they needn't be mutually exclusive. If events are a function of behavior, and events manifest Poisson clumping, then that just means that human behavior, in this context, is described by Poisson clumping. I think it would be difficult to defend the thesis that human beings choosing lines in a grocery store doesn't have a behavioral component. Great answer! +1 – J D Mar 30 '24 at 22:05
3

I'll take a shot at this.

A, Multiples of 2: 0, 2, 4, 6, 8, 10, 12, 14, 16, 18, ...
B, Multiples of 3: 0,.,3,...,6.....9......12,...,15,...,18 ...
C, Multiples of 4: 0,....,4,.....8,........12,......,16, ...
D, Multiples of 5: 0,.....5,.........10,........, 15,.........

If shopping times are 2, 3, 4, and 5 minutes on avearge, there will be waves every 4 minutes, 6 minutes, 8 minutes, 10 minutes, 15 minutes, 16 minutes, 18 minutes, and the largest wave will be every 12 minutes (3 groups of shoppers converge on the cash register) for every batch of customers arriving at the same time.

It's got to do with composite and prime numbers. Please google magicicadas (17 year lifecycle).

Agent Smith
  • 3,642
  • 9
  • 30
  • But why would customers arrive in batches? – TKoL Mar 29 '24 at 18:44
  • 1
    @ TKoL It's turtles all the way down ... cogito. Good question! Maybe we should use the Poisson distribution or something. – Agent Smith Mar 29 '24 at 18:45
  • @TKoL If shoppers want to pick up supplies on their lunch hour, then that would be a large batch distributed over the lunch hour. Same logic for people shopping on their way home from work. If one Googles a grocery store, the distribution of customers over the hours of business is shown as a bar graph. – Idiosyncratic Soul Mar 29 '24 at 19:15
  • @IdiosyncraticSoul that's not the kind of reasoning being used in this answer. He's talking about people arriving at the same time, not distributed across a lunch hour – TKoL Mar 29 '24 at 19:19
  • @TKoL, apologies, humblest, if the answer didn't satisfy you. It isn't necessary for the customers to arrive at the same time. Suppose 0 here represents 12 noon. Person A can arrive at 12:10 and person B at 12:09 and person C can arrive at 12:08. They'll all be at the cash register at 12:12. Please refer to the answer (edited to add A, B, C, D). – Agent Smith Mar 29 '24 at 19:28
  • @AgentSmith that has the same problem as before, both rely on the customers arriving in some non random pattern. Why start with that assumption? In one example, they arrive simultaneously. Why? In your comment here, they're arriving not simultaneously, but miraculously with the right timing to get to the register at the same time - why? – TKoL Mar 29 '24 at 19:32
  • 1
    @TKoL, this would be a simplified model of the event described by the OP. I would've loved to refine it to a point where it convinces you, but I lack the tools. Sorry. Secondly, imagine a shop that's open for only 1 minute, that's 60 seconds. If customers arrive at a rate of 1/s and 90 customers are "interested", some of them will arrive simultaneously. It isn't miraculous. – Agent Smith Mar 29 '24 at 19:44
  • 2
    There is a similar patterning with ocean waves on a beach: after 9 waves there are two big ones. Sit and watch. On a rough day it would be more chaotic and unpredictable, but still, waves pile on each other with some regularity. Something to do with a Poisson distribution... – Scott Rowe Mar 30 '24 at 02:03
  • 1
    Thank you for the answers, this is very interesting. I will be spending more time there and calculating the amount people who arrive to make more out of this. I will maybe return with the result if your answer correctly predicts the clumps of people. – Oneprime Mar 30 '24 at 09:25