Science never ends with an anecdote; otherwise, it would be sociology. But sometimes science begins with an anecdote.
I first became interested in circadian rhythm – why some people are night owls while others are morning larks – when I lived in Christchurch, New Zealand, for one year. I lived on Riccarton Road, one of the main drags in Christchurch. Down Riccarton Road from where I lived, there was a supermarket called Countdown that was open 24 hours a day, 7 days a week.
I would often go to Countdown to shop at three o’clock in the morning. Every time I did, I noticed that the place was crawling with Asian customers. This was before the recent explosion of Asian immigration to New Zealand, so Asians were still a small minority in Christchurch back then. I would not see many Asians anywhere in Christchurch most of the time, except at three o’clock in the morning in a 24-hour supermarket.
I began wondering then if this was because Asians were more nocturnal than other races. I had not thought about possible race differences in circadian rhythms until I lived in New Zealand and repeatedly encountered a large number of Asians at three o’clock in the morning. I had always shopped at three o’clock in the morning in 24-hour supermarkets all my adult life, but I did not remember seeing a disproportionate number of Asians before.
It was many years before I finally solved the mystery of late-night Asian shoppers, by eventually formulating the Savanna-IQ Interaction Hypothesis and collaborating with Kaja Perina on a paper on intelligence and circadian rhythm that I summarize in my previous post.
So are Asians more nocturnal than others? I don’t have data on Kiwis, but data on a large representative sample of young Americans do suggest that they are. Compared to all others, Asians go to bed significantly later on weeknights (00:43 vs. 00:08) and on weekends (01:27 vs. 01:12). However, Asians do not wake up significantly later than others either on weekdays (07:38 vs. 07:32) or on weekends (10:20 vs. 10:18). In general, races differ significantly from each other in what time they go to bed, both on weeknights and on weekends, but they do not differ significantly from each other in what time they wake up, either on weekdays or on weekends.
So it appears that my casual observations at Countdown in Christchurch so many years ago may have some empirical basis. Compared to other races, Asians do appear to stay up (and, presumably, shop groceries, among other things) later into the night. The bivariate association between being Asian and nocturnality (what time they go to bed at night), represented in the graphs above, remains significant even when childhood IQ is controlled for. Even net of childhood intelligence, Asians go to bed significantly later than others every night. So it is not because they are on average more intelligent than others that Asians are more nocturnal.
However, the association disappears once we control for all the other social and demographic variables included in the original analyses reported in the previous post (age, sex, current marital status, parental status, education, income, religion, whether currently in school, and the number of hours worked). In fact, the only significant effect of race in the multiple regression analyses is being black. Compared to whites (the reference category), blacks go to bed later on weekends (but not on weeknights) and wake up later on weekends (but not on weekdays).