Question: Birthdays and Death Days-Is There a Connection? Is the timing of death random or does it depend on significant events in one's life? That's the question University of California at San Diego sociologist David Phillips and his colleagues attempted to answer. Previous research had shown a possible connection between the timing of death and holidays and other special occasions. This study focused on the connection between birthday and death day. The researchers studied death certificates of all Californians who had died between 1969 and 1990. Because of incomplete information before 1978, we report only on the part of their study that included the years 1979 to 1990. They limited their study to adults (over 18) who had died of natural causes.
They eliminated anyone for whom surgery had been a contributing factor to death because there is some choice as to when to schedule surgery. They also omitted those born on February 29 because there was no way to know on which date these people celebrated their birthday in non-leap years. Because there is a seasonal component to birthdays and death days, the researchers adjusted the numbers to account for those as well. They determined the number of deaths that would be expected on each day of the year if date of birth and date of death were independent of each other. Each death was then classified as to how many weeks after the birthday it occurred. For example, someone who died from 0 to 6 days after his or her birthday was classified as dying in "Week 0," whereas someone who died from 7 to 13 days after the birthday was classified in "Week 1," and so on.
Thus, people who died in Week 51 died within a few days before their birthdays. Finally, the researchers compared the actual numbers of deaths during each week with what would be expected based on the seasonally adjusted data. Here is what they found. For women, the biggest peak was in Week 0. For men, the biggest peak was in Week 51. In other words, the week during which the highest number of women died was the week after their birthdays. The week during which the highest number of men died was the week before their birthdays. Perhaps this observation is due only to chance. Each of the 52 weeks is equally likely to show the biggest peak. What is the probability that the biggest peak for the women would be Week 0 and the biggest peak for the men would be Week 51? Using Rule 3, the probability of both events occurring is (1/52) = (1/52) = 1/2704 = .0004. As we will learn in Chapter 18, unusual events often do happen just by chance. Many facts given in the original report, however, add credence to the idea that this is not a chance result. For example, the peak for women in Week 0 remained even when the deaths were separated by age group, by race, and by cause of death. It was also present in the sample of deaths from 1969 to 1977. Further, earlier studies from various cultures have shown that people tend to die just after holidays important to that culture.