When I first became a TA, I noticed there was a grandmother problem. Despite claims that our society was aging, it quickly became clear to me that within a few years, our grandmothers would die out. During a period of one year, I recorded the number of students who had lost a grandparent (almost always the matriarch), and found, shockingly, that nearly 1/5 of the students had a grandmother die each year. Now, at first, this might seem unlikely. After all, wouldn’t that mean that after 5 years most grandmothers would be dead?
Not really, remember that this is a moving sample. Remember, also, that there is not a one-to-one correspondence between grandmothers and students. In some case, a single lost grandmother might impact dozens, possibly scores, of college students. Also remember that each student has two biological grandmothers.
You see how complex this becomes. Yet we can deduce that for the average student at least one grandmother will likely pass on during their years in the university. It is a sad, statistical fact.
Now, we all know that correlation does not mean causation, and that increased university attendance need not cause a septuagenarian epidemic. Yet, there is some evidence that grandmothers die at even higher rates just before exams. Now, any trained social scientist will see this as a measurement error. Of course, grandmothers die at all times during the semester, it is simply that an instructor is unlikely to be alerted to this fact. (This is the flaw in this otherwise exemplary study.)
The conclusion, however, is even more bleak. My informal measurement of 20% per year was based only on data collected immediately before and after exams. When we include other periods, the number is closer to 40% or 50%, meaning that all but the smallest number of students will lose both grandmothers during their college years. As a result, as college attendance increases, we are likely to see male life expectancy to far outstrip female life expectancy in the US. As these findings become general knowledge, more women will refuse to have children, or will insist that they have poor educations. The effects of this are clearly long-term, but the potential harm to society is extreme. This area demands further study.
This entry was posted in Uncategorized and tagged Teaching. Bookmark the
permalink. Trackbacks are closed, but you can .
Dead Grandmothers
When I first became a TA, I noticed there was a grandmother problem. Despite claims that our society was aging, it quickly became clear to me that within a few years, our grandmothers would die out. During a period of one year, I recorded the number of students who had lost a grandparent (almost always the matriarch), and found, shockingly, that nearly 1/5 of the students had a grandmother die each year. Now, at first, this might seem unlikely. After all, wouldn’t that mean that after 5 years most grandmothers would be dead?
Not really, remember that this is a moving sample. Remember, also, that there is not a one-to-one correspondence between grandmothers and students. In some case, a single lost grandmother might impact dozens, possibly scores, of college students. Also remember that each student has two biological grandmothers.
You see how complex this becomes. Yet we can deduce that for the average student at least one grandmother will likely pass on during their years in the university. It is a sad, statistical fact.
Now, we all know that correlation does not mean causation, and that increased university attendance need not cause a septuagenarian epidemic. Yet, there is some evidence that grandmothers die at even higher rates just before exams. Now, any trained social scientist will see this as a measurement error. Of course, grandmothers die at all times during the semester, it is simply that an instructor is unlikely to be alerted to this fact. (This is the flaw in this otherwise exemplary study.)
The conclusion, however, is even more bleak. My informal measurement of 20% per year was based only on data collected immediately before and after exams. When we include other periods, the number is closer to 40% or 50%, meaning that all but the smallest number of students will lose both grandmothers during their college years. As a result, as college attendance increases, we are likely to see male life expectancy to far outstrip female life expectancy in the US. As these findings become general knowledge, more women will refuse to have children, or will insist that they have poor educations. The effects of this are clearly long-term, but the potential harm to society is extreme. This area demands further study.
Share this: