Death be not proud!

(With apologies to poet John Donne) 'For, those, whom thou think'st die at examination time, Die not’

By Gordon Henry, Assistant Professor of Psychology, IU Southeast

At the end of spring semester, one IU professor takes a satiric jab at an academic phenomenon that has afflicted educators for centuries: final examination familial mortality. His solution? More research!

The phenomenon of which I speak is seemingly unaffected by such variables as the department or school out of which I teach (primarily business/management or psychology) or the particular course currently ongoing (ranging from introductory psychology to graduate-level business seminars). Specifically, I can expect that, for every exam administered in my courses, a certain number of friends and relatives of my students will die.

The reported causes of deaths among this population cover the gamut of events that can end a human’s existence: cancers, tumors, heart attacks, ruptured aneurism, and most popularly, automobile and other types of accidents. The causes of death of family and acquaintances of my students generally mirror societal trends. Fortunately, it has been my experience that the homicide rate among those close to my students has been somewhat below the national average. Despite my inability to accurately guess the creative ways in which these people will expire, the number of deaths that will occur just prior to the administration of one of my exams is absolutely predictable.

I can reliably count on one death for every 25 students per exam. That is, 4 percent of my students will have a friend or relative expire just before any particular exam. As examples, Tables 1 and 2 contain the death-related data collected in two of my recent classes. If extrapolated to all institutions of higher learning in the United States, the total number of deaths that occur in close temporal proximity to course examinations is staggering.

According to the National Center for Education Statistics, there were 14, 966,091 students enrolled in post-secondary institutions of learning in the fall of 1998. About 4 percent (or 598,644) of these students could expect a loved one to die each time an examination is administered. An informal survey of my current colleagues led me to the conclusion that my average of six examinations per class during a semester is abnormally high, and that three exams is much more typical. So, approximately 1,795,932 (598,644 x 3) deaths of people known to students occur just prior to examination administrations in the United States alone!

It is entirely probable that some of the deceased (victims?) were known and/or loved by more than a single student. I am sure, however, that this estimate of the number of deaths is conservative given those inevitably occurring when other assignments such as research papers, case analyses, class presentations, etc., are due. I certainly do not wish to imply a functional relationship between events that I have shown to be only correlation-ally related. But the consistency of the relationship between test administrations and death rates, in my experience, certainly suggests an empirical question.

Might we, as academicians, be able to significantly impact death tolls by strategically manipulating the timing, difficulty, etc., of our examinations? Certainly, the potential universal impact of such a possibility deems the relationship worthy of further investigation.

My recommendation, then, is for academicians to collaboratively design and engage in lines of research that examine the exact nature of the relationship between exams and death rates. For example, all the business professors at a certain institution may agree to schedule their exams on the same days one semester.

The reported deaths of students’ friends and family members would then be tracked throughout the semester. If the number of deaths consistently fluctuates in accordance with the pre-arranged exam schedule, then some light would be shed on the possible functional nature of the relationship between those two variables.

The variety of designs that such lines of research could take is infinite. Unfortunately, a number of years spent engaged in this investigative endeavor must likely pass before any strong causal conclusions and potential societal payoff could be made.

Table 1, Course A

'Deaths' Occurring in Course A prior to Examinations

N=67	# Deaths	%Deaths
Exam 1	2	3.0
Exam 2	4	6.0
Exam 3	3	4.5
Final exam	0	0.0
(no makeups)

 

Table 2, Course B

'Deaths' Occurring in Course A prior to Examinations

N=11	# Deaths	%Deaths
Exam 1	0	0.0
Exam 2	1	9.1
Exam 3	0	0.0
Final exam	1	9.1
(no makeups)