THE BLOG
06/06/2013 11:09 am ET Updated Aug 06, 2013

Education: What Numbers Count

The misuse of numbers in policy discourse is nowhere more pronounced than on the subject of education. One glaring example is provided by the anxious discussion about college graduation rates, tuition and debt. Almost everyone seems to be irked by some aspect of the present state of affairs. Tuition has been rising to levels that have made the costs of higher education unmanageable for many families. Many students accumulate considerable loan debt in order to finance their education. Many work full or part-time. Quite a few both work and have debt. Graduation rates within a 4 or 6 year period understandably decline. That logic can be ameliorated only marginally through the provision of scholarships since the reduced financial support for public higher education crimps scholarship funds along with everything else.

The first numerical error is to segregate one set of numbers from other sets representing closely related aspects of the problem. We have the odd phenomenon of many university administrators designing plans to raise graduation rates through the setting of targets without paying attention to the evident fact that working students simply cannot take a full course load. Instead, they concentrate on marginal factors such as students staying enrolled to take a few extra credits or spending too much time partying. This is an example of studied inattention to numbers that are inconvenient and ignoring the simple arithmetic that reveals their practical significance. The second error involves searching for answers along unorthodox avenues that promise magical solutions. On-line learning is now being aggressively promoted as the magic cure. It is the darling of cost conscious administrators, business interests moving quickly to carve out a lucrative profit center, and the irrepressible technological optimists who are ever ready to seize on the latest innovation to cure anything and everything that ails us - death aside, for the time-being. The most obvious logical flaw in these calculations is that the product is not being held constant. To study in isolation using depersonalized means of communication is simply not the same as "going to college." This is very much like the notion that today's airline ticket is the same as the one bought in 1973. A few basic technical courses probably can be taught with reasonably satisfactory results to highly motivated, mature students - but that is about it. To base forecasts of educational outcomes and employment prospects on an unproven, likely fallacious assumption of equivalency is irresponsible and dishonest.

Where do numbers figure in the distance learning panacea cum fad? One, as noted, is that a course is a course, that credit hours are credit hours. A second is the ignoring of the inconvenient truth that to achieve even the lowest level of satisfactory results depends on the heavy use of low paid teaching assistants and adjuncts to monitor, to test and to grade. That merely extends and accentuates a trend toward reliance on an unstable, exploited work force that already has degraded higher education in the United States. A fair accounting of investments in human resources would shift the financial cost-benefit calculations of distance learning markedly. A third numerical error is the over reliance on numbers themselves. Specifically, to take into consideration only those elements of reality that can be counted and to leave aside all the important intangible elements is to deform reality and distort the truth. Numbers are bought into play either so as to certify what is not statistically certifiable or to exclude it. There is one thing that we can be certain of: on-line learning will be declared a success by all the self-interested parties. That is preordained. The right people can always come up with the right numbers to confirm the rightness of their bias. The precedent already has been set by charter schools. Abundant evidence tells us that they are no panacea; indeed, they divert scarce resources from public systems while producing results inferior to the conventional schools. Their students fall short of public students by 3% - despite the systematic exclusion of "problem' students and other distortions. These numbers, though, change nothing in cultivated perceptions nor do they counter the strong currents running in the charter school direction.

University life is rife with examples of misused numbers. Take evaluations - which all of us do since the current ethos has everybody evaluating everyone else, on paper anyway. If we can't hold presidents accountable for acting like scoundrels, or financial barons from swindling us, at least we can take rigorous measures to make sure that we're getting our money's worth from teachers and holding students' noses to the grindstone. Think of Grade Point Averages (GPAs). A 3.5 is not a 3.5 is not a 3.5. There is a world of difference between a transcript studded with grades ranging from D to A compared to one where all grades cluster around 'B'. Elementary. But a fundamental difference that is ignored more than it is recognized. The same with course evaluations of professors. An instructor whose individual evaluations range from '1' to '5' is obviously a very different sort of professor than one who receives all "3s." But try telling that to a department chair or to a Dean. Any faculty member can recall meetings where some diligent colleague, scrutinizing evaluation data for a promotion candidate on his laptop, pronounces with gravity: "last Spring the course surveys averaged 3.85, there was some improvement in the Fall to 4.01 and then serious slippage to 3.72" - all in the confident assumption that something of supreme importance has been revealed. Publication performance also is susceptible to numerical illusion. Many departments place greatest weight on the number of publications - especially articles. They are counted; the more the better. Yes, there often is a weighting of some kind that differentiates among journals as to their (presumed) quality, i.e. toughness of criteria for acceptance. What almost never gets considered is the variation of supply-demand ratios from one field to another. That is to say, it is easier to publish in an elite journal in some fields of specialization than in others. Then, what about co-authored articles? Never mind - what is wanted is a few simple numbers that sustain the institutional bias.

The examples that we have cited point up a further oddity about the place of numbers in public discourse. Although exalted as the definitive statement of reality, they are blithely ignored when inconvenient. Hence, the burying of charter schools' failures, as noted above. So, too, the disregard for the enormously important fact that nation-wide funding for public elementary education has dropped by 4% since 2007 in real terms (U.S. Census Bureau). This does not include the draconian cutbacks and closings by Rahm Emanuel in Chicago. So, too, the even more drastic drop in funding for public higher education over a far longer period of time - 8%. (State Higher Education Executive Officers association). That is a quarter-century low, while enrollments continued their climb to record highs - and, the percentage of educational revenue supported by tuition has climbed steadily from 23.2 percent in 1986 to 43.3 percent in 2011. Why pay attention to these statistics, though, when we already have discovered the wonder drugs of charter schools and distance college learning. Doubt it? We'll soon have the numbers - any numbers - to 'prove' it. We'll also have graduates with a depleted ability to make sense of them.

When faced with a problem, a logical mind proceeds through three stages. First, define precisely what are the issues you are trying to interpret and analyze. Then assemble as much useful data as possible. This is the point at which numbers should enter the picture. Finally, assess the relevance of the numbers as applied to the pre-defined problem. Were only it so!