Via a mailing list, probably via this Tyler Cowen post, an awful statistic about adjunct faculty:

35 years ago there were 44% more tenured faculty than adjuncts. Today there are 76% more adjuncts than tenured faculty, via @chronicle

— Ángel Cabrera (@CabreraAngel) April 25, 2014

This is awful in two ways. First, it's obviously a sad comment on the state of the college teaching profession. More importantly, though, it's a classic abuse of statistics, using a really confusing presentation of the numbers to exaggerate an effect that doesn't really need it.

Here, let's try this as a poll, since the way this is stated basically amounts to an eighth-grade algebra word problem. Answer before scrolling down to read the explanation:

(A 10-percentage-point increase here means going from, say, 12% to 22% of the total. Which is not the right answer, by the way.)

SPOILER SPACE

SPOILER SPACE

SPOILER SPACE

SPOILER SPACE

SPOILER SPACE

SPOILER SPACE

SPOILER SPACE

SPOILER SPACE

SPOILER SPACE

If you want numbers to make a sensible comparison, you have to solve some equations-- if we say that the number of adjunct faculty 35 years ago is x, then the number of tenure-track faculty is 1.44x, and the fraction of adjunct faculty is:

$latex P(adj) = \frac{x}{x+1.44x} =0.41 $

The modern calculation has the number of tenure-track faculty as y and the number of adjuncts as 1.76y, so:

$latex P(adj) = \frac{1.76 y}{y+1.767} =0.64 $

In other words, according to these poorly sourced statistics (I didn't turn this up on the Chronicle of Higher Ed site or Twitter feed), the adjunct fraction increased from 41% to 64%. It ought to be enough to note that adjuncts have gone from a sizable minority of faculty to a substantial majority, but the "X% more than" formulation is just confusing, in order to sound more scary. Nobody reports numbers this way unless they're engaged in "How to Lie With Statistics" hackery.

No matter how good the cause, responsible academics should be better than this.

- Log in to post comments

Phew...

As a math adjunct, I'm glad I got the answer right. However after having just given a near 8 hour calculus marathon, my brain nearly farted on me and said: "wait, that's a near quadruple increase (a bit less than half to near double), why is 120% the highest answer?!?"

There's another slightly dodgy aspect of this, namely that it's "tenured" not "tenure-track." That might be a slight inaccuracy to keep the original stat tweetable, or it might significantly increase the non-adjunct fraction.

Confusing presentation of the numbers, indeed. Unfortunately, the phrasing in the tweet above (and poll) doesn't lead to the conclusion below. Tenured faculty + adjunct faculty does not equal total faculty.

(My comment snuck in while Will's was in moderation...)

If you add tenure-track but not-(yet)-tenured faculty to the equation, that adds an extra term to the denominator of each of the fractions, which will lower the adjunct percentage in both cases, and tend to bring them closer together. So it is a slight inaccuracy in the conclusion, but only a quantitative issue-- the stat used is still confusing in a way that seems to exaggerate the problem.

(Which, I will repeat, is a very real problem, and something that needs to be addressed. But terrible and confusing stats don't really help with that.)

It need not be merely quantitative! A sudden shift in college hiring patterns could have created an assistant professor pool that dwarfs both stated populations. From this data alone, no clear sense if we have 5 % adjunct or 60 % adjunct. Of course, applying logic to the the problem, we can probably assume that those numbers have remained a relatively stable proportion of the total.

35 years ago is after the great golden age of faculty expansion and into "In a few years, a wave of retirements will open a lot of tenure-track jobs" so I don't think there would've been a vast number of assistant professors then. Today, as we know, there aren't a huge number of assistant profs, either. So I think it's likely to be a fairly small effect on both ends.

Having now seen the original figure from which this statistic seems to have been derived-- in this tweet, I think "tenured" is short for "tenured and tenure-track." So the point is moot.

One flaw in those kinds of data is that the number of faculty in one category does not translate into the number of sections or (more importantly) the number of students they teach. At my college there are about 3 times as many adjunct faculty as f-t t-t faculty, but they only teach half of the classes.

Regarding your remarks @#6:

The Great Hiring ended circa 1969 in physics, but did not end abrubtly in all fields. However, it would be accurate to assume that t-t but untenured faculty made up less than 10%, and probably less than 5%, of the facutly at that time.

However, one major bias is that I know that a teaching assistant doing recitations was not counted as "adjunct faculty" in 1979. We weren't even considered employees as regards social security taxes. Today, those exact same positions are much more likely to be classed as adjunct faculty.

Your guess at the current fraction of t-t faculty who have yet to get tenure could be biased by where you teach. There has, indeed, been a great wave of retirements. More than half of the t-t faculty at my college were hired in the last 5 years. (Faculty with a CC teaching load do not stick around into their 70s or 80s like you see at large universities.) I also know that there is a large number of not-yet-tenured faculty at a nearby R1, including more than 15% of the physics department.