Nonpartisan Education Blog

Campbell’s Law is like Campbell’s Soup: Ubiquitous and Innocuous

You became familiar with Campbell’s Law when only a few days old and by age two had mastered it. As a parent, you would have witnessed your children discovering, learning, and employing Campbell’s Law even before they could form coherent sentences.

Campbell’s Law is obvious. It is a truism. It postulates that when sentient beings are aware of the fact that their behavior is being measured–and they may be judged by that measurement—they may try to manipulate the measurement. Infants may be more prone to cry if they know it will get them what they want. Job-seekers may magnify their successes and minimize their failures on their resumes. Lonely Hearts may post their most flattering photos on the computer-dating website instead of the least flattering.

Campbell’s Law applies whenever there is measurement, judgment, and awareness, which is pretty much all the time and in every aspect of our lives.

Despite its ubiquity and banality, US educators cite Campbell’s Law as a rationale for eliminating external standardized testing that is, after all, sometimes used, fairly or not, to judge their performance. Given that Campbell’s Law applies whenever an educational test is administered–particularly so when a test has consequences–and given that Campbell’s Law postulates that measurement will be “corrupted” (i.e., that sentient actors may try to game the system in their favor), we should not measure in situations where Campbell’s Law applies, they say.

Problem is, Campbell’s Law almost always applies. Those educational tests called “no-stakes” still harbor plenty of incentive for manipulation, particularly among educators with less-than-pure morals. Cheap or lazy education administrators might prefer to use the same test forms year after year to save money, while the tested students inform the not-yet-tested students about the content of the test. Personally ambitious education administrators might use the same test forms year after year purposely with little security so that student scores rise over time; those administrators can brag about the score increases as evidence of their managerial prowess.

Educators who cite Campbell’s Law as a rationale for eliminating external standardized testing may not tell you that Campbell’s Law applies equally well, if not even more forcibly, in school classrooms absent external standardized testing. Are teacher evaluations free of the vagaries of Campbell’s Law? Of course not. For example, grades are susceptible to inflation with ordinary teachers, as students get to know a teacher better and learn his idiosyncrasies and how to manipulate his opinion. That is, students can hike their grades for reasons unrelated to academic achievement by gaming a teacher’s personality or grading system.

There are a number of problems with teacher evaluations, according to numerous researchers. Teachers tend to consider “nearly everything” when assigning marks, including student class participation, perceived effort, progress over the period of the course, and comportment. All of which are available for exploitation by manipulative students willing to employ Campbell’s Law. Actual achievement in mastering the subject matter is just one among many factors. Indeed, many teachers express a clear preference for non-cognitive outcomes such as “group interaction, effort, and participation” as more important than test scores. It’s not so much what you know, it’s how you act. Being enthusiastic and group-oriented gets you into the audience for TV game shows and, apparently, also gets you better grades in school. And non-cognitive measures can easily be faked.

Teachers always “teach to” something. If they are not teaching a required curriculum matched to a standardized test, what are they teaching? …and why should we be so sure that it is better than the required curriculum? In the absence of common standards and tests, teachers still teach to something, but that something is arbitrary; it is whatever the teacher personally prefers. And, why should the taxpayers pay for that?

What teachers and schools do in the classroom absent any adherence to common standards is not necessarily any “broader” or “deeper” than what happens with common standards. Indeed, it is likely to be “narrower” as it is determined by nothing more than one individual’s personal preferences. Nor is it necessarily any better, more profound, or more beneficial to the students. It’s merely more arbitrary.

Cheating in regular classroom work has become epidemic. The overwhelming majority of students admit to cheating in polls. Teachers and schools are ill-equipped to monitor or detect most cheating. Meanwhile, the Internet makes cheating far easier than it used to be.

Testing opponents argue that teachers have an incentive to cheat on high-stakes tests and no incentive to cheat otherwise. Nonsense. Social promotion and grade inflation provide the contrary evidence. In surveys, the majority of teachers claim overwhelming pressure to give high grades to and promote undeserving students.

We will not eliminate the influence of Campbell’s Law by eliminating external high-stakes standardized testing. Campbell’s Law will still apply in the un-monitored, un-measured classrooms that anti-testing critics idolize. But, its effects will be hidden. The primary result of an elimination of external testing: the public will no longer have access to objective evaluations of their children’s education. Our only evaluation source will be educators themselves, who can tell us whatever they wish.

— Richard P Phelps