Outside the Forensic Box: The Scientist as Eyewitness

Source: Matthew J. Sharps

Nevil Maselkyne was a great astronomer. Appointed in the 18th century as the British Astronomer Royal, he had previously traveled widely, often under horrible conditions, to conduct worldwide astronomical observations. He was probably the first in the world to measure the Earth's mass; he was a great scientist in every way.

Which is why it's so strange that he fired his assistant, David Kinnebrooke, in 1796.

Poor Kinnebrooke died at age 30 in 1802. Whether his dismissal at the hands of the great man Maselkyne had anything to do with it is hard to say; but at the same time, we are very aware, in the modern world, of the nasty effects of stress on the human body and nervous system (e.g., Sharps, 2022).

Yet the whole thing is very sad. Kinnebrooke, an innocent victim of the psychological laws of observational astronomy, as well as of the choleric Maselkyne, demonstrated a seminal but frequently ignored fact of experimental psychology.

What poor Kinnebrooke had inadvertently exhibited was the personal equation in science—the fact that we don't all see the same things in the same way.

This poses a hell of a problem for science, which is supposed to be completely objective, immune to influences of personality and other extraneous factors. Isaac Newton, a bit of a mystic himself, helped to contribute to the idea that science is always perfectly objective—if you do science right, everybody everywhere will have no choice but to agree on the results.

Really?

Gregor Mendel was on to a good thing in genetics—but he still appears to have fudged some of his data to get the equations to work out right. There are many similar examples in science history, not necessarily nefarious. Perhaps with the best of intentions, data can be so messy and theory so straightforward—so let's ignore the bits that don't fit.

Bad idea.

Maselkyne and Kinnebrooke were using the venerable "eye-and-ear" method to establish precise stellar transit times. The given observer noted the clock time to the nearest second, then counted seconds to the beat of a pendulum while watching a given star "move" across his telescope's field. Very simple, fantastically precise—what could go wrong?

It never seems like a good idea to ask that question.

Kinnebrooke's observations were always at least an appreciable fraction of a second different from those of Maselkyne. Every time.

We now know that this discrepancy would have occurred for anybody, but let that go for the moment. We of the modern world, perhaps somewhat accustomed to the public exercise of massive egos, might think that Maselkyne simply assumed Kinnebrooke to be wrong because, hey, Maselkyne was Maselkyne, and Kinnebrooke wasn't.

It's certainly possible that ego, on the part of Maselkyne, was involved here; during his tenure as Astronomer Royal, it pretty much rained assistant astronomers who somehow didn't measure up. Yet this quintessentially modern view may do an injustice to the great astronomer's memory. Like practically every other educated person of the 18th and early 19th centuries, Maselkyne assumed science was purely objective, essentially immune to the individual differences of different scientists. Assuming direct observation, in itself, to be an invariant gold standard (as many later scientists would also assume), he dismissed the heretic Kinnebrooke, who failed to measure up.

The problem is that there was nothing wrong with Kinnebrooke's observations. Two decades later, Bessel at Konigsberg would show that different individuals' observations, even in these simple contexts, would frequently vary by over a second. This is absolutely normal; a personal equation, based on weird biopsychological phenomena, lay between great astronomers and the precise nature of objective reality.

Psychology Essential Reads

Is a Texas Bitcoin Miner Making People Sick?

What Is a Scientific Model?

It works like this: You see the phenomenon at second 1.0, and I see it at second 1.85. If you're Maselkyne, you get to fire me, but both of us are in the grip of the personal equation, the psychological factors that contribute to error variance in any given situation. It's a huge problem: Our individual nervous systems are never going to let us get it exactly right.

Many factors may contribute to this phenomenon—differences in stimulus anticipation, or complication, involving the mental integration of both visual and auditory stimuli in the "eye-and-ear" method. Many factors may intrude, but the fact is that we're pretty much stuck with the personal equation.

So much for the purely objective science suggested by the earlier work of Newton and perhaps by anybody today "believing" in science, as opposed to actually practicing it.

Eighteenth-century science was absolutely clear on the idea that purely objective information would set everything right forever. It is certainly true that science, properly conducted, is the most powerful method we have for understanding the world around us. Yet scientists are also witnesses to science, and we now know that the principles of eyewitness psychology can have an appreciable effect on scientific results.

The personal equation, as a component of observer bias, pervades the sciences from astronomy to zoology, with medicine in between (e.g., Brinkman & Podolsky, 2021). Of course, this factor also extends to the science of psychology. This is a big problem; to paraphrase an authoritative source (Gregory [Ed.], 1987, pg. 611), if the importance of this whole nightmare had been better realized, the objective faith in stimulus-response paradigms that characterized much of 20th-century psychology might never have happened.

Today, we still try to find ways around the personal equation. Larger numbers of observers have been suggested "in the hope that individual variation would be diluted by still more observers" (Brinkman & Podolsky, 2021). In strict experimental work, the problem may be attenuated by concepts of error variance. We hope.

Yet in criminal investigation and in arenas of similar importance, such as medicine, no such statistical remedies may be available. We are still dependent on the faculties of the individual observer.

The personal equation, persistent and ubiquitous, soldiers on. Both practitioners and consumers of science, eyewitnesses to important findings, might benefit substantially from the dose of humility that this fact provides.