Symptoms

It is hard to know where to start, perhaps the most obvious flaws (especially with the lecturers at Victoria) occur in the mathematical field; graphs are drawn without axes ("this is learning," says the lecturer and sloshes a great curve of white across the blackboard) or, if that luxury is allowed, they are very rarely labelled!

More generally, we may assert no facet of personality can be assessed quantitatively, for reasons that will appear later. But does this worry the psychologist ?—not a bit.

In order to introduce "mathematical accuracy" into his results, he simply rates them on seven-point or 100-point scales.

(One lecturer at Victoria, who shall be nameless, thought that marking essays out of 10 was unrealistic. His solution: the essays were marked A, B, C, D, E, and each letter was extended to plus or minus—A minus distinct from B plus—giving in effect a 15-point scale).

It may be objected that the exact sciences also use this idea (e.g. cloud-cover is rated on an eight-point scale) but we cannot permit the analogy since a physicist's ratings may always be verified if required, albeit in a more laborious way, but anything in the way of verification of strictly subjective ratings such as maturity or intelligence using another, more general system, is impossible.

We cannot even permit definitions using statistics, e.g. "intelligence"—"A general factor entering all abilities", because (a) The relative ability of the candidate depends on the particular test used to measure that ability, and (b) The units of abilities (i.e. the "spreads" in frequency vs. ability curves) are completely arbitrary.