Measurement: does whatever exists exist in some amount, and can all mental qualities be measured?

Edward L. Thorndike, answered from the texts and cited to the page.

The position in brief

My guiding maxim is this: whatever exists at all exists in some amount; and whatever exists in some amount can be measured. If a child has more of a quality — more knowledge of arithmetic, more reading skill, more intelligence, more of a virtue even — than another child, then there is a difference of degree, and a difference of degree is precisely what measurement captures.1

The refusal to measure the things of the mind is not high-mindedness; it is the refusal to be exact, a preference for vague words over definite quantities. So I built scales — for handwriting, for reading, for arithmetic, for intelligence itself — that let us say how much, and compare, and track improvement, and test whether our teaching works.

A science of education is impossible without measurement, as a science of anything is; and the objection that the mind is too fine or too sacred to be numbered is the enemy of knowledge. Measure what you can, define your terms by your measures, and let the numbers discipline the loose talk.234

Sources

  1. In the instance just quoted A could say: " The 600 children in my school imder the old method made an average gain of 4 per cent, in a year in arithmetical knowledge, 3 per cent, in interest and zeal, etc. Under the new method the figures are 6 per cent, 7 per cent., etc." To some extent we can already make accurate objective measurements of mental traits.
    Educational Psychology, pp. 18–19
  2. Children who went to school two thousand days will be found to have been from the start more interested in scholarly work, and more capable at it, than those who went only one thousand days. The young men who learn the carpenter's trade are selected by its attractions, and are different from those who are selected by the different attractions of clerical work.
    Education A First Book, pp. 211–213
  3. In the case of merit of handwriting, the differences between A and B, B and C, and A and C do not even pretend to be put B B THE MEASUREMENT OF INDIVIDUAL DIFFERENCES 159 in terms so as to allow comparison. As the records stand, anyone must get the differences transposed into terms of some unit A — before he can calculate- — at all. Since the measures of A — L A, B and C are objective, this can be done.
    Educational Psychology Vol 3 Mental Work and Fatigue, pp. 159–160
  4. So, again, to describe the business ability of a certain man, we may distrust the possibility of telling his ability on scales of amount, and simply call him the Napoleon of finance.
    Educational Psychology Vol 2 The Psychology of Learning, pp. 73–74

Where the argument is made

  • An Introduction to the Theory of Mental and Social Measurements (1904/1913)
  • The Measurement of Intelligence (1927)
  • Educational Psychology (1913-14)

How the argument was received

Thorndike founded educational and psychological measurement, and the principle that mental qualities differ in degree and can be scaled underlies the whole enterprise of testing and assessment. The maxim is also the target of a standing and serious critique: that measurement can reify a complex, multidimensional quality into a single number, mistake the score for the thing, and lend spurious exactness to constructs (like a unitary "intelligence") that the measurement itself helped invent — the charge Stephen Jay Gould pressed in The Mismeasure of Man. Modern psychometrics both extends Thorndike's program and is chastened by the critique, distinguishing good measurement from the naive assumption that anything named can be validly scored. The principle stands; its uncritical application is what is contested.