A Panorama of Statistics

9 The beguiling ways of bad statistics II

In the previous chapter, we described four settings in which bad statistics can beguile us. Here, now, are four more.

When assembling evidence to support some controversial proposition (e.g. during a political debate), any number is better than no number

In beginning a public lecture in 1883, the British physicist William Thomson (later Lord Kelvin) offered the following dictum: ‘… when you can measure what you are speaking about, and express it in numbers, you know something about it; but when you cannot measure it, when you cannot express it in numbers, your knowledge is of a meagre and unsatisfactory kind …’ (online at [9.1]).

Thomson was underlining the indispensability of measurement to the progress of the physical and natural sciences. Today, however, his assertive message is being advanced indiscriminately – bewitching (for example) many of the social sciences and, in particular, the practice of government and public administration.

No social or economic policy proposal is felt to be adequately supported nowadays, unless there are numbers among the evidence. More unfortunately, this sentiment has been reinterpreted by lax thinkers in many areas of public debate to imply that if there are numbers among the evidence, those numbers should suffice to clinch the matter. From such an illogical position, it is then only a short hop to the view that it is not the intrinsic fitness‐for‐purpose of the numbers that is the vital ...