The Press. MONDAY, SEPTEMBER 30, 1901. USE AND MISUSE OF STATISTICS.

Press, Volume LVIII, Issue 11083, 30 September 1901, Page 4

The Press. MONDAY, SEPTEMBER 30, 1901. USE AND MISUSE OF STATISTICS.

Th© ordinary investigator who goes to statistics for figures in support of his arguments generally gets what he looks for. The figureg can nearly always be made to suit. We have a conspicuous example of this in the letter from tbe Rev. John Mackenzie, which we publish in another column. This gentleman starts off with tho rooted conviction that crime is increasing in New. Zea--land, in consequence of the absence "of the reading of the Bible from tha State schools. H. goes to the New Zealand "Year-book" for figures in support of his argument, and he finds the following table :;■*- ' c__-_a_Es be_?ojlb magistbates. Proportion per 1000-----Y ' of mean Year. Number. population. 1882 ... 22,108 ... 43.41 -1885 ... 22,709 ... 39.61 1888 l ... ' 18.829 ... 30,88 1891 ... 17,189 ... 27.29 1894 ... 17,3_3 .... - 26.55 .. 1897 .:. 19,949 ... 27.65 1898 . ... -2,234 ... 30.19 Now, tlie reverent! gentleman, in making use of this table, does not start where the book" does, and tell us that the proportion of charges was 43.41 per 1000 in 1882, and was 30.19 in 1898. That would show a decrease of crime instead' of an increase, and would be fatal to hjs arguments.,. He picks out the lowest percentage in this fluctuating table, namely 25.55 per 1000 in 1894, and comparing that with 1898, boldly produces it as proof that there has been an increase of crime "during the last ten years," and our 'ungodly" schools are the cause. He adopts precisely the same course in regard to summary convictions, and convictions in the Sup_eme Court. Instead of starting in 1883 with the "Yearbook," or in 1891, which he took for the charges before magistrates, he now picks out 1891, finding that better" adapted for his purpose. Of course figures can be made to prove anything twisted in this fashion, but it is scarcely honest argument, and ws take leave to say that a clergyman must" indeed be blinded by his partisan bias to allow himself to make use of such methods. For ourselves, we do not see that there is any necessary connection at all between ths criminal statistics and Bible reading in schools. An increase of the police force, or an increase of activity on their part, would increase the nmnbsr of apprehensions and convictions. A few years of prosperity almost invariably leads to more drinking, and this swells our criminal statistics. Sa far as the crime statistics in the "Yearbook" bear on the question at all, howeverj thay tell against Mr. Mackenzie rather than

in his favour.. The Registrar-General, sum* ming up the figures during the years 1888, I 1891, 1894, and 1898, distinctly states thata comparison of. the number of per- ■ sons summarily convicted during ' the year "shows a slight decrease in the pro- i portion of these convictions per 1000 of the population on the total, and also in regard to offences against the person. Offences against property show a slight rise. Dealing with the convictions in the Superior Courts for all offences, the proportion per 1000 of population is a very little higher for 1898 than for 1888." But a fairer, test is not to take the number of convictions, but of persons imprisoned after' conviction. We want to know whether we are breeding more criminals in this colony, not whether tbe criminals we possess axe being committed frequently to gaol Mr -oackenzie, judging by his letter, evidently thinks that if ten people were each committed ten times to prison for small offences during the year, it would shojr a higher degree of criminality, and -be ''a greater reflection on our education system than if the same ten persons were each committed to penal servitude for life. This is not so, and taking the number of the dis- - tinct persons imprisoned after conviction there has been, as the Registrar-General tells us, "since 1889 a decrease of 17 per

-■..-.-.- . -. _ < -cent, in the number, and a reduction of "12 per 10,000 in the proportion to population." But a fairer test still is the percentage of crime among ihe New Zea-land-born over'fifteen yeacrs of.age. If our system of secular education has any effect at all in. increasing crime, it is among this section of the population that it would be seen, not in immigrants from other countries. Now the fact distinctly stated by the Registrar-General is that, while the New Zealand-born over fifteen years formed 44 per cent, of the total population, New Zealanders of all ages constituted less_ than 15 per cent of .the total number received in gaoL We quoted this fact in our previous remarks on this but Mr Mackenzie, has ignored it as completely as he did the years in the Registrar-General's tables which did not square with his theories. There are, as we said before, numerous good arguments in favour of Bible-reading in schools, some of which the reverend gentleman has used in his letter which we publish to-day. He should confine himself to these general arguments in future,. and leave statistics severely alone. Either he has not mastered the A B O of their proper use, or els. he is utterly devoid of conscientious scruples as -to the way in which they should be employed.