l82 ECOLOGY AND LIFE HISTORY OF THE COMMON FROG 



but that it was too laborious. Perusal of these papers suggests that 

 in each case, the authors have arrived at their conclusions independently, 

 for none of the authors I have just cited refers to the work of the 

 others, although all are in fact mutually relevant. 



I have alluded above to the idea, frequently held, that statistics is a 

 subject that is difficult and beyond the grasp of many people. I believe 

 that there must almost always be something wrong in this idea. To 

 begin with, statistical tests differ enormously in their difficulty. The 

 very important chi-squared test in a 2 X 2 table is so simple that a child 

 can do it. What must be distinguished is the difference between the 

 difficult, the intricate and the laborious. Few statistical operations 

 require elaborate mathematics: most need only simple arithmetic. 

 There is sometimes difficulty in focusing the concepts so as to make a 

 choice of the appropriate method, particularly in the analysis of 

 variance. Some operations are intricate, but not difficult. For 

 example, the solution of the simultaneous equations that led to the 

 equation in Chapter 8 by the Doolittle method required four sheets of 

 mathematical paper pasted together, so that all the intricate workings 

 could be put on one sheet. But all these workings were simple 

 arithmetic, and this part of the operation did not take many hours. On 

 the other hand, the correlation tables from which the variances and 

 covariances needed for this operation took many months to complete. 

 This work was very laborious, fairly intricate, but quite easy. The 

 next error that I suggest may be found is that a statistical operation is 

 thought of as similar to the calculations that come at the end of a day's 

 research, when a few percentages summarize the results. In fact, an 

 elaborate statistical operation is the research itself, and may take 

 months or years. Disappointment is bound to follow if this is not 

 realized, but the time needed does not make it difficult but only 

 laborious. 



Statistical Books 



This subject is so important that it deserves a paragraph to itself. 



Being myself a poor mathematician, I think I know a difficulty ex- 

 perienced by those who resemble me in this respect. They have not 

 found the right books. I fmd that one statistical topic is most easily 

 understood from a book by author A, but another is, for me, better 

 explained by author B. It is quite possible that with two other topics, 

 the position of the two authors would be reversed, and other people 



