How statistics can help

For to be possessed of good mental powers is not sufficient the principal matter is to apply them well. The greatest minds are capable of the greatest vices as well as of the greatest virtues, and those who proceed very slowly may, provided they always follow the straight road, really advance much faster than those who, though they run, forsake it. 

Descartes, Discourse on the Method of Rightly Conducting the Reason and Seeking for Truth in the Sciences, Part I. 

Let us assume, however, that you agree with Descartes — after all, you are stiU reading — and think that not everything that looks obvious is really so obvious. In this case, if you are involved with experimentation, be it in academic fife, in industry, or in a research and development laboratory, we believe this book can be very useful to you. With it you can learn to perform experiments and draw the appropriate conclusions from them more economically and efficiently. 

In the following chapters we will discuss some relatively simple and easy to use experimental methods. These techniques might even appear obvious after you think a little about them, but this does not detract from their merit or effectiveness. To make this point clearer, let us consider a 

Suppose a chemist wishes to obtain the maximum jdeld fi om a certain reaction, and that only two variables, the temperature and the concentration of a certain reagent control this reaction. In the nomenclature that we shall adopt in this book, the property of interest, in this case the reaction 3deld, is the response. The variables that — in principle, at least — affect the response (that is, the temperature and the concentration) are the factors. The function describing this influence is called a response surface. The objective of the research worker is to find out the values — the levels — of the two factors that produce the largest possible response. How would you set about solving this problem  

---