Practical Applications of Some Recent Developments in Statistical Methods

FEW, if any, statisticians can manage to keep up to date and cover all the different journals where articles which might be of value to them may possibly appear. Certainly those whose work is concerned with the solution of practical problems cannot spend their time, both covering the literature and applying themselves to their everyday work. It seemed therefore, that it might be helpful to a few members if I were to do some of the preliminary sorting out of recent publications for them. Obviously it is not possible for me to be familiar with more than a small part of the material, but because of the peculiar nature of my work I have formed certain views about the usefulness of some of the newer techniques. Naturally this review cannot be anywhere near comprehensive and it must have also a very biased point of view which derives from the fact that it is possible only to scan the literature for items which offer possibilities of application in one's own field. I may therefore have missed some very important developments and I hope that this is going to be pointed out in the discussion. I have hopes that perhaps it might be possible for this society to have a meeting such as this every year to review the development of the preceding five or six years. My discussion this evening will deal with all developments in non-mathematical terms and cover not so much the methods as the sorts of problems to which the methods might be applied. I am sure that most people are, like myself, inherently lazy, but that if sufficient reward is offered they are capable of quite a lot of effort. I am assuming that this is so. I hope to dangle the carrot of what might be achieved by some of the newer methods which could not be achieved (so far as I have been able to judge) by the older. I hope that if this carrot proves sufficiently attractive, members will pursue it. There is, of course, little correlation between the different techniques published during the last six years. It would be pointless in an essentially practical discussion to seek refuge in a presentation based on date order. I propose therefore to sort out the material into three classes.