ABSTRACT
Many statistical projects arise from requests for help from specialists in other disciplines. Such specialists will have varying degrees of expertise in statistics, and the ability to communicate effectively with them is most important. Such work is usually called consulting, although Cox (1981) has questioned the overtones to this word and advocated greater emphasis on collaboration. It is certainly true that full participation in a collaborative study is more rewarding than trying to give a quick answer to a cookbook question. However, in my experience, the statistician must be prepared to give advice at a variety of levels and the following remarks are concerned with statistical consulting in its widest sense.
