Martini, Manhattan, Scotch and soda? What's your preference? Two members of the Social Relations Department will soon be happy to serve your choice of liquor in an attempt to test the effects of alcohol on fantasy.

A letter signed by David C. McClelland, professor of Psychology, and Michael Kahn, teaching fellow in Social Relations, offers four rounds of before-dinner cocktails to participants in the survey.

On four separate occasions the subject will be served a specified number of cocktails "in the privacy and convenience of his own home." Thirty minutes later he will be shown four pictures and asked to write a brief description about each of them. A different number of cocktails will be served at each sitting, and all liquor will be provided by the researchers.

The letter claims that "no one has ever before studied the effects of alcohol on fantasy, although it seems sensible to assume that these are after all the most important and interesting effects the drug has."

Part of the letter defends giving the tests in the subject's own home. It objects to the standard liquor-effects test given "in a sterile and impersonal laboratory with specified amounts of medicinal alcohol out of a graduated flask." The new system, the letter asserts, will make it possible to test the effects of alcohol in situations ranging from heavy-drinking parties to quiet cocktail affairs.

Although the letter was sent out over his signature, McClelland said last night that it partially "misrepresents the true nature of the experiment."

"We are investigating the effects of many social situations," McClelland asserted, and this letter places too much emphasis on the liquor phases of the experiment. Kahn, who McClelland said may have authorized the letter, was not available for comment.