DRAFT POLL

In response to queries about the accuracy of the CRIMSON Poll on the draft and the Vietnam war, the following explanations may help clarify the issue.

The most basic objection to the poll is that the questionnaire reached only 43 per cent of the class. People who filled out the poll, the argument continues, were more highly motivated to get their answers recorded than those who did not bother to answer the questions. Therefore, the logic concludes, most of the people who are against the war had their answers tabulated, while the "silent middle"--the 57 per cent who weren't polled--are probably a good deal more conservative. The poll was therefore not representative of the class.

The procedure followed in administering the poll was aimed at eliminating this particular kind of objection. First, questions were phrased so that they did not evoke a radical response any more than a conservative one. Second, the poll was handed out in House dining halls during lunch and dinner of three consecutive days in mid-December. Each student, as he entered the dining hall, was asked whether or not he was a senior. If he said that he was, he was handed a poll and asked to fill it out. Only about ten or fifteen seniors refused to take the poll--the other 529 filled it out regardless of political penchant. Once they had completed the poll, their name was checked off a master list in order to prevent repeats.

But then why did only 43 per cent of the class answer the poll, if all of the Houses were covered for three days? To begin with, not all the seniors eat in the Houses. Furthermore, some of them had already left for Christmas vacation. But most importantly, only one of the entrances into the dining hall was covered in each House.

Take Adams House, for example. There are two entrances into the dining hall, one from the Gold Coast and the other from Plympton Street. We only covered the Plympton Street entrance. Are the people we missed who enter Adams House dining hall from the Gold Coast significantly less radical than those who entered from Plympton Street? We think not.

Of course, it is possible to argue that the population which eats in the dining halls is a very peculiar sample. For this reason we contacted students in Dudley House who do not usually eat in the Houses. In order to avoid over-representing Dudley, we set the average number of seniors who answered from each House as the target number of Dudley House students to be contacted.

But in all fairness it is possible that pollsters became tired of asking everyone who walked into the dining halls whether he was a senior (it is extremely embarrassing to ask a tutor if he is a senior), and the poll may be a little more selective than the above would indicate. But not much.

The second main objection to the poll is more obvious. Percentages given were not of the total number of people who answered the poll, but of a slightly smaller number who answered the question being referred to. This should explain what appeared to be numerical miscalculations of about one per cent.

In retrospect, the CRIMSON article should have been more explicit about procedures and we thank our many careful readers for having pointed out the shortcomings of the poll's explication.