Scientific Poll Results

Alright all - the results are in. Previously I asked for a bit of help in deciding whether or not I should wear a bow-tie to the Interns-are-no-longer-the-lowest-on-the-totem-pole party, and they are pretty one sided. There are, however some important issues that must be discussed in relation to the reliability and validity of results.

(this may be a tad boring to some - you can safely skip to the next paragraph) First, it is likely that the results are highly reliable and we can safely assume that if we poll the same population again we will get roughly the same results. The question was clearly stated and exhaustive. We do, on the other hand, have some validity issues. First, the sample was not randomly selected, but a sample of convenience, and not even a snowball design. Even if it was random, we only have a n of 7, which is well short of the typical minimum of 30 in order to insure statistical significance. Second, while the question was clear, it was also poorly constructed. Quite obviously it is a leading question that may have increased the likelihood of socially desirable responses, despite the fact that it was an anonymous poll.

In other words, it is not clear whether or not the poll has validity despite being reliable. As such, I will have to defer to some other authority on the subject in order to conclusively reject the null proposition - that I should wear a bow-tie. Because most authorities are very constrained by demands on their time, I will have to use the closest convenient authority. In this case, I happen to live with a Dr. Although she only knows medical stuff - somewhat useful at times - she will have to suffice. Below are the results of the poll, but final adjudication will have to wait at this time.