Flawed Survey

Oct 09, 2019

To the Editor:

There are serious flaws with the survey reported in your Sept. 25 issue (“EJ Elementary Public’s Choice in Local Survey”).

First, it was not a random poll (which in itself disqualifies it, in terms of accurately reflecting public opinion), but the interpretation of the results shows a general lack of understanding of statistics.

With a sample size of just 150, the results would be, at best, accurate to plus or minus 8.16 percent. For example, on the question of how to dispose of the property of whatever school is closed, the three sets of opinions (44 percent versus 35 percent versus 33 percent) form a statistical dead heat. Symptomatic of the problem is that these numbers add up to 112 percent!

When the real vote is counted, on Dec. 10, the results may come out quite differently, but I hope they at least add to 100 percent.

Tom Sales

Roselle, N.J.


Comments (0)
If you wish to comment, please login.