The campaign of Democratic gubernatorial candidate Mike McWherter has released results of a recent statewide poll that shows that the majority of Democrats are still undecided.

56% of respondents said they were undecided. If one is to be 99% confident in the poll, the poll probably has a margin of error of about +/- 5% based on the number of Democratic primary voters and the sample size used.

If one is to be 95% confident in the poll, the poll probably has a margin of error of about +/- 4%.

The range of undecided Democratic primary voters could be anywhere from 51% to 61%.

I calculated the margin of error based on the number of Democratic primary voters in the 1998 Democratic gubernatorial primary: there were 298,466 voters in that primary. Given that this population has probably grown significantly since 1998, the margin of error is probably closer to 6%.

In a Democratic party primary public opinion poll that is considered scientific, researchers must be concerned with getting the proper sample size of possible voters. Social scientists tend to aim for a proper sample size that would yield a margin of error of +/- 3%, and researchers would say that the results have a high level of confidence.

I cannot find the data used by McWherter’s poll, so I am approximating their margin of error (or “confidence interval”)

So, I did some calculations using SurveySystem.com’s “Sample Size Calculator.”

Click here to input the numbers yourself.

So, if you were to do a public opinion poll assessing the status of Democratic voters in the Democratic gubernatorial primary, you would want to know how many Democratic voters are likely to vote in the primary. The only data I could find was for the 1998 Democratic primary: 298,466 people voted in the 1998 Democratic gubernatorial primary.

There are two other numbers you need to consider before you realize how many people you need to call to find out who is leading in a reliable public opinion poll.

First, there is the confidence interval. The confidence interval is the plus-or-minus figure usually reported in newspaper or television opinion poll results. For example, if you use a confidence interval of 4%, and 47% percent of your sample picks an answer you can be “sure” that if you had asked the question of the entire relevant population, between 43% (47-4) and 51% (47+4) would have picked that answer.

The next number to consider is the confidence level. The confidence level tells you how sure you can be. It is expressed as a percentage and represents how often the true percentage of the population who would pick an answer lies within the confidence interval. The 95% confidence level means you can be 95% certain; the 99% confidence level means you can be 99% certain. Most researchers use the 95% confidence level.

The McWherter campaign’s poll had a sample size of 602 Democratic primary voters. I wanted to find out if that number is adequate to generalize about Democrats in Tennessee. So, enter a general population size of 298,466 (the number of people who voted in the gubernatorial primary in 1998), enter a margin of error (or confidence interval) of 3%, and enter a confidence level of 95%. The McWherter campaign’s poll, if calculated in 1998, should have had a sample size of 1,063.

The McWherter campaign’s poll had a sample size of 602, though.

That would be for 1998, and one can assume that the number of Democratic primary voters in Tennessee has grown since 1998.

The McWherter campaign’s poll does not have a 3% margin of error (more likely a 4%), and we can not even be 95% sure of its results at this margin of error.

To be 99% confident in the McWherter poll with a +/- 3% margin of error, the McWherter poll would have had to have called at least 1,838 Democratic primary voters, almost three times the sample size of the poll.

I found there to be a lack of confidence in the McWherter campaign’s poll, which his campaign touts as giving him front-runner status in the Democratic gubernatorial primary.

The McWherter campaign commissioned the poll from GarinHartYang Research Group (GHYRG).