More Dan Jones Antics

(c) 2006 Deseret News, used here under the fair use laws.Okay, blaming Dan Jones for all of the problems with the Deseret News’ report that Utahns don’t want tax cuts for businesses isn’t enrtirely fair.  The Deseret News should carry some blame too.

Specifically, I am concerned with the graphic presented in the article (also shown in this post).  I hate to repeat many previous posts slamming Dan Jones and the Deseret News.  But I must bring up at least three points on this.

1. Note that this has a margin of error (see the small print) of +/- 5%. This means that it is statistically believed that the “real” answer falls somewhere between 47-57% of the population actually don’t support the tax cut. However, if you take the fact that you combined 2 percentage points (+/- 5% times 2) then it really is somewhere between 42-62% don’t support the bill. A real statistician wouldn’t have reported this, because they can’t say anything accept that the people are undecided.

2. Look at the last few words of the question, “saving businesses up to $50 million.” Well, the liberals have managed to convince most people that business is bad, and so they here this last part and they don’t like the idea of businesses getting a cut. They voted with emotion because of the biased question. They were tricked out of voting with their heads and realizing that a tax cut to business is a tax cut to the business user, and we are all business users.

If you doubt my assumptions, I invite you to look at the other graphic that they present in the article. The top portion is very revealing.  Again look it the last few words of the question, which reads, “saving . . . electrical producers and their ratepayers $3.4 million dollars.”  The participant realizes that he/she is a ratepayer, and they want to save $3.4 million, so they suddenly favor it.

You might further argue that they use similar verbiage in the question about cable TV.  I would argue that the difference is reflected in the fact that not all participants were cable subscribers.  So, they leaned towards the businesses are bad model drummed into our heads.

3. Finally, and I have said this in previous posts, but note that they only surveyed 406 people. Any real statistician knows that for the entire population of Utah they need at least 2000 participants to have a completely representative sample.

Just some thoughts from my statistical 101 days and from when I taught a 300 level college level course in research methods.

Possibly Related Posts:

  • Delicious
  • Facebook
  • Digg
  • Reddit
  • StumbleUpon
  • Twitter

2 comments to More Dan Jones Antics

  • David

    OK, just a counterpoint to consider. The Deseret News would not likely pay for a more thorough poll that would be required to really answer the question. For the level of accuracy/detail most people expect from a newspaper a poll like this is sufficient.

    The good news is that the question, sample size, and margin of error are all included so you can decide that it is crap.

    • However, where is the responsibility of the Deseret News to avoid publishing such crap in the first place. For me, it should be very high. However, the Deseret News continues to show that truth is not a value to them.

Leave a Reply