Kevin MacMillan: I wouldn’t read too much into online polls
April 22, 2014
Every so often, I get questions about our online polls. Comments center on if there is any outside influence to choosing questions, if we're hurting a project or sensationalizing an issue by asking a question — or something along the lines of, "Did you see the results of your poll on (insert issue)? Unbelievable! You've got to do a story."
Our polls can be found at http://www.tahoedailytribune.com, and depending on what page you're on (Regional vs. North Shore/Truckee vs. South Shore), you're asked a different question. They're just like any online poll you see on CNN, Fox News, ESPN, Rolling Stone, etc. — we ask a question, you provide a response through multiple choice.
Our most recent poll asked, "Which addiction do you feel is the biggest problem for Truckee/Tahoe?" With April being Alcohol Awareness Month, and a slew of panel discussions, prescription drug take-backs and other events happening in our region, I felt it was a worthy question to ask.
On their face, the answers reflect several opinions. Of the 1,324 votes, 27.57 percent (356 votes) point to meth being the biggest problem, while 23.87 percent (316 votes) vote for alcohol, and 16.92 percent (224 votes) indicate "we don't have an addiction problem."
A related question was published to the North Shore/Truckee pages: "Do you consume alcohol?" The results from 750 votes are perhaps a bit more eye-opening: 31.07 percent (233 votes) leads the way with "Yes, every day," while 23.2 percent (174 votes) is for "I choose not to consume alcohol." Third place, 18.8 percent (141 votes), is for "Yes, maybe once or twice a week."
I say "on their face" because our online polls — much like those similar to the aforementioned news sites — really should be taken only at face value.
If you look below each poll, you'll read the following disclaimer: "Please Note: Poll results are not scientific and are for entertainment purposes only."
That's a fancy way of us saying that, frankly, you shouldn't believe everything you read on the Internet.
Not only are our polls random, but digital technology these days has Swiss cheese for a defense system. From hackers to identity-theft crimes to emails warning us of Nigerian princesses in peril, there are loopholes everywhere.
The same can be said for our web polls. There is nothing stopping someone from having a lot of time on his or her hands, developing some sort of algorithm and devising a way for a computer to vote thousands of times.
Or, putting it another way, if a poll is up for 21 days, you can vote 21 times (once a day) from the same computer. Or, conceivably, 210 times a day if you run around to 10 computers. And then, multiply that by whatever number of friends and family who you convince to run around to the same number of computers, and you're going to have a lot of votes.
Point being, you really can't put much stock into a random, non-scientific poll. Anyone who does is either lazy or, more than likely, choosing to represent only portions of an issue to make his or her opinion stronger.
With all this in mind, I present the following example: In a poll published March 14 that was live for 39 days, 1,613,473 votes were registered for the yes-or-no question, "Are you in favor of Measure R, the proposed Truckee trails sales tax increase on the June primary ballot?"
Yep. That's right. More than 1.6 million votes. And 97.63 percent (1,575,302) answered "Yes," an option that received 41,455.3 votes per day.
Measure R will be on the June 3 primary ballot. To put that 1.6 million in perspective, only 3,867 people voted in the Measure J election in 2012. Not only was that during the general election (which historically sees more voters than a primary), but it was for a far-more controversial tax measure.
So, did enough people to fill the city of Philadelphia (estimated to have a population of 1.5 million-plus as of 2012, America's fifth-densest city) legitimately bestow faith in why Truckee's little Measure R is a good idea?
I seriously doubt it. So, while I'm not accusing anyone of deliberately fudging the numbers and am merely using Measure R as an example, whenever you see someone quoting an online poll, make sure you check to see what kind of survey it is before you form an argument around its results.
— Kevin MacMillan is managing editor of the Sierra Sun and North Lake Tahoe Bonanza newspapers; he may be reached for comment at email@example.com. Follow him on Twitter @Kevin1MacMillan.