Validity of borough mayor's online poll questioned

May 21, 2018

 

Results from online surveys should be taken with a grain of salt, says Ivan Moore of Alaska Survey Research.
Credit Wiki Commons

An increased presence on social media has been a cornerstone for borough mayor Charlie Pierce since he took office. His administration has posted videos and updates and now, a poll. While online polls might be interesting or entertaining, rarely are they valid.

 


If you check out the borough’s Facebook page, you’ll see a poll asking your opinion on some borough budget stuff. Specifically, what you think is the best option from a list of possibilities. Do you like a sales tax increase or a bed tax or the use of savings to balance the budget, or...well, it actually doesn’t really matter what’s on the list.

"I don't need to see it," Ivan Moore said when I asked if he'd like a link to the survey.

Moore runs Alaska Survey Research in Anchorage, and he’s been conducting polls and surveys there for two decades. Basically, he says online polls don’t have any of the controls that lead to valid results. Moore says pollsters have to make sure that the responses they get actually represent the group that’s being surveyed. In this case, borough tax payers.

“And that is extremely difficult, if not impossible to do with online surveys unless you invite your people and restrict the access to the survey (to) the people you invite. Otherwise, you get people selecting themselves, going on and doing it, and your results end up being unrepresentative and essentially for entertainment purposes only.”

Moore may not have needed to see the poll to understand its shortcomings, but it’s important to parse them out. The mayor’s poll states that he presented a balanced budget to the assembly, that the assembly shot down. Moore says that’s a no-no. In truly random sampling, that kind of context isn’t provided. When his company conducts a phone survey, the first question is simply would you like to participate?

“We don’t tell you ahead of time what the survey is about. So you either say yes or you say no, but you don’t say that based on whether you’re interested in the subject matter, because you don’t know what the subject matter is.”

The mayor’s survey is a series of options, some new, some old and you choose just one. Using borough savings is an option, though it’s not truly savings that would be used. The mayor proposed using money from a land trust fund, and the assembly said no. Another option is to increase property tax on every property owner. The wording on that option has changed since the survey went live just after 3 o’clock Saturday morning. The other options were a bed tax, a cigarette tax, an increase in the sales tax or continue curbing spending and keep cutting.

Of the roughly 400 people who have weighed in, the results are pretty evenly split, with the property tax option getting the least support. Assembly member Kelly Cooper shared her concerns about the survey on Facebook, and the property tax option in particular, which initially said property owners and homeowners. That could be confusing, as oil and gas properties contribute a large portion of property taxes.

“When I saw that it was a sponsored ad and being paid for by borough taxpayer dollars, I felt the need to weigh in. One of the questions was changed midway through the poll, it was reworded. Had this been on the mayor’s (personal) page, I never would have jumped in there and given my input. But when it’s the borough Facebook page, it needs to be accurate and have more clarity on what those questions are and what they mean. It’s just messaging to try and get the results he’s hoping to see.”

Calls to the mayor’s office weren’t returned in time for this story, but on that Facebook post, Pierce states that the poll is "excellent and has been very well received."

Ivan Moore says there are a couple ways to go about either conducting or presenting a poll like this properly. One, establish the right parameters in which you conduct the poll so that it’s valid. As far as presentation goes, he says local TV stations, like KTUU, have been doing it right since online opinion polling became a thing.

“They would always, religiously, when they presented the results on the news, say this is not a scientific survey. Which is basically a particular way of saying these results need to be taken with a grain of salt.”