How Does Political Public Opinion Polling Work in Hawaii?
A crash course in Hawaii political polling with Ward Research's Becki Ward.
Rebecca “Becki” Ward, president of Ward Research, is one of the top polling specialists in Hawaii. With the political season gearing up, we asked Ward to walk us through some polling basics.
HONOLULU Magazine: How do you get the phone numbers of the people who are called? Is there a database that your company maintains of people who want to be called?
Becki Ward: The phone numbers are generated two ways primarily: No. 1, using Random Digit Dialing, by taking working phone numbers and alternating the last digit to create new numbers (some of which may be unlisted), and No. 2, through the purchase of lists (voter lists, for example).
We do not keep a database of people willing to do phone surveys, although we do keep a list of those who ask us never to call again. (Note that market research and polling is not subject to FCC Do Not Call regulations, but we keep our own list in order to respect Hawaii residents' preferences.)
HM: Do you call cell phones? What data do you have about the number of people who have only mobile phones or land lines?
BW: Definitely. Telephone surveys without cell phone calling generally cannot generate samples that are representative of the greater population anymore.
We have estimates that suggest that only about 60 percent of households statewide have landlines. And those households become harder and harder to reach, with people not home, screening calls, etc.
HM: What’s the shelf life of a poll? Are there certain questions/pieces of data you can gather in February that would still be fresh in, say, May?
BW: There is no guarantee that any data collected in February would still be accurate in May, because things can happen in that period of time to change awareness, attitudes and behaviors. Having said that, there are also times when nothing happens to change these and the data remains stable. But you have to collect data again to know whether it has changed or not.
HM: What polling pitfalls or practicalities are unique to Hawaii? (People often say that our strong Asian influence makes people reluctant to speak to strangers on the phone about their beliefs.)
BW: There is, sometimes, a reticence to speak among some of the Asian ethnicities in Hawaii. So there are some factors that will increase their comfort level and encourage better participation in telephone surveys—things like an 808 area code on caller ID and an interviewer from Hawaii with a voice that puts them at ease.
HM: What’s a push poll? And does your company do them?
BW: A push poll is one that is really seeking to create public opinion, through the use of loaded language in the questions and other questionable practices (skewed sampling, for example). Then the results of the poll are often released publicly, to suggest that public opinion is different than it truly is. So, no, we do not do push polls. In fact, when we poll for the media (in The Hawaii Poll, for The Honolulu Star-Advertiser and Hawaii News Now), all of the questions are provided word-for-word in the reporting and on the website, along with all of the data, for full transparency.
HM: Whenever I get a call from a polling company, it seems the caller always predicts the call will take 10 minutes. It always takes longer. Is there a length of call that most people will actually respond to?
BW: We try not to exceed 15 minutes survey length and, of course, shorter is always preferable. The challenge is that getting a qualified respondent on the phone is taking longer and longer these days, which increases the costs of polling. So clients, who are paying more than they used to pay, want to get more information from the expenditure. As pollsters, we're challenged with balancing the experience of the people we interview (and the deterioration of data quality in longer surveys) with the desire of clients for greater value.
HM: What is the most common misperception people have about polls? Or the question you always wish people would ask you?
BW: The greatest misconception about political polling—and we're getting into that season—is that we're trying to predict the outcome of an election. Polls represent a snapshot in time. We're asking, "If the election were held today, who would you vote for,” not trying to predict the outcome when votes are cast one month later, for example. Because, again, things can happen in between the time of the poll and the election to change voter awareness, beliefs and preferences.