Why poorly controlled polls are consistently misleading
If online polls are to be believed, from two-thirds to three-fourths of the first people to respond to almost any polling question pertaining to pit bulls favor strong breed-specific legislation to protect themselves, their children, and their animals from attack. Yet time and again the polling numbers reverse as the numbers of respondents rise.
This is no new pattern. It has been seen time and again throughout the 20-year history of online polling, and may now be seen on multiple online polling web sites almost every day.
One might surmise that pit bull advocates are merely slow readers, and therefore take more time to respond.
Or one might surmise that pit bull advocates typically mobilize a militant cohort nationwide and perhaps worldwide effort to stuff the electronic ballot boxes in response to any indication that breed-specific legislation might gain political support. This would be more accurate, since there are in fact numerous pit bull advocacy Facebook pages, listservs, and websites promoting just such electronic ballot box-stuffing efforts.
Of most concern to the electronic ballot box stuffers seems to be breed-specific legislation, of any sort, which might inhibit the commerce of backyard pit bull breeders. This means practically all pit bull breeders, since keeping large numbers of pit bulls together in puppy mill conditions tends to present an occupational safety hazard to the puppy millers that most of them conspicuously avoid. Pit bulls are the breed most often impounded in mass neglect cases, as ANIMALS 24-7 has documented in 32 years of logging cases throughout the U.S. and Canada, except in puppy mill raids, which rarely impound any pit bulls.
Online polls are advertising vehicles
The nature of online polling plays into electronic ballot-box stuffing, because––with rare exceptions––online polls are not really designed to gather useful, accurate, or even vaguely representational data. Online polls exist to collect “hits,” which in turn expose the participants to advertising.
Designing an online poll to exclude multiple responses from the same individual is as easy as allowing only one response per email address, or even better, only one response per IP address. Programs such as Survey Monkey put the tools to do so in anyone’s hands, free of charge, but most online pollsters avoid such technology because the more enthusiastic a respondent becomes about voting more often than the dead in a Chicago graveyard, the more often the respondent will see the ads that accompany the polls. Since the price and salability of online advertising are dictated by the “hit” count, anything that inflates the “hit” count is a plus for the pollsters, no matter how nonsensical the outcome of the numbers.
“Do you want a pit bull’s owner for your neighbor?”
Observed New Jersey resident Serafima Yan recently, contemplating the polling trends, “Those polls do not work. Real polls should be done door-to-door with just one question: Do you want a pit bull’s owner for your neighbor?
If the approach Yan recommended has ever been taken in a door-to-door survey, ANIMALS 24-7 is unaware of it. And going door-to-door can be quite risky. Ask any mail carrier: 2,782 mail carriers were bitten by dogs in 1995, at a time when far more carriers delivered door-to-door, while 5,699 mail carriers were bitten in 2011, several of them mauled quite severely.
In 2009, however, ANIMALS 24-7 asked almost the question Yan suggested, in a poll designed to test the accuracy of online polling. ANIMALS 24-7 had noticed that uncontrolled online surveys pertaining to pit bulls were usually posted soon after publication of local reportage about dog attack fatalities and disfigurements, about two-thirds of which involve pit bulls.
A tightly controlled small survey
What ANIMALS 24-7 wondered was whether the usual early poll tilt in favor of breed-specific legislation reflected residual opinion, likely to be found in any controlled survey at any time, or only reflected the shocked reaction to the local incidents. Finding a definitive answer might have required polling thousands. But a tightly controlled small survey of people from places where there had not been recent pit bull attacks or legislative debate might provide clues, ANIMALS 24-7 reasoned, and ANIMALS 24-7 has been doing tightly controlled small surveys to investigate various humane issues for decades.
What ANIMALS 24-7 asked was:
As a general preference, I would prefer not to live next door to a pit bull terrier.
Yes No No opinion
ANIMALS 24-7 surveyed several dozen volunteers from among the membership of two professional societies having no direct involvement with animal issues. The volunteers were not told in advance what they would be asked about. They were asked only three basic questions, but membership directory information permitted tracking many biographical variables.
Except for having more formal education than most Americans, the respondents in composite mirrored U.S. demographic norms, including in urban/rural balance, geographic distribution, income range, and response from visible minorities. More men responded than women, reflective of the membership of the societies, so proportional weighting was used to achieve gender balance.
68%, including 71% of the men and 62% of the women, agreed that they would prefer not to live next door to a pit bull. 28%––29% of the men but only 21% of the women––did not object to living next door to a pit bull.
Four percent of the respondents had a pit bull, about equal to the rate of keeping pit bulls in the general population. 76% of respondents had pets, far above the U.S. norm of 57%; 24% had children, all of whom also had pets; 24% had neither pets nor children.
No men were undecided about living next to a pit bull, but 18% of the women were undecided, all of whom had pets but no children living at home.
Among all respondents with pets, 69% would prefer not to live next door to a pit bull. Among all respondents with children at home, 80% would prefer not to live next door to a pit bull. Among all respondents who had ever had children, 86% would prefer not to live next door to a pit bull.
ANIMALS 24-7‘s 2009 survey findings were validated on August 14, 2012 when Miami voters overwhelmingly defeated a well-funded and aggressively promoted ballot measure meant to repeal the 23-year-old Miami-Dade County pit bull ban.
The repeal attempt attracted just 37% support–the most lopsided failure of a ballot measure endorsed by major national humane organizations in at least a couple of decades.
The limited available polling data suggested that my 2009 survey data, coming from a nationwide small sample, accurately predicted the splits in public opinion in Miami three years later.
Probably the 2009 findings would still be generally applicable today, since the volume of pit bull attacks on both humans and animals has more than doubled, visibly increasing public concern.
Please donate to help our work: