About All Those Airport Surveys
Polling indicates that building a new single terminal at Kansas City International Airport is unpopular, yet we seem inundated with surveys that purport to show that opinions are changing. It’s tough to say, because we don’t necessarily know if the information is trustworthy. What we do know, however, is not comforting.
In opinion survey research, the number of people surveyed is less important than randomness. A survey of 2,000 people who themselves decided to complete an online questionnaire, for example, may be less valuable than a survey of 200 people who were contacted at random. Without adequate randomization, survey results may over-represent the views of a group of passionate partisans.
Unscientific survey data—data lacking adequate randomization—has played a big role in the debate about whether or not to build a new single terminal at MCI. Some of it is due to passionate partisans; some of it is due to questionable reporting. Much if not all of it creates the sense that building a new terminal is more popular than it is.
Consider some questionable reporting. The Kansas City Business Journal published a front-page piece indicating that the average TSA wait time at MCI was 29 minutes. Anyone who has flown from Kansas City would be suspicious of that figure, and other average wait times should have suggested to reporters that the data was problematic. (Jacksonville and Phoenix airports both listed an average TSA wait time of zero!). Kevin Koster, who served on the Mayor’s Airport Terminal Advisory Group, was skeptical, and he followed up with his own research (see here and here). In short, the TSA reports that MCI’s actual average wait time was 3.63 minutes. The discrepancy is due to the fact that the Business Journal data came from a website in which travelers report their own wait times without any independent verification. The Business Journal basically relied on an online survey that has little or no scientific validity—perhaps travelers whose wait times were uncharacteristically long were overrepresented among the website’s respondents.
Even some of the other, likely scientifically valid polling that has been reported fails to meet the ethical standards set by the American Association of Public Opinion Researchers (AAPOR) because it does not disclose all the questions asked.
Then there is the Aviation Department’s own online survey. At the City Council’s August 10 Business Session, department officials provided an overview of their presentations to groups across Kansas City about the need for a new terminal. After their presentation (which included a long discussion of the Aviation Department’s own scientifically invalid survey of session attendees), Councilman Jermaine Reed offered a startling admission [starts 19:26],
I can tell you that every time I fly I certainly try to get on the survey and make sure I mark everything bad about the airport. I don’t say anything good. I put no, no, one, one, zero, zero. What can we improve? Everything. In my comments so … I shouldn’t probably tell you that. If you see the comments it is probably me.
Proponents for a new terminal are aware of public misgivings. Sadly, rather than having the serious and legitimate concerns of skeptics addressed, we have seen favoritism, secrecy, and now questionable polling that creates the misleading impression that the public is on board with their plans. Kansas City deserves better.