Q: I’m not a statistician, but I remember that when doing a survey, a 30 percent return was statistically valid.
So I’m wondering with North Kato, a city of 13,000 and about 5,500 households, how can survey responses from just 500 people or so be even close to being valid?
Wouldn’t a 30 percent return be something like 1,500 responses? Wouldn’t a survey firm know what it takes to be statistically valid?
And what system was used to “randomly” pick their 1,500 people? How many were non-English-speaking? Did they have translators for Spanish or Somali speakers or were minority groups ignored? Did they call only landlines or how did they get cell phone numbers?
Just curious how the city pays for a survey that is FAR from being statistically valid at all.
A: The survey in question, done earlier this year, was initiated by the city of Mankato as part of its five-year strategic planning process. North Mankato and Mankato Area Public Schools accepted Mankato’s invitation to join the polling effort, and all three used the same Boulder, Colorado-based polling firm — National Research Center Inc. Mankato paid $18,000 for the citizen survey, the district paid $6,000 and North Mankato spent more than $10,000.
The municipal questionnaires covered a wide variety of topics, ranging from whether people felt safe on the streets to whether they were satisfied with garbage service. In Mankato, the attitudes of residents will be one factor as the City Council sets its priorities for the next half-decade.
And in North Mankato — during some recent debates about whether new restrictions should be placed on citizens’ ability to offer complaints about the city or about city staff during council meetings — City Administrator John Harrenstein has cited the survey as evidence that the vast majority of residents are happy with their…