Nothing Is Ever Easy

Conducting a survey is not a simple task. There are how-to courses--online and probably in classrooms, too--to teach people how to design a survey, how to get survey responses, and how to analyze survey results. One wonders if any of those courses warns against what happened with the Seventh Street Park Survey, meant to find out what kinds of improvements the community wants for the park.

Back on August 18, at the end of a Hudson FORWARD meeting, Sheena Salvino and Branda Maholtz of the Hudson Development Corporation (HDC) handed out a few paper copies of the Seventh Street Park Survey. One of the items on the survey as it was distributed that night proved problematic. It was this one, which asked respondents to rate, on a scale of 1 to 10, the improvements they wanted for the park.



The item was confounding. People were unsure how to answer if they wanted more than "Remove Fencing Around Fountain" but less than "Completely New Concept." What number represented simplify and restore it? Salvino and Maholtz recognized the problem and changed the item, so that the scale went from "Historic Restoration" to "New Concept/Redo."

Some less dramatic changes were made to the survey as well. Respondents who didn't live in Hudson were asked to tell where they did live, and the options for the question "How do you use 7th Street Park?" were reordered, moving "Getting Children outside" closer to the top.

Changing a survey after its distribution has already begun seems an unwise thing to do, and in this case, it has had some very wide ranging consequences. Although only three or four people at the Hudson FORWARD meeting completed the survey and handed it back, one person at the meeting took the survey with him, made numerous copies, and distributed the copies to others. Yesterday, when seventy-five of those surveys were delivered back to HDC completed, the immediate reaction was that those surveys could not be counted because they had been completed using the wrong form. Predictably (and understandably) that response caused a brouhaha that reportedly involved the mayor and the city attorney before it finally calmed down. It seems now that a way has been figured out to analyze the results even though they are coming in on two different versions of the survey.

When Gossips spoke with Salvino yesterday, she spoke of her concern to preserve the integrity of the survey. It would seem that the integrity was already compromised when items were changed after some paper copies had been distributed. Still, the problems of analyzing the results should not be unsolvable. If the paper surveys are to be scanned by machine, which seems unlikely, the change in order of the possible responses to one of the questions will require that item to be analyzed by hand. 

The "Scale of 1-10" item presents a bigger problem. There are asterisks on the survey indicating which items the respondent is required to complete. On the original survey, the "Scale of 1-10" item had no asterisk; on the revised survey, it does. People completing the original paper survey could have just skipped that item, and perhaps it should be skipped on all surveys when analyzing the results. There are many other questions that get it the same information.  

The sad thing is that, after this snafu, people may be unlikely to have much confidence in the survey's findings, whatever they are.
COPYRIGHT 2014 CAROLE OSTERINK