Oct

19

 Many full moons ago Mr. Bollinger asked me if I could find the source for the remark about Joplin that I quoted. I don't remember what I wrote then, but the gist of it was that Joplin was the only one of the ragtime players who had the guts to write down what he had improvised the night before on the piano. I still have not found the source of the remark; but I have no doubt that it was true.

I also have no doubt that we are not going to be seeing Bollinger and Joplin's kind of candor about methods of thinking - where "public opinion" is concerned. The current response rate for telephone polling is - allegedly - 9%. That is the number the Pew people disclosed about their own polling in 2012; and it is now taken as current gospel. In 1997 when the Pew people also disclosed their response rate, the number was 37%. It is likely that the actual response rate now is roughly 1 in 20; but that is a mere guess. The actual polling may be even less.

But, have no fear, the Pew people assure us that their numbers are still good because the results track with the data acquired by the government in its census and other surveys, where the response rate is unquestionably high enough to be accurate. "Pew Research Center devotes considerable effort to ensuring that our surveys are representative of the general population. For individual surveys, this involves making numerous callbacks over several days in order to maximize the chances of reaching respondents and making sure that an appropriate share of our sample are interviewed on cellphones. We carefully weight our surveys to match the general population demographically."

Thurston Trowel III replies: 

What is your point on low response rate?

A low response rate merely means the polling outfit has to make more phone calls to get the requisite representative sample.

It does not reflect negatively on the scientific accuracy of a poll in any way, if this is what you were suggesting. No idea if that is the case but it seemed to be where you were going and apologies if wrong.

As to Joplin writing down improvised ideas, musicians everywhere for decades have recorded sketches as part of the creative process of developing ideas. I do this all the time. Whether one does so on paper, magnetic tape, hard drive or cell phone is largely irrelevant, and a matter of personal preference and convenience.

Stefan Jovanovich retorts: 

Dear Third Shovel: Joplin had the courage to share his method at a time when the people playing ragtime were selling it as magic. They were not sharing the tricks of their trade. Mr. Bollinger remains exceptional in much the same way. The solution you offer– the polling people just have to make more calls - sounds logical until you ask the question the broker posed to his client. Who are the new people going to be? The reliability of opinion polling has been founded on the assumption that the sample will be homogeneously responsive, that the response rates will not vary no matter what the respondents answers are. But that is no longer the case. Making greater effort becomes a statistical frontal assault; the added costs gain no ground. Then there is the further issue of outright bias. How do you know what the proper weighting should be if you choose to ignore the only data that you have– the results of the most recent elections before this one? 


Comments

Name

Email

Website

Speak your mind

Archives

Resources & Links

Search