Statatics from Election Results

[Bill (Smithville NJ)]

(11-08-2015 03:17 AM)Brian D Wrote:  Somewhat related: each election season there are published polls with a lot of questions, and you can often find the raw data and use linear algebra to discover how many people of each party were polled, and curious glitches/anomalies/flat-out-lies such as how many people of each party answered each question. In some polls you discover wildly fabricated results such as people switching parties between questions, or a negative number of voters of some party, etc.

Hi Brian,

I spent some time wading through one of the early presidential polls to see what it was based on. If I remember correctly, it was an on-line poll of 330 people. The participants were selected from a pool of people who had participated in a minimum of one on-line poll per day for the past year. This got me thinking: Who are these people that take the time to answer a poll daily?

Of course, I find polls a pain and refuse to answer polls and questionnaires as matter of personal policy. Every time I take my car in for service, the dealer mails be a questionnaire about my service visit. The questions are not designed to give any valid information back to the service department. My favorite question was: How well do you feel the mechanic did his work? How would I know? I spent my time in the service waiting area and never met the mechanic or watched him work.

Bill
Smithville, NJ