Latest Chance News

Statistics really don't work well with humans where the answers are based upon the caprice of the respondent. There never is enough weight placed on lying, answering to please the asker, or plain ornery folks who want to skew the results. I looked on their site where they showed the responses of people judging a debate. They correctly pointed out that there were two questions that were identical but worded in a positive/negative form. The questions were- who won the debate? equal 17% neither 7%. It seems that fewer people wanted to sound negative half empty but preferred half full.

Random sample criteria are also very hard to determine with people. Even after studying statistics, I still never really understood how to determine those criteria. Maybe that's why all these polling companies keep their data and methods so secret. They don't want you to know how they picked the people. It surely isn't tossing 300 mil names in a hat and picking a few thousand. What are the criteria?

I liked statistics with objects. Seems much more reliable. Process control, even though human operators sometimes mess up statistics, was also an area I enjoyed. Overall quality control was a toughy. Humans again. Anyway. I'm beginning to ramble...

Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.