Survey Quality Questioned by Public...

Grandpa,

I'd like to know a little more about the design of the community survey for Grand Terrace. Is the survey scientific? What statistical tests will be used? (relationship between 2 measurement variables {correlation, regression}, differences between means {independent, related samples}, or comparison of frequency distributions)? Will the sample of 300 be random? Normal? How can 300 phone calls generated via telephone with a population of residents (who work in the day time) ensure a non-biased population sample? Is the sheer nature of phone calls already biased? How many people use their cells phones instead of home phones, and won't that skew that data?

These are the questions that need to be asked.

Gramps Answers:

Yep, 100%... Statistics Can Lie
Remember the Basic Lesson:
All Drug Abusers drank Milk, thus, all Milk Drinkers will become Drug Users.
Of course this logic is wrong, as wrong as surveys with directed leading questions.