uses and misuses of statistics. use vs. misuse describe data compare two or more data sets ...
TRANSCRIPT
Chapter 1.5
Uses and Misuses of Statistics
Use vs. Misuse
Describe data Compare two or
more data sets Determine if a
relationship exists between variables
Test hypotheses Make estimates
about population characteristics
Sell products that don’t work properly
Attempt to prove something true that is really not true
Grab our attention by evoking fear, shock and outrage
“There are three types of lies – lies, damn lies,
and statistics”
“Figures don’t lie, but liars figure”
Describing Misuse
Using very small samples“three out of four dentists recommend Crest toothpaste” If only four dentists were surveyed, that could
just be chance. The result could be quite different if 100 were surveyed
Suspect Samples
It is important to know how the subjects were
selected Studies using volunteers have a built in bias
because1. Particular socioeconomic background
(unemployed)2. Spending time in another environment
(college students and retirees)3. Generalizing results to both men and women4. Opinion polls that require phone or mailed
response
Volunteer Samples
When intact groups are used for a sample,
they may not be representative of the entire population (classroom)
When all customers within a certain time frame are used for a sample, they may not be representative of the entire population (time)
Convenience Samples
There are four commonly used terms that are
loosely called averages. (mean, median, mode, and midrange)
Depending on the data set, these can vary greatly
Ambiguous Averages
Statistical distortion can occur when different
values are used to represent the same data.
For example, a politician might say “During my administration, expenditures increased a mere 3%” While someone running for the same position might say “during my opponent’s administration, expenditures increased a whopping $6,000,000”
Changing the Subject
A claim that used detached statistics is one in
which no comparison is made.
For example, “our brand of crackers has one-third fewer calories”… than what?
Detached Statistics
Many claims attempt to imply connections
between variables that may not actually exist.
For example:“Eating fish may help to reduce your cholesterol”“Studies suggest that using our exercise machine will reduce your weight”“Taking calcium will lower blood pressure in some people”
Implied Connections
Misleading Graphs
Survey questions can be phrased in ways the
influence the way people answer them
For example: a survey question like “Do you think the Union School District should build a new football stadium?”
If it were phrased “Do you favor increasing school taxes so that the Union School
District can build a new football stadium?”
Faulty Survey Questions
Yes, when used properly….
It’s like a car. If you use it to get to where you need to go, then it’s good. If you use it to run over people you don’t like, not so good…
So are statistics good?
http://
www.youtube.com/watch?v=FP6ZP0DM1po
http://www.politifact.com/
For example