Part I. Producing Data Chapter 2. Samples, good and bad

advertisement
STA1013 section 7 (2003 Fall) Seo-eun Choi
chapter 2 summary
2003. 8. 27 Wednesday
Part I. Producing Data
Chapter 2. Samples, good and bad
• BAD samples : voluntary sample, convenience sample
• GOOD samples : simple random sample, systematic sample, cluster sample, stratified sample
2.1 How to sample badly
• voluntary response sample : Town Talk example - allow people to call in rather than
actively select its own sample
it chooses itself by responding to a general appeal. eg) write-in or call-in opinion
polls
• convenience sample : orange example - selection of whichever individuals are easiest
to reach
• result: biased result - it favors systematically favors certain outcomes
2.2 Simple random samples
Problem, problem...
• voluntary response sample : people choose whether to respond
• convenience sample : the interviewer makes the choice
• PROBLEM : in both cases, personal choice produces bias
• REMEDY : allow impersonal chance to choose the sample (no favoritism by the
sampler, no self-selection by respondents)
Choosing a sample by chance attacks bias by giving all individuals an equal chance to be
chosen.
All have the same chance to be in the sample.
ANY SIMPLE WAY? use hat – simple random sampling
SRS (simple random sample) of size n
• it consists of n individuals from the population
1
• n individuals are chosen in such a way that every set of n individuals has an equal
chance to be the sample actually selected
• It gives each individual an equal chance to be chosen.
• It gives every possible sample an equal chance to be chosen.
What if population is SO HUGE?? - use random digits
1. Label : label individuals in population
2. Table : choose n numbers from Table A (pp.545,546) starting any line between 101
and 150
2.3 Can you trust a sample?
• We have more confidence in results from an SRS, because it uses impersonal chance
to avoid bias.
• Was sample chosen at random?
• No bias is the best, little bias is the better.
2.4 Exploring the web
• www.misterpoll.com
• www.randomizer.org
2.5 Statistics in summary
• sample : in order to get information about population
• biased samples : convenience sample, voluntary response sample
• bias : it favors systematically favors certain outcomes
• how to attack bias? : use randomness (eliminate favoritism)
• simple random sample : simplest sample of random sample
2
Download