1. Errors in Sampling Sampling Errors- Errors caused by the act of taking a sample. Makes sample results inaccurate. Random Sampling Error Errors caused by the chance in selecting a random sample. Nonsampling Error Errors not related to the act of selecting a sample from the population. They can even be present in a census.

2. Bad Sampling methods Ex: bias samples, voluntary response samples, convenience samples, etc.

3. Common Types of Errors: Undercoverage When some groups of the population are left out of the sample On purpose or by accident Examples: Call-in poll, stopping people @ mall, mailings, standing by back door of school, etc. Sampling Error

4. Processing Errors * Nonsampling error * Mistakes * Examples: - doing arithmetic wrong - typos - recording wrong numbers/info - losing data

5. Response Error * When subjects give an incorrect response * Examples: - lying (especially with sensitive questions) - remembering info incorrectly - don’t understand question - etc. * Nonsampling error

6. Nonresponse Error * Failure to obtain data from an individual in the sample * Happens b/c subjects refuse to respond or can’t be contacted * Examples: - not answering phone/ hanging up phone - not sending back mailing - absent on day of poll - refuse to write answers on a survey * Nonsampling error

7. Wording of questions * When a question: - is confusing - uses big words or technical language that most people don’t understand - uses a word that has more than one meaning and doesn’t clarify - ARE SLANTED towards one response (based on the question alone or a statement with the question * Usually sampling error

8. What to do about nonsampling errors: * For nonresponse… just select another individual * Weight the responses based on who responds Example: If twice as many rural homes respond than urban homes, give more weight to the responses from the urban.

9. Complete the book problems with a partner

10. Ways to take samples Example: At a university, there are 1000 male professors and 500 female professors. To take an accurate sample of the university faculty’s opinions, we decide to take a sample of 100 male professors and 50 female. This sample of 150 professors is a good representation of the faculty. Why is this not an SRS? What was their process?

11. Stratified Random Sample (NOT simple random): 1)Divide population into groups with something in common (called STRATA) Example: gender, age, etc. 2) Take separate SRS in each strata and combine these to make the full sample - can sometimes be a % of each strata

12. Try p. 119 #68

13. Probability Sample- * Chosen by chance * Give a chance to each individual, but not always an equal chance. * We must know what chance each sample has Example: The lottery

14. Questions to ask before you believe a poll: p. 121

15. Try the book problems