1.5 Sampling Can Go Wrong

All of the stuff that we have done so far makes sampling and statistics seem simple, but some problems can arise.

Example

A prediction poll mistake of the 1948 presidential election proclaimed that Thomas Dewey was the winner over Harry Truman. Have you ever heard of President Dewey? No, but we've all heard about President Truman. Something went wrong with the survey.

Two things can go wrong: sampling errors and nonsampling errors.

Definitions

Sampling Errors

These are errors caused by the act of selecting a sample. They cause sample results to be different from the results of a census. (Note: a census occurs when the sample is the entire population.)

Nonsampling Errors

These are errors that are not related to the sample from the population, and these anomalies may also be present in a census.

Sampling Errors

1. Random Sampling Error: This results from chance selection in the simple random sample (error is due to chance).
2. Nonrandom Sampling Error: This results from improper sampling
   • convenience sampling
   • selecting from a yearbook or a phone book
   • door-to-door surveys during the day

Nonsampling Errors

1. Missing Data:
   • refusal to answer survey
   • subject is not available for survey
2. Response Errors:
   • subject may lie or remember incorrectly
   • subject may guess if they do not understand question
3. Processing Errors:
   • math errors
   • coding data incorrectly
4. Effects of Data Collection Procedure:
   • timing
   • exact wording
   • media
     • mail (low response)
     • telephone (fast and economical)
     • personal interview (more in-depth responses)

Homework

(Scott Street's section only)

Pages 42-46

1.40, 1.43, 1.46, 1.49 (a-c)

(Solutions)

Please direct all questions regarding STAT-110 to your instructor or to the director of STAT-110, Dr. Todd Ogden at ogden@stat.sc.edu.

Mail comments regarding this presentation to W. Scott Street, IV at street@stat.sc.edu.

© 1996 by W. Scott Street, IV