| 1. Is the sample representative of the Population ? | |
| - Is the
"sampling
frame" representative? - Is the sample biased? Statistical bias
is broader than personal
bias.
A very common bias is "nonresponse bias." |
|
| 2. Is the sample large enough? | |
| Sample Size for Forecasting a Close Election | |
| 3. Do the questions tempt respondents toward biased ansuers? | |
| 4. What is the agenda of the people collecting and presenting the data? | |
| 5. Do the respondents know the answers to the questions? | |
| 6. Is there a standard to compare the results against? | |
| -Weakness of before-and-after | |
| 7. What are the alternative explanations for the results? | |
| No Explanation
(Random Chance) Transient Explanation Other Explanations |
|
| 8. Is there a statistical "lie" in the presentation of the results? | |
| Note on How to Lie with Statistics | |