AP Statistics Confidence Intervals Flashcards
Terms : Hide Images [1]
| 8774557258 | What conditions must be checked if σ is given? | -SRS: data must be random -Normality: given or CLT (20 or more values) -Independence: N≥10n | 0 | |
| 8774557259 | What conditions must be checked if σ is unknown? | -SRS: data must be random -Normality: single-peaked, bell curve -Independence: N≥10n | 1 | |
| 8774557260 | What conditions must be checked if estimating proportion? | -SRS: data must be random -Normality: np≥10, n(1-p)≥10 -Independence: N≥10n | 2 | |
| 8774557261 | Conclusion | We can say with C% confidence that mean/proportion is within the interval (Estimate±margin of error) | 3 | |
| 8774557262 | Test statistic when estimating µ (σ given) | z* | 4 | |
| 8774557263 | Test statistic when estimating µ (σ unknown) | t* | 5 | |
| 8774557264 | Test statistic when estimating p | z* | 6 | |
| 8774557265 | Confidence interval | estimate±(test statistic*standard error) | 7 | |
| 8774557266 | Standard error when estimating µ (σ given) | σ/√n | 8 | |
| 8774557267 | Standard error when estimating µ (σ unknown) | s/√n | 9 | |
| 8774557268 | Standard error when estimating p | √(p^(1-p^)/n) |  | 10 |
| 8774557269 | How to find the test statistic | -Use confidence level to find tail areas (1-C/2) -use invT or invNorm | 11 | |
| 8774557270 | Degrees of freedom | -n-1 -needed only when using t test statistic | 12 | |
| 8774557271 | Paired t procedures | -x and s come from the difference in the pairs -solve as normal -conclusion is the interval of difference | 13 | |
| 8774557272 | Comparing two means | -estimate will be sample 1 est - sample 2 est -standard error becomes (√(σ²₁/n₁)+(σ²₂/n₂)) -must square standard deviations (replace σ with s if σ is unknown) -must specify which population was subtracted from which in conclusion | 14 | |
| 8774557273 | Comparing two proportion | -estimate will be sample 1 prop - sample 2 prop -standard error: √(p^₁(1-p^₁)/n₁)+(p^₂(1-p^₂)/n₂) -must check normality conditions for both populations | 15 |