Flashcards
Statistics 1b 1
Cards in this deck (10)
- 1approaches to statistical inference-Frequentist
approach
-Bayesian
approach - 2Frequentist approachFrequentist approach:
ensures that we draw correct conclusions in a
fixed percentage of the studies,
in the long run - 3Bayesian approachBayesian approach:
quantifies evidence in a particular
data set for a particular hypothesis - 4Which approach is more useful (Bayesian frequentist)One approach can be more useful than the other , depending on your goals
- 5When you draw many, many samples, the sample mean will be Normally distributed withjämän
- 6What informs us about how accurate our estimation is?margin of error
- 7decreasing a sample size by 1/4, changes the margin of error by..doubling it
- 8What can you technically do if the margin of error is too big?- lower the CI (smaller C and therefore smaller z)
- Increase the sample size
reduce the st dev by carefully controlling the measurement process or by focussing only on a portion of a large pop. - 9What do you have to do to the sample size to to make the margin of error half its original value?multiply sample size by 4
- 10conditions to use CI- SRS (note that undercoverage and nonresponse rates etc can create errors in the data
- no bias
- try to search and correct for outliers before using the formula. If not possible, another formula less sensitive to outliers should be used
- pop needs to be (relatively) normal
The higer the sample size, the more accurate the formula
- pop st dev is known
Information
- Description
- Flashcards on Statistics 1b 1, created by Adrian Mayer on 15/01/2018.
- Created by
- Adrian Mayer
- Views
- 1
- Rating