Test #2 Flashcards
Event
-an outcome/set of outcomes (subset of S)
Law of Large Numbers
-x bar gets closer to mu as n increases
Central Limit Theorum
- no matter the distribution, the sample mean will have an approx normal distribution when the sample size is large
- n of 25-30 is usually large enough
Confidence Level C
-probability that the interval will capture the true parameter in repeated samples (i.e. success rate)
P-value
- probability that the z value would take as extreme/more extreme a value than the observed value, if Ho were true
- the smaller the value, the stronger the evidence against Ho
Finding P from z statistic
- less than: look up z and find probability
- greater than: look up z, find probability, subtract f/one
- two sided:
- Z negative: look up z, find probability, double
- Z positive: look up z, find probability, subtract f/one, x2
Significance Level (a)
-Probability that the test rejects Ho when it’s really true
Comparing a and p
- P less than a: reject Ho, results statistically significant at level a
- P greater than a: do not reject Ho, results statistically insignificant
Margin of error gets smaller when:
- n gets larger (more info about mu)
- SD gets smaller (less variation)
- z* gets smaller (means smaller confidence level, cheap)
Statistical vs practical significance: n
- When n is large, tiny deviations f/Ho (little practical sign) will be statistically significant
- When n is small, large deviations f/Ho (great practical sig) might go undetected (statistically insignificant)
Statistically significant means:
-Probably not due to random chance. Not necessarily practically important
Type I Error
Reject Ho when it’s true
Type II Error
-Fail to reject Ho when Ha is true
Probability def
-The study of random behaviour; unpredictable in the short run but can have regular pattern in long run
Manageable # of probabilities, assigned to each one
discrete probabilities
