Confidence Intervals for 1 propn,1 mean, diff in 2 mean, diff in 2 propn, hypothesis tests
Show Answers:
Confidence Interval for : One Proportion
Conditions:
(Binomial conditions)
2 possible outcomes
Fixed number of trials
Constant probability of success
Independant trials
(Assumptions)
Random sample
Greater than or equal to 5
Confidence Interval for: One Mean
Conditions:
(Binomial conditions)
2 possible outcomes
Fixed number of trials
Constant probability of success
Independant trials
(Assumptions)
Random sample
Greater than or equal to 5
Confidence Intervals for: Difference in Two Proportions Conditions:
2 independant random samples
Normal approximation to the binomial holds - more than 5, both samples large
Z test for differences in two means
(Variences Known)
Conditions:
Two independant random samples
If sample sizes are small, samples are drawn from normal populations
Population variances are known
t test for difference in two means
(Variances unknown)
Conditions:
Both populations are normally distributed
Two independant random samples
Population variances are unknown
If either population is not normal, need large sample sizes to appeal to CLT
What is P value?
A p value is the probability of getting a test statisitic at least as extreme as the test statisitic from the sample given that the null hypothesis is true
What does the test statistic measure?
The test statisitic measures the number of standard errors that the sample statistic is away from the value of the population parameter in the null hypothesis