Population parameter
Describing the whole population
Statistic
From a sample
Standard deviation
Uses mean or proportion from population
Standard error
Uses sample mean or proportion
When to use T* vs Z* values
Use t* values (confidence level + dF) for means while use z* values (probability) for proportions
Sampling distribution of sample statistic
Distribution of sample means or proportions is normal-shaped even if the population itself isn’t normal. Described by the central limit theorem.
Confidence Interval (CI)
Statistic +- (multiplier)(standard error)
Margin of error
(multiplier)(standard error)
Conservative margin of error
Using 0.5 as p value
Test statistic
(Sample statistic - null)/(standard error)
Standard error for two props
Combined proportion
Type I error
Reject H0 when it is actually true (Lower confidence level/higher alpha)
Type II error
Fail to reject H0 when its actually false (Higher confidence level/lower alpha)