APST 207 Lecture Notes - Lecture 7: Standard Error, Confidence Interval
Document Summary
Note: remember your p-value interpretation depends on the chart you use. To get more extreme: positive z-score (1-p-value)*2 negative z-score p-value* Z scores on a single value: z= (x- )/s. Z-score on a sample mean: z = ( - )/ sex. Using normsdist to calculate the probabilities they"re married to. If the z score is negative normsdist can be used as is times. If the z-score is positive normsdist can be subtracted from then times 2. Example: using gpa sample of unr sophomores: mean is 2. 71, population of sophomores. To find more extreme we substract this value from 1 (because the z-score is positive) and take it times 2 = ( 1-. 9799)*2 = . 04. The population could have a new mean ( ) We could be dealing with a different population. If it"s a different population it would have it"s own population mean ( ) and standard error (se) Ou sample could be extreme by chance.
