PSYCH 100A Lecture Notes - Lecture 6: Type I And Type Ii Errors, Null Hypothesis, Standard Deviation
Document Summary
What you compute: standard error: sigma/ square root of n. Used for: comparing a sample mean, xbar, to a known population values. Establish your expectation about where groups from the population will fall. Set boundaries for where 95% of scores will fall, only valid for a 2-tail test. 95% ci: xbar +/- (z critical value x standard error) What you got - expected/ spread of sampling distribution. If you score is outside the boundary then reject the null hypothesis otherwise fail to reject the null hypothesis. What you compute: standard error= s/ square root of n. Already used n-1 when calculating standard deviation. Used for comparing a sample mean, to a known population mean, but we are using an estimate of the population standard deviation, s. Because we are using s we must use the t-table. Set boundaries for where 95% of scores will fall only valid for 2-tail tests. 95% ci: xbar +/- (t critical value x standard error)