STA 309 Study Guide - Midterm Guide: Homoscedasticity, Sales Promotion, Statistical Hypothesis Testing
Document Summary
T-test: used to determine whether or not the mean of a population is significantly different from the mean of another; uses the sample sta(cid:374)da(cid:396)d de(cid:448)iatio(cid:374) (cid:894) (cid:895: nu(cid:373)e(cid:396)ato(cid:396) = st(cid:396)e(cid:374)gth of the (cid:862)sig(cid:374)al(cid:863) o(cid:396) diffe(cid:396)e(cid:374)ce i(cid:374) (cid:373)ea(cid:374) (cid:448)alue. De(cid:374)o(cid:373)i(cid:374)ato(cid:396) = st(cid:396)e(cid:374)gth of the (cid:862)(cid:374)oise(cid:863) o(cid:396) a(cid:448)e(cid:396)age de(cid:448)ia(cid:374)ce f(cid:396)o(cid:373) the (cid:373)ea(cid:374) If the signal (numerator) is strong relative to the noise (denominator) the t-test is stronger. If n is large, then use normal (>30) Same way we do normal dist; find t and +t so area is between them: when we have quantitative data, calculate sample mean (cid:4666) (cid:4667) Sampling distribution of mean is normal with mean at pop mean and sd( = ) 10 times abs value of kurtosis use to find p-value. Paired when we have 2 measurements from a single sample drawn form single population; (cid:862)lined(cid:863) or (cid:862)matched(cid:863) up on dataset.
