MATH 1F92 Lecture Notes - Lecture 12: Null Hypothesis, Type I And Type Ii Errors, Simple Random Sample
Document Summary
When we want to make inferences on two quantitative data sets that are dependent or (cid:858)(cid:373)at(cid:272)hed pairs(cid:859), (cid:449)e ha(cid:448)e to use an entirely different approach. For inferences on two independent means, our point estimate for the difference between the two population means was (x bar1 x bar2) with a mean of mu1 mu2 and a standard. This (cid:449)as see(cid:374) as esti(cid:373)ati(cid:374)g the (cid:862)differe(cid:374)(cid:272)e i(cid:374) 2 (cid:373)ea(cid:374)s(cid:863) deviation of sqrt(s1 (mu1 mu2). Now, since we have matched pairs, and therefore have dependence, we can no longer take the avgs separately. This will be referred to as estimating the (cid:862)(cid:373)ea(cid:374)s of the differe(cid:374)(cid:272)es(cid:863) ((cid:373)ud). To obtain a sampling distribution for (the sample average of differences): we take the avg of the differences in each sample, this will be known as our . This is our: we take the difference between the 1st sample data and the 2nd sample data.