STAB22H3 Lecture Notes - Lecture 10: Ipsos, Combined Oral Contraceptive Pill, Epidemiology

• School: UTSC
• Department: Statistics
• Course: STAB22H3
• Professor: Caren Hasler

Document Summary

Sampling distribution: statistics (sample count, sample mean x, sample proportion p) can change from one sample to another, they are random, therefore they have a distribution, called the sampling distribution. Summary: if x ~ n ( ; ), then: xn is normally distributed, with (cid:373)ea(cid:374) , with standard deviation (sd) < , the sd of xn decreases as n increases. That is, xn esti(cid:373)ates with greater precision when n is larger. Su(cid:373)(cid:373)ar(cid:455): suppose x has (cid:373)ea(cid:374) and sd . For large n: xn is normally distributed, with (cid:373)ea(cid:374) , with standard deviation (sd) < , with sd that decreases as n increases. That is, x(cid:374) esti(cid:373)ates with greater precision when n is large. Central limit theorem (clt: draw a srs of size (cid:374) fro(cid:373) a populatio(cid:374) with (cid:373)ea(cid:374) a(cid:374)d sd . Normal approximation for counts and proportions: draw a srs of size n from a large population having population p of success.