Ksample pop < 0:00 quantitative data cak heighe calc 0.很敏 很多令 Sample mean X mean heighe X Categorical data sample proportion distribution of female proportion Gender • 样本统计量 Sample statistics Sampling distribution. 抽样分布 distribuition of individual data point Include all individual population Include a distribution small number of individual ⇒ sample. distribution distribution of sample statistics Include s.s. from all possible sample 在满足特殊条件的情况下 灭和戶显现特殊分饰/* 利用这种特殊分布,做推断. Statististical inference 对于今 当:1) SRS.样本被随机抽出。 unbiasness无前性. ⇓ E(P)=P.←全体百分比 样本百分比的均值 当220.1N: 10% condition. ⇓ Independent sampling = PCI-P n 当292/10日 24-p) 2101 Large wome P~Binan PNN condition 当)SRS. 2) 10% condition 3) Large count 三条件满足时, P ↑~NCP. J) 当D SRS MX = Mx. pop, size Sample 2n<0.1N 6x = x Size 3) 13,90 或 population enormed 成立时 지 In →XNE. 文~N(x) 2x0.1. N. independent sample. ΣΧ 几 ΣΧ Var (x) = Vav() == Var(EXⅰ) == (Var(x)+Var(X2)+-+Var() 2 ==.n. Var(x) Var(X) 42 4dx)= ↑ stdix) Std(x) N 6x = 6x ~~ n N independent Σχin Xi identically Σχί - X= x = n distribution N~ 当ん>,30. or population is normally distributed XiONEXi~NEX~~ 文~N(Mx,会) 对于f. DSRS ⇒ ECP)=P. P=XX成功的个数。 X~B(u,p) n E(x)= np std(x > ppc1-p). ECP)==P 样本百分比 std(p) = std (ㄧㄤˇ)=/std(x)) 当up7x0.n(1-p) 7,10 X~B ⇒ XNN い X n P رہرہ। P=A~N PNNLP, JP) n P=0.80 N=800. N=60. 0.52. (1)定作假设. Hypothesis. Testing 19.验条件. (3)算值. (4)做判断. 假设检验 小定假设. 他觉得 P20.8 你觉得 P<0.8 ✓ 先定HA⇒ Ha: P<0.8 再定Ho Ho:声=0.8 こ 通过假定 Ho成立时出 现了不该出现 的现象推 Ho. 证明 Ha Ho永远是“三”. Ha畏">""<"或"" 確是“……” 2.验条件. 1) SRS. V 2) 10% n=60<0.1N=80 V V 3). np=60x0.527/10.. ncl-p)=60×0.48710- 3) 算P-value.作文定 Ho 成立. 08x0.2 60. P-value是当Ho成立时,取到该样本 与比样本更符和的方向样本的概率 PCP<0.52) Lower O = normaled upper 12 6: N