作者zxc95 ()
站內Statistics
標題[問題] 假設檢定
時間Sun Dec 20 00:35:57 2009
Some nurses in County Public Health conducted a
survey of women who had received inadequate prenatal
care. They used information from birth certificates
to select mothers for the survey. The mothers that
were divided into two groups : 14 mothers who said
they had 5 or fewer prenatal visits and 14 mothers
who who said they had 6 or more prenatal visits.
Let X and Y equal the respective birthweights of
the babies from these two sets of mothers and assume
that the distribution of X is N(μx,σ^2)and the
distribution of Y is N(μy,σ^2).
請問如果我假設H0:μx-μy=0
那我H1要怎麼假設?
是要設小於0還是大於0?
我承認有點不懂題意
是要實驗孕婦在懷孕期間受到關照程度
來跟嬰兒體重看是否有顯著差異嗎?
請教高手
謝謝
--
※ 發信站: 批踢踢實業坊(ptt.cc)
◆ From: 219.70.175.5
推 Chsieh:如果H0是那樣的話 就把=改成≠不就好了? 12/20 00:58
→ yhliu:可用雙邊對立假說(≠0), 或亦可用單邊對立假說. 12/20 11:00
→ yhliu:就後者而言, "<" 似乎合適, 較少受關照平均出生體重較低. 12/20 11:02
→ yhliu:就習題或考試而言, 題目中應有相關描述可資辨別應採用何者. 12/20 11:03
→ zxc95:我懂了 謝謝 12/21 23:04