T检验与Z检验的区别

概念区别：T检验，亦称student t检验（Student's t test），主要用于样本含量较小（例如n> x=[497,506,518,524,488,517,510,515,516];

>> [h,sig,ci,zval]=ztest(x,500,15,0.01,0)

h =     0                          %接受

sig =    0.0432                         % 为真条件下P( )的值

ci =  497.2320  522.9903         % 未知时 的置信水平为0.95的双侧置信区间

zval =    2.0222                %Z统计量 的值.

所以认为包装机工作正常.

【例3-2】(例3.5) 某部门对当前市场的价格情况进行调查. 以鸡蛋为例, 所抽查的全省20个集市上, 售价分别为 (单位: 元/500克)

3.05,    3.31,    3.34,    3.82,    3.30,    3.16,    3.84,    3.10,    3.90,    3.18,

3.88,    3.22,    3.28,    3.34,    3.62,    3.28,    3.30,    3.22,    3.54,    3.30.

已知往年的平均售价一直稳定在3.25元/500克左右, 在显著性水平 下, 能否认为全省当前的鸡蛋售价明显高于往年?

>> x=[3.05,3.31,3.34,3.82,3.30,3.16,3.84,3.10,3.90,3.18,...

3.88,3.22,3.28,3.34,3.62,3.28,3.30,3.22,3.54,3.30];

>> [h,sig,ci,tval]=ttest(x,3.25,0.025,1)

h =     1

sig =    0.0114

ci =    3.2731       Inf

tval =     tstat: 2.4763       df: 19

所以认为全省当前的鸡蛋售价明显高于往年.