MA3251-STATISTICS AND NUMERICAL METHODS

UNIT-1 TESTING OF HYPOTHESIS

FORMULA
Test for 𝐻0 𝐻1 𝜈 Test Statistic
∑𝑥
𝑥̅ − 𝜇 𝑥̅ =
𝜇≠𝑎 𝑛
𝑍= 2
Two tail 𝜎 𝑛 ∑ 𝑥 − (∑ 𝑥)2
or
( ) 𝑠2 =
√𝑛 𝑛2
𝜇>𝑎 If 𝜎 is not known Or
Single Mean 𝜇=𝑎 -
One tail then 2
∑ 𝑥2
or 𝑥̅ − 𝜇 𝑠 = − (𝑥̅ )2
𝑍= 𝑛
𝜇<𝑎 𝑠 Or
( )
One tail √𝑛 2
∑(𝑥 − 𝑥̅ )2
𝑠 =
𝑛
𝑥̅1 − 𝑥̅2 ∑ 𝑥1
𝑍= 𝑥̅1 =
𝑛1
𝜎12 𝜎22 ∑ 2 2
√ 𝑛 𝑥 1 − (∑ 𝑥1 )
𝑛1 + 𝑛2
1
𝑠12 =
𝑛12
If 𝜎1 ≠ 𝜎2 is not Or
known then ∑ 𝑥12
2
𝑥̅1 − 𝑥̅2 𝑠1 = − (𝑥̅1 )2
𝜇1 ≠ 𝜇2 𝑍= 𝑛1
Two tail 𝑠2 𝑠2 Or
√ 1+ 2 ∑(𝑥1 − 𝑥̅1 )2
or 𝑛1 𝑛2 𝑠12 =
Difference of 𝜇1 > 𝜇2 𝑛1
𝜇1 = 𝜇2 - If 𝜎1 = 𝜎2 = 𝜎(𝑆𝑎𝑦)
Mean One tail ∑ 𝑥2
Large Samples

then 𝑥̅2 =
or 𝑥̅1 − 𝑥̅2 𝑛2
𝜇1 < 𝜇2 𝑍= 𝑛2 ∑ 𝑥 2
2 − (∑ 𝑥2 )
2
One tail 1 1 𝑠22 =
𝜎√ + 𝑛22
𝑛1 𝑛2
If 𝜎1 = 𝜎2 is not Or
known then ∑ 𝑥22
𝑥̅1 − 𝑥̅2 𝑠22 = − (𝑥̅2 )2
𝑛2
𝑍=
Or
𝑠2 𝑠2 ∑(𝑥2 − 𝑥̅2 )2
√ 1+ 2
𝑛2 𝑛1 𝑠22 =
𝑛2
𝑃≠𝑎
Two tail
or 𝑝−𝑃
Single 𝑃>𝑎 𝑍= 𝑃 = 1−𝑄
𝑃=𝑎 - 𝑥
Proportion One tail √𝑃𝑄 𝑝=
or 𝑛 𝑛
𝑃<𝑎
One tail
𝑃1 ≠ 𝑃2
Two tail
or 𝑝1 − 𝑝2 𝑛1 𝑝1 + 𝑛2 𝑝2
Difference of 𝑃1 > 𝑃2 𝑍= 𝑃=
𝑃1 = 𝑃2 - 1 1 𝑛1 + 𝑛2
√𝑃𝑄 (
𝑛1 + 𝑛2 )
Proportion One tail
𝑄 = 1−𝑃
or
𝑃1 < 𝑃2
One tail

Prepared by Mr.S.Meeankshi Sundaram, Department of Mathematics

 ∑𝑥
𝑥̅ =
𝑛
2
𝑛 ∑ 𝑥 − (∑ 𝑥)2
𝑠2 =
𝑛2
Or
∑ 𝑥2
𝜇≠𝑎 𝑥̅ − 𝜇 𝑠2 = − (𝑥̅ )2
𝑡= 𝑛
Two tail 𝑆 Or
( )
or √𝑛 ∑(𝑥 − 𝑥̅ )2
If 𝜎 is not known 2
𝜇>𝑎 𝑠 =
Single Mean 𝜇=𝑎 𝑛−1 𝑛
One tail then
or 𝑥̅ − 𝜇
𝑡= 𝑛 ∑ 𝑥 2 − (∑ 𝑥)2
𝜇<𝑎 𝑠 𝑆2 =
( ) 𝑛(𝑛 − 1)
One tail √𝑛 − 1 Or
∑(𝑥 − 𝑥̅ )2
𝑆2 =
𝑛−1
Or
𝑛𝑠 2
𝑆2 =
𝑛−1
∑ 𝑥1
𝑥̅1 =
𝑛1
2
𝑛1 ∑ 𝑥1 − (∑ 𝑥1 )2
𝑠12 =
𝑛12
Small Sample ‘t’ Test

Or
∑ 2
𝑥1
𝑠12 = − (𝑥̅1 )2
𝑛1
𝜇1 ≠ 𝜇2 Or
∑(𝑥 2
Two tail 2 1 − 𝑥̅1 )
𝑠1 =
or 𝑥̅1 − 𝑥̅2 𝑛1
Difference of 𝜇1 > 𝜇2 𝑡= ∑ 𝑥2
𝜇1 = 𝜇2 𝑛1 + 𝑛2 − 2 1 1 𝑥̅2 =
Mean One tail 𝑠√𝑛 + 𝑛 𝑛2
1 2 2
or 𝑛2 ∑ 𝑥2 − (∑ 𝑥2 )2
𝜇1 < 𝜇2 𝑠22 =
𝑛22
One tail Or
∑ 2
𝑥2
𝑠22 = − (𝑥̅2 )2
𝑛2
Or
∑(𝑥 2
2 2 − 𝑥̅2 )
𝑠2 =
𝑛2
2 2
2
𝑛 1 1 + 𝑛2 𝑠2
𝑠
𝑠 =
𝑛1 + 𝑛2 − 2
∑𝑑
𝑑̅ =
𝑛
2
𝑛 ∑ 𝑑 − (∑ 𝑑)2
𝑠2 =
𝑑̅ 𝑛2
𝑡= Or
Paired 𝜇=0 𝜇>0 𝑛−1 𝑠 ∑ 𝑑2
( ) 𝑠2 = − (𝑑̅ )
2
√𝑛 − 1 𝑛
Or
2
2
∑(𝑑 − 𝑑̅ )
𝑠 =
𝑛

Prepared by Mr.S.Meeankshi Sundaram, Department of Mathematics

 𝑛1 ∑ 𝑥1 2 − (∑ 𝑥1 )2
𝑆12 =
𝑛1 (𝑛1 − 1)
Or
∑(𝑥 2
2 1 − 𝑥̅1 )
𝑆1 =
𝑛1 − 1
Small Sample ‘F’ Test

Or
2
𝑛1 𝑠12
𝑆1 =
Equality of 𝜈1 = 𝑛1 − 1 𝑆12 2 𝑛1 − 1
𝜎12 = 𝜎22 𝜎12 ≠ 𝜎22 𝐹 = 2 , 𝑆1 > 𝑆22 2 2
Variance 𝜈2 = 𝑛2 − 1 𝑆2 2
𝑛 2 ∑ 𝑥2 − (∑ 𝑥2 )
𝑆2 =
𝑛2 (𝑛2 − 1)
Or
∑(𝑥 2
2 2 − 𝑥̅2 )
𝑆2 =
𝑛2 − 1
Or
𝑛2 𝑠22
𝑆22 =
𝑛2 − 1
(𝑂𝑖 − 𝐸𝑖 )2
𝑛−1 𝜒2 = ∑
𝐸𝑖
𝑖
𝑎: 𝑏: 𝑐: 𝑑
(𝑂𝑖 − 𝐸𝑖 )2 𝑎
𝑛−1 𝜒2 = ∑ 𝐸𝑖 = 𝑁 ×
𝐸𝑖 𝑎+𝑏+𝑐+𝑑
𝑖
𝑁 = ∑𝑓
(𝑂𝑖 − 𝐸𝑖 )2 Uniform Distribution
𝑛−1 𝜒2 = ∑ 𝑁
𝐸𝑖 𝐸𝑖 =
𝑖 𝑛
Goodness of
Fit Binomial Distribution
𝐸𝑖 = 𝑁 × 𝑛𝑐𝑖 𝑝𝑖 𝑞 𝑛−𝑖
Chi Square

2
(𝑂𝑖 − 𝐸𝑖 )2
𝑛−1 𝜒 =∑ 𝑞 = 1−𝑝
𝐸𝑖
𝑖 𝑥̅ ∑ 𝑓𝑥
𝑝 = , 𝑥̅ =
𝑛 ∑𝑓
Poisson Distribution
(𝑂𝑖 − 𝐸𝑖 )2 𝑒 −𝜆 𝜆𝑖
𝜒2 = ∑ 𝐸𝑖 = 𝑁 ×
𝑛−2 𝐸𝑖 𝑖!
𝑖 ∑ 𝑓𝑥
𝜆 = 𝑥̅ , ̅̅̅
𝑥=
∑𝑓
(𝑂𝑖 − 𝐸𝑖 )2 𝐸𝑖 =
(𝑚 − 1)(𝑛 − 1) 𝜒2 = ∑ (𝑟𝑜𝑤 𝑡𝑜𝑡𝑎𝑙)×(𝑐𝑜𝑙𝑢𝑚𝑛 𝑡𝑜𝑡𝑎𝑙)
𝐸𝑖 𝑔𝑟𝑎𝑛𝑑 𝑡𝑜𝑡𝑎𝑙
Independence 𝑖

of Attributes 𝜒2 = 2×2 a b
1 𝑁 (𝑎𝑑−𝑏𝑐)2 c d
(𝑎+𝑏)(𝑎+𝑐)(𝑑+𝑏)(𝑑+𝑐)
𝑁 =𝑎+𝑏+𝑐+𝑑

Prepared by Mr.S.Meeankshi Sundaram, Department of Mathematics