Hypothesis Testing for a Single Sample Cheatsheet

Alternative OC Curves
Hypothesis Test Null Hypothesis Test Statistic Rejection Region
Hypothesis Parameter
Hypothesis Test on 𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑖𝑓𝑧0 < −𝑧𝛼 𝑜𝑟 𝑧0 > 𝑧𝛼
𝐻1 : 𝜇 ≠ 𝜇0 𝑋̅ − 𝜇0 |𝜇 − 𝜇0 |
Mean (Case I: Normal 𝑍0 = 2 2
𝐻0 : 𝜇 = 𝜇0 𝐻1 : 𝜇 > 𝜇0 𝜎 𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑖𝑓 𝑧0 > 𝑧𝛼 𝑑=
Distribution with 𝜎
𝐻1 : 𝜇 < 𝜇0 √𝑛 𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑖𝑓 𝑧0 < −𝑧𝛼
Known Variance)
Hypothesis Test on
Mean (Case II: 𝐻1 : 𝜇 ≠ 𝜇0 𝑋̅ − 𝜇0 𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑖𝑓𝑧0 < −𝑧𝛼 𝑜𝑟 𝑧0 > 𝑧𝛼
𝑍0 = 2 2 |𝜇 − 𝜇0 |
Unknown Distribution, 𝐻0 : 𝜇 = 𝜇0 𝐻1 : 𝜇 > 𝜇0 𝑠 𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑖𝑓 𝑧0 > 𝑧𝛼 𝑑=
𝐻1 : 𝜇 < 𝜇0 𝑠
Variance Unknown, √𝑛 𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑖𝑓 𝑧0 < −𝑧𝛼
Large Sample Size)
Hypothesis Test on
Mean (Case III: 𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑖𝑓 𝑡0 < −𝑡𝛼,𝑛−1 𝑜𝑟 𝑡0 > 𝑡𝛼,𝑛−1
𝐻1 : 𝜇 ≠ 𝜇0 𝑋̅ − 𝜇0 2 2 |𝜇 − 𝜇0 |
Normal Distribution 𝑇0 =
𝐻0 : 𝜇 = 𝜇0 𝐻1 : 𝜇 > 𝜇0 𝑠 𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑖𝑓 𝑡0 > 𝑡𝛼,𝑛−1 𝑑=
with Unknown
𝐻1 : 𝜇 < 𝜇0 𝑠
Variance, Small √𝑛 𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑖𝑓 𝑡0 < −𝑡𝛼,𝑛−1
Sample Size)
𝐻1 : 𝜎 2 ≠ 𝜎0 2 𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑖𝑓 𝜒0 2 < 𝜒 21−𝛼,𝑛−1 𝑜𝑟 𝜒0 2 > 𝜒 2 𝛼,𝑛−1
Hypothesis Test on (𝑛 − 1)𝑠 2 2 2 𝜎
𝐻0 : 𝜎 2 = 𝜎0 2 𝐻1 : 𝜎 2 > 𝜎0 2 𝜒0 2 = 𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑖𝑓𝜒0 2 > 𝜒 2 𝛼,𝑛−1 𝜆=
Variance 𝜎0 2 𝜎0
𝐻1 : 𝜎 2 < 𝜎0 2 𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑖𝑓𝜒0 2 < 𝜒 21−𝛼,𝑛−1
𝐻1 : 𝑝 ≠ 𝑝0 𝑝̂ − 𝑝0 𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑖𝑓𝑧0 < −𝑧𝛼 𝑜𝑟 𝑧0 > 𝑧𝛼 |𝑝 − 𝑝0 |
Hypothesis Test on 𝑍0 = 2 2 𝑑=
𝐻0 : 𝑝 = 𝑝0 𝐻1 : 𝑝 > 𝑝0 𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑖𝑓 𝑧0 > 𝑧𝛼
Proportion
𝐻1 : 𝑝 < 𝑝0 √𝑝0 (1 − 𝑝0 ) √𝑝(1 − 𝑝)
𝑛 𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑖𝑓 𝑧0 < −𝑧𝛼 𝑛
𝑘
𝐻0 : 𝑑𝑖𝑠𝑡𝑟𝑖𝑏𝑢𝑡𝑖𝑜𝑛 𝐻1 : 𝑑𝑖𝑠𝑡𝑟𝑖𝑏𝑢𝑡𝑖𝑜𝑛 𝑖𝑠 (𝑂𝑖 − 𝐸𝑖 )2
Goodness of Fit Test 2
𝜒0 = ∑ 𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑖𝑓 𝜒0 2 > 𝜒 2 𝛼,𝑘−𝑝−1
𝑖𝑠 ~𝑑𝑖𝑠𝑡𝑟𝑖𝑏𝑢𝑡𝑖𝑜𝑛~ 𝑛𝑜𝑡 ~𝑑𝑖𝑠𝑡𝑟𝑖𝑏𝑢𝑡𝑖𝑜𝑛~ 𝐸𝑖
𝑖=1
𝑟 𝑐 2
(𝑂𝑖𝑗 − 𝐸𝑖𝑗 )
Contingency Table 𝐻0 : 𝑟𝑜𝑤 𝑎𝑛𝑑 𝑐𝑜𝑙𝑢𝑚𝑛 𝐻1 : 𝑟𝑜𝑤 𝑎𝑛𝑑 𝑐𝑜𝑙𝑢𝑚𝑛 𝜒0 2 = ∑ ∑
𝐸𝑖𝑗 𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑖𝑓 𝜒0 2 > 𝜒 2 𝛼, (𝑟−1)(𝑐−1)
Test 𝑎𝑟𝑒 𝑖𝑛𝑑𝑒𝑝𝑒𝑛𝑑𝑒𝑛𝑡 𝑎𝑟𝑒 𝑛𝑜𝑡 𝑖𝑛𝑑𝑒𝑝𝑒𝑛𝑑𝑒𝑛𝑡 𝑖=1 𝑗=1
𝐸𝑖𝑗 = 𝑛𝑢̂𝑖 𝑣̂𝑗
 Hypothesis Testing for Two Samples Cheatsheet

Alternative OC Curves
Hypothesis Test Null Hypothesis Test Statistic Rejection Region
Hypothesis Parameter
Hypothesis Test on
Difference of Means 𝑋̅1 − 𝑋̅2 − ∆0 𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑖𝑓𝑧0 < −𝑧𝛼 𝑜𝑟 𝑧0 > 𝑧𝛼 |∆ − ∆0 |
𝐻1 : 𝜇1 − 𝜇2 ≠ ∆0 𝑑=
(Case I: Normal 𝑍0 = 2 2
𝐻0 : 𝜇1 − 𝜇2 = ∆0 𝐻1 : 𝜇1 − 𝜇2 > ∆0 𝜎1 2 𝜎2 2 𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑖𝑓 𝑧0 > 𝑧𝛼 √𝜎1 2 + 𝜎2 2
Distribution, √ +
𝐻1 : 𝜇1 − 𝜇2 < ∆0 𝑛1 𝑛2 𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑖𝑓 𝑧0 < −𝑧𝛼 n = n1 = n2
Population Variances
Known or Estimated)
Hypothesis Test on 𝑋̅1 − 𝑋̅2 − ∆0
𝑇0 = 𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑖𝑓𝑡0 < −𝑡𝛼,𝑣 𝑜𝑟 𝑡0 > 𝑡𝛼,𝑣
Difference of Means 1 1 |∆ − ∆0 |
𝐻1 : 𝜇1 − 𝜇2 ≠ ∆0 𝑆𝑝 √ + 2 2
(Case II: Normal 𝑛1 𝑛2 𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑖𝑓 𝑡0 > 𝑡𝛼,𝑣 𝑑=
𝐻0 : 𝜇1 − 𝜇2 = ∆0 𝐻1 : 𝜇1 − 𝜇2 > ∆0 2𝑆𝑝
Distribution, 𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑖𝑓 𝑡0 < −𝑡𝛼,𝑣
𝐻1 : 𝜇1 − 𝜇2 < ∆0 (𝑛1 − 1)𝑠1 2 + (𝑛2 − 1)𝑠2 2 n* = n1 + n2 – 1
Population Variances 𝑆𝑝 = √ 𝑣 = 𝑛1 + 𝑛2 − 2
Unknown and Equal) 𝑛1 + 𝑛2 − 2
𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑖𝑓𝑡0 < −𝑡𝛼,𝑣 𝑜𝑟 𝑡0 > 𝑡𝛼,𝑣
Hypothesis Test on 2 2
Difference of Means 𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑖𝑓 𝑡0 > 𝑡𝛼,𝑣
(Case III: Normal 𝐻1 : 𝜇1 − 𝜇2 ≠ ∆0 𝑋̅1 − 𝑋̅2 − ∆0 𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑖𝑓 𝑡0 < −𝑡𝛼,𝑣
𝑇0 = 𝑠 2 𝑠 2
2
Distribution, 𝐻0 : 𝜇1 − 𝜇2 = ∆0 𝐻1 : 𝜇1 − 𝜇2 > ∆0 𝑠 2 𝑠 2 ( 1 + 2 )
𝐻1 : 𝜇1 − 𝜇2 < ∆0 √ 1 + 2 𝑛1 𝑛2
Population Variances 𝑛1 𝑛2 𝑣=
2 2 2
Unknown and 𝑠 𝑠 2
( 1 ) ( 2 )
Unequal) 𝑛1 𝑛
+ 2
𝑛1 − 1 𝑛2 − 1
𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑖𝑓𝑡0 < −𝑡𝛼,𝑣 𝑜𝑟 𝑡0 > 𝑡𝛼,𝑣
Hypothesis Test on 𝐻1 : 𝜇𝐷 ≠ ∆0 ̅ − ∆0
𝐷 2 2
𝐻0 : 𝜇𝐷 = ∆0 𝐻1 : 𝜇𝐷 > ∆0 𝑇0 = 𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑖𝑓 𝑡0 > 𝑡𝛼,𝑣
Difference of Means 𝑆𝐷
𝐻1 : 𝜇𝐷 < ∆0 𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑖𝑓 𝑡0 < −𝑡𝛼,𝑣
(Paired T-Test) √𝑛
𝑣 =𝑛−1
𝜎1 2
𝐻1 : ≠ 𝜆0
𝜎2 2 𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑖𝑓𝑓0 < 𝑓1−𝛼,𝑛 𝑜𝑟𝑓0 > 𝑓𝛼,𝑛 𝜎1
Hypothesis Test on −1,𝑛2 −1 −1,𝑛2 −1
𝜎1 2 𝜎1 2
𝑆1 2 2 1 2 1 𝜆=
Ratio of Two 𝐻0 : = 𝜆0 𝐻1 : > 𝜆0 𝐹0 = 𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑖𝑓𝑓0 > 𝑓𝛼,𝑛1−1,𝑛2−1 𝜎2
Variances 𝜎2 2 𝜎2 2 𝑆2 2 n = n1 = n2
𝜎1 2 𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑖𝑓𝑓0 < 𝑓1−𝛼,𝑛1−1,𝑛2−1
𝐻1 : < 𝜆0
𝜎2 2
𝑝̂1 − 𝑝̂2 − ∆0
𝑍0 = 𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑖𝑓𝑧0 < −𝑧𝛼 𝑜𝑟 𝑧0 > 𝑧𝛼
Hypothesis Test on 𝐻1 : 𝑝1 − 𝑝2 ≠ ∆0 1 1
𝐻0 : 𝑝1 − 𝑝2 = ∆0 𝐻1 : 𝑝1 − 𝑝2 > ∆0 √𝑝̂ (1 − 𝑝̂ ) (+ ) 2 2
Difference of 𝑛1 𝑛2 𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑖𝑓 𝑧0 > 𝑧𝛼
Proportions 𝐻1 : 𝑝1 − 𝑝2 < ∆0 𝑥1 + 𝑥2 𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑖𝑓 𝑧0 < −𝑧𝛼
𝑝̂ =
𝑛1 + 𝑛2