Table of Hypothesis Tests (Chapters 9-14)

Ch. Sample Standard deviation Needed observations value of test Critical value calculation Two-tail Other required formula
& 𝝈𝟏 , 𝝈𝟐 statistics or statistics hypothesis
type parameters
𝜎, 𝑥̅ , 𝜇, 𝑛 𝑥̅ − 𝜇
𝑧= t-table 𝐻 : 𝜇=𝜇
Mean 𝜎 known 𝜎/√𝑛
α, 𝑑𝑓 = ∞
𝐻 : 𝜇≠𝜇
One-sample
𝜎 unknown 𝑠, 𝑥̅ , 𝜇, 𝑛 𝑥̅ − 𝜇
𝑡= t-table 𝐻 : 𝜇=𝜇
9 𝑠/√𝑛
α, 𝑑𝑓 = 𝑛 − 1
𝐻 : 𝜇≠𝜇
𝜎 known 𝜎 , 𝑝, 𝑝̅ 𝑝̅ − 𝑝 𝑝̅ − 𝑝
𝑧= ,𝑧 = t-table 𝐻 : 𝑝=𝑝
Proportion 𝜎 𝑝(1 − 𝑝)
𝑛 α, 𝑑𝑓 = ∞
𝐻 : 𝑝≠𝑝
One-sample

𝜎 known 𝑥̅ − 𝑥̅ − (𝜇 − 𝜇 )
𝜇 =𝜇 𝑧= t-table 𝐻 :𝜇 =𝜇
Independent 𝜎 𝜎
+ α, 𝑑𝑓 = ∞
𝑛 𝑛 𝐻 :𝜇 ≠𝜇
Means 𝜎 unknown 𝑥̅ − 𝑥̅ − (𝜇 − 𝜇 ) ( ) ( )
𝜇 =𝜇 𝑡= t-table 𝐻 :𝜇 =𝜇 𝑠 = ,
1 1
Two-sample 𝑠
𝑛
+
𝑛 (∑ )
∑
α, 𝑑𝑓 = 𝑛 + 𝑛 − 2 𝐻 :𝜇 ≠𝜇 𝑠 = ,

10 (∑ 𝑥 )
∑𝑥 −
𝑛
𝑠 =
𝑛 −1
Dependent 𝜎 unknown 𝑑̅ − 𝜇 𝑑 = 𝑥 − 𝑥 , 𝑑̅ =
∑
,
𝜇 =0 𝑡= t-table 𝐻 :𝜇 =0
means 𝑠 /√𝑛
Two-sample ∑
(∑ )
α, 𝑑𝑓 = 𝑛– 1 𝐻 :𝜇 ≠0 𝑠 =

Independent 𝜎 known 𝑧 𝑝̅ = , 𝑝̅ = ,
𝑝 =𝑝 t-table 𝐻 :𝑝 =𝑝
Proportion 𝑝̅ − 𝑝̅ − (𝑝 − 𝑝 ) 𝑥 +𝑥
=
Two-sample 𝑝 (1 − 𝑝 ) 𝑝 (1 − 𝑝 ) 𝑝 =
+ α, 𝑑𝑓 = ∞ 𝐻 :𝑝 ≠𝑝 𝑛 +𝑛
𝑛 𝑛
 𝝈𝟏 , 𝝈𝟐 𝝈𝟏 = 𝝈𝟐 𝑠
𝐹= , 𝑠 >𝑠 F table 𝐻 : 𝜎 =𝜎 , ∑
(∑ )
equal 𝑠 𝑠 = ,𝑠 =
variances
df_numerator=n1 – 1 𝐻 : 𝜎 ≠𝜎 (∑ )
∑
Two-sample df_denominator= n2 – 1

11 Means, 𝑀𝑆𝑇 MST = SST/(k − 1)

𝜇 =𝜇 𝐹= F table 𝐻 :𝜇 =𝜇 =𝜇
𝑀𝑆𝐸 MSE = SSE/(n − k)
Three or =𝜇 =⋯
df_numerator=k– 1 𝐻 : 𝑛𝑜𝑡 𝑎𝑙𝑙 𝑚𝑒𝑎𝑛𝑠 𝑆𝑆 𝑡𝑜𝑡𝑎𝑙 = (𝑥 − 𝑥̅ ) ,
more samples
df_denominator=n-k 𝑆𝑆𝐸 = ∑(𝑥 − 𝑥̅ ) ,
𝑎𝑟𝑒 𝑒𝑞𝑢𝑎𝑙 𝑆𝑆𝑇 = 𝑆𝑆 𝑡𝑜𝑡𝑎𝑙 − 𝑆𝐸𝐸

12 Test df Population 𝑟√𝑛 − 2

𝑡= t-table 𝐻0: 
correlation r correlation ∑(𝑥 − 𝑥̅ )(𝑦 − 𝑦)
√1 − 𝑟 𝑟=
coefficient  is zero. (𝑛 − 1)𝑠 𝑠
α, 𝑑𝑓 = 𝑛– 1 𝐻1:  ≠ 

13 Global test: the 𝑆𝑆𝑅/𝑘

𝐹= 𝐹 distribution table
multiple 𝑆𝑆𝐸/[𝑛 − 𝑘 − 1] 𝐻0: 𝛽1 = 𝛽2 =
regression ⋯ = 𝛽𝑘 = 0
df_numerator=k 𝐻1: Not all 𝛽’𝑠 are 0
model
df_denominator=n-k-1
Individual tes Evaluating 𝑏 −0 𝑏 from the multiple
𝑡= t-table
Individual 𝑆 𝐻0: 𝛽𝑖 = 0 regression coefficient;
Regression 𝐻1: 𝛽𝑖 ≠ 0 𝑆 from standard error of
α, 𝑑𝑓 = 𝑛 – (𝑘 + 1)
Coefficients regression

14 Goodness-of- Chi-square 𝑓, 𝑓
B.6 Chi-Square table
𝐻0 : there is no
𝑓𝑜 is an observed
fit test 𝜒 (𝑓 − 𝑓 ) frequency in a particular
𝜒 = difference between
𝑓 Degree of freedom = 𝑘– 1 category.
𝑓𝑜 and 𝑓𝑒.
𝑘 is the number of 𝑓𝑒 is an expected
categories. 𝐻1 : there is a frequency in a particular
difference between category.
𝑓𝑜 and 𝑓𝑒.