Quiz 4 Formula Sheet

Legend X−μ0
t n−1=
∑: Sum; ∏: Product; s
ϴ: population parameter;
√n
For large samples (n > 30), sometime a z-
µ: population mean; test is used in place of a t-test:
2
σ ( X ) : Population (ex ante) variance of X X−μ 0
z= , if variance is known
σ
σX: Population (ex ante) standard deviation of X √n
σXY: population (ex ante) covariance of X & Y X−μ 0
z= , if variance is unknown
s
ρXY: population (ex ante) correlation coeff. of X & Y √n
Formulas
- Hypothesis statements - Tests of the Differences between Means:
independent samples
H0: θ = θ0 versus Ha: θ ≠ θ0 (a “not equal to” If populations are normally distributed with
alternative hypothesis) identical variance, then we use the pooled
H0: θ ≤ θ0 versus Ha: θ > θ0 (a “greater than” estimate of variance, s2p:
alternative hypothesis) ( X 1− X 2 )−( μ1 −μ 2 )
t=
H0: θ ≥ θ0 versus Ha: θ < θ0 (a “less than”
( )
2 2 1/ 2
sp s p
alternative hypothesis) +
n1 n2
- ( n1−1 ) s 21+ ( n2−1 ) s22
2
Sample statistic−Population parameter under H 0 s=
p
Test statistic= n1 +n2−2
Standard error of the sample statistic
The degrees of freedom is n1 +n 2−2
- Critical Value (Rejection Point): For large If populations are normally distributed with
sample with known 𝝈2: unequal and unknown variances ( s2i ):
For a one-tailed test, zα ( X 1− X 2 )−( μ1 −μ 2 )
t=
For a two-tailed test, zα/2.
( )
2 2 1/ 2
s 1 s2
+
Critical Value: otherwise n1 n2
For a one-tailed test, it is denoted by tα,(n-1)
( )
2 2 2
s1 s2
+
For a two-tailed test, tα/2,(n-1). n1 n 2
df =

( ) ( )
2 2 2 2
s1 s2
- The p-value test n1 n2
For a one-tailed test, reject if p < α. +
n1 n2
For a two-tailed test, reject if p < α. - Tests of the Mean Differences: dependent
samples, note that ∑ is the sum
- Tests Concerning a Single Mean
When the sample is large or when the ,
sample is small but the population is d−μd 0
t=
normal, the test statistic for μ is: sd
Quiz 4 Formula Sheet
Test stat is t-distributed with df = n-1

- Test of a single variance:

H0: σ 2=σ 20 versus HA:σ 2 ≠ σ 20
H0: σ 2 ≤ σ 20 versus HA:σ 2> σ 20
H0: σ 2 ≥ σ 20 versus HA: σ 2< σ 20
chi-square test statistic with df = n-1
2 ( n−1 ) s 2
χ=
σ 20
- Tests of the Equality of Two Variances
H0: σ 21=σ 22 versus HA:σ 21 ≠ σ 22
H0:σ 21 ≤ σ 22 versus HA: σ 21> σ 22
H0: σ 21 ≥ σ 22 versus HA:σ 21< σ 22
F test statistic with df1 = n1-1 and df2 = n2-1,
2
s1
F= 2 , with s21 ≧ s22
s2
- Correlation Analysis:
Hypothesis Testing: H0: 𝜌xy = 0, HA: 𝜌xy ≠ 0.
r √ n−2
The test statistic t= , df = n-2
√ 1−r 2