Statistics for management and Economics, Seventh Edition

Formulas
Numerical Descriptive techniques
Population mean

=
N
x
N
i
i
1
Sample mean
n
x
x
n
i
i

1
Range
Largest observation - Smallest observation
Population variance
2
=
N
) x (
N
i
i

1
2

Sample variance
2
s =
1
) (
1
2

n
x x
n
i
i
Population standard deviation

=
2

Sample standard deviation

s =
2
s
Population covariance
N
) y )( x (
N
i
y i x i
xy

Sample covariance
1
1

n
) y y )( x x (
s
n
i
i i
xy
Population coefficient of correlation
y x
xy

Sample coefficient of correlation
y x
xy
s s
s
r
Slope coefficient
2
1
x
s
) y , x cov(
b
y-intercept
x b y b
1 0

Probability
Conditional probability
P(A|B) = P(A and B)/P(B)
Complement rule
P(
C
A
) = 1 P(A)
Multiplication rule
P(A and B) = P(A|B)P(B)
Addition rule
P(A or B) = P(A) + P(B) - P(A and B)
Bayes Law Formula
) A | B ( P ) A ( P . . . ) A | B ( P ) A ( P ) A | B ( P ) A ( P
) A | B ( P ) A ( P
) B | A ( P
k k 2 2 1 1
i i
i
+ + +

Random Variables and Discrete Probability Distributions

Expected value (mean)
E(X) =

x all
) x ( xP
Variance
V(x) =


x all
) x ( P ) x (
2 2

Standard deviation
2

Covariance
COV(X, Y) =

) y , x ( P ) y )( x (
y x

Coefficient of Correlation
y x
) Y , X ( COV



Laws of expected value
1. E(c) = c
2. E(X + c) = E(X) + c
3. E(cX) = cE(X)
Laws of variance
1.V(c) = 0
2. V(X + c) = V(X)
3. V(cX) =
2
c
V(X)
Laws of expected value and variance of the sum of two variables
1. E(X + Y) = E(X) + E(Y)
2. V(X + Y) = V(X) + V(Y) + 2COV(X, Y)
Laws of expected value and variance for the sum of more than two variables
1.

k
i
i
k
i
i
X E X E
1 1
) ( ) (
2.

k
i
i
k
i
i
X V X V
1 1
) ( ) ( if the variables are independent
Mean and variance of a portfolio of two stocks
E(Rp) = w1E(R1) + w2E(R2)
V(Rp) =
2
1
w V(R1) +
2
2
w V(R2) + 2
1
w
2
w COV(R1, R2)
=
2
1
w
2
1
+
2
2
w
2
2
+ 2
1
w
2
w

1

Mean and variance of a portfolio of k stocks

E(Rp) =

k
i
i i
R E w
1
) (
V(Rp) =

+
+
k
i
k
i j
j i j i
k
i
i i
R R COV w w w
1 1 1
2 2
) , ( 2
Binomial probability
P(X = x) =
)! x n ( ! x
! n

x n x
) p ( p

1
np
) p ( np 1
2

) p ( np 1
Poisson probability
P(X = x) =
! x
e
x

Continuous Probability Distributions
Standard normal random variable

X
Z
Exponential distribution
/ 1
x
e ) x X ( P

>
x
e 1 ) x X ( P

<
2 1
x x
1 2 2 1
e e ) x X ( P ) x X ( P ) x X x ( P

< < < <
F distribution
2 1
, , 1 A
F

=
1 2
, ,
1
A
F
Sampling Distributions
Expected value of the sample mean
) (X E
x
Variance of the sample mean
) (X V
n
x
2
2


Standard error of the sample mean
n
x


Standardizing the sample mean
n
X
Z
/

Expected value of the sample proportion

)

(P E
p
p

Variance of the sample proportion

n
p p
P V
p
) 1 (
)

(
2

Standard error of the sample proportion
n
p p
p
) 1 (

Standardizing the sample proportion
n p p
p P
Z
) 1 (

Expected value of the difference between two means

) (
2 1
X X E 2 1
2 1

x x
Variance of the difference between two means
2
2
2
1
2
1 2
2 1
2 1
) (
n n
X X V
x x

+

Standard error of the difference between two means
2
2
2
1
2
1
2 1
n n
x x

+

Standardizing the difference between two sample means
2
2
2
1
2
1
2 1 2 1
) ( ) (
n n
X X
Z

Introduction to Estimation
Confidence interval estimator of

n
z x
/

2
t
Sample size to estimate

2
2

,
_

W
z
n
/

Introduction to Hypothesis Testing

Test statistic for

n /
x
z

Inference about One Population

Test statistic for

n / s
x
t

Confidence interval estimator of

n
s
t x
/ 2
t
Test statistic for
2

2
2
2
1

s ) n (

Confidence interval Estimator of

2

LCL =
2
2
2
1
/
s ) n (

UCL =
2
2 1
2
1
/
s ) n (

Test statistic for p

n / ) p ( p
p p
z

1
Confidence interval estimator of p
n / ) p ( p z p
/
t 1
2
Sample size to estimate p
2
2
1

,
_

W
) p ( p z
n
/
Confidence interval estimator of the total of a large finite population
1
]
1

t
n
s
t x N
/ 2
Confidence interval estimator of the total number of successes in a large finite population
1
1
]
1

t
n
p ( p
z p N
/
1
2
Confidence interval estimator of

when the population is small

1
2

t
N
n N
n
s
t x
/
Confidence interval estimator of the total in a small population

,
_

t
1
2
N
n N
n
s
t x N
/
Confidence interval estimator of p when the population is small
1
1
2

t
N
n N
n
) p ( p
z p
/

Confidence interval estimator of the total number of successes in a small population

,
_

t
1
1
2
N
n N
n
) p ( p
z p N
/
Inference About Two Populations
Equal-variances t-test of
2 1

,
_

2 1
2
2 1 2 1
1 1
n n
s
) ( ) x x (
t
p

2
2 1
+ n n
Equal-variances interval estimator of
2 1

,
_

+ t
2 1
2
2 2 1
1 1
n n
s t ) x x (
p /
2
2 1
+ n n
Unequal-variances t-test of
2 1

,
_

2
2
2
1
2
1
2 1 2 1
n
s
n
s
) ( ) x x (
t


1 1
2
2
2
2
2
1
2
1
2
1
2
2
2
2 1
2
1

n
) n / s (
n
) n / s (
) n / s n / s (

Unequal-variances interval estimator of

2 1

2
2
2
1
2
1
2 2 1
n
s
n
s
t ) x x (
/
+ t

1 1
2
2
2
2
2
1
2
1
2
1
2
2
2
2 1
2
1

n
) n / s (
n
) n / s (
) n / s n / s (

t-Test of
D

D D
D D
n / s
x
t

1
D
n
t-Estimator of
D

D
D
/ D
n
s
t x
2
t
1
D
n
F-test of
2
2
2
1
/
F =
2
2
2
1
s
s
1
1 1
n and 1
2 2
n
F-Estimator of
2
2
2
1
/
LCL =
2 1
2
2
2
2
1
1
, , /
F
s
s

,
_

UCL =
1 2
2
2
2
2
1
, , /
F
s
s

,
_

z-Test and estimator of

2 1
p p
Case 1:

,
_

2 1
2 1
1 1
1
n n
) p ( p
) p p (
z
Case 2:
2
2 2
1
1 1
2 1 2 1
1 1
n
) p ( p
n
) p ( p
) p p ( ) p p (
z

z-estimator of
2 1
p p
2
2 2
1
1 1
2 2 1
1 1
n
) p ( p
n
) p ( p
z ) p p (
/

+

t

Analysis of Variance
One-way analysis of variance
SST =

k
j
j j
) x x ( n
1
2
SSE =

j
n
i
j ij
k
j
) x x (
1
2
1
MST =
1 k
SST

MSE =
k n
SSE

F =
MSE
MST
Two-way analysis of variance (randomized block design of experiment)
SS(Total) =

b
i
ij
k
j
) x x (
1
2
1
SST =

k
i
j
) x ] T [ x ( b
1
2
SSB =

b
i
i
) x ] B [ x ( k
1
2
SSE =


+
b
i
i j ij
k
j
) x ] B [ x ] T [ x x (
1
2
1
MST =
1 k
SST

MSB =
1 b
SSB

MSE =
1 b k n
SSE
+
F =
MSE
MST
F=
MSE
MSB
Two-factor experiment
SS(Total) =

a
i
b
j
r
k
ijk
) x x (
1 1 1
2
SS(A) =

a
i
i
) x ] A [ x ( rb
1
2
SS(B) =

b
j
j
) x ] B [ x ( ra
1
2
SS(AB) =


+
a
i
b
j
j i ij
) x ] B [ x ] A [ x ] AB [ x ( r
1 1
2
SSE =

a
i
b
j
r
k
ij ijk
) ] AB [ x x (
1 1 1
2
F =
MSE
) A ( MS
F =
MSE
) B ( MS
F =
MSE
) AB ( MS
Least Significant Difference Comparison Method
LSD =

,
_

+
j i
/
n n
MSE t
1 1
2
Tukeys multiple comparison method
g
n
MSE
) , k ( q

Chi-Squared Tests
Test statistic for all procedures

k
1 i
i
2
i i 2
e
) e f (
Simple Linear Regression
Sample slope
2
1
x
xy
s
s
b
Sample y-intercept
x b y b
1 0

Sum of squares for error
SSE =

n
i
i i
) y y (
1
2
Standard error of estimate
2

n
SSE
s

Test statistic for the slope

1
1 1
b
s
b
t

Standard error of
1
b
2
1
1
x
b
s ) n (
s
s

Coefficient of determination
2 2
2
2
y x
xy
s s
s
R

2
1
) y y (
SSE
i
Prediction interval
2
2
2 2
1
1
1
x
g
n , /
s ) n (
) x x (
n
s t y

+ + t

Confidence interval estimator of the expected value of y
2
2
2 2
1
1
x
g
n , /
s ) n (
) x x (
n
s t y

+ t

Sample coefficient of correlation
y x
xy
s s
s
r
Test statistic for testing

= 0
2
1
2
r
n
r t

Multiple Regression
Standard Error of Estimate
1

k n
SSE
s

Test statistic for

i

i
b
i i
s
b
t

Coefficient of Determination
2 2
2
2
y x
xy
s s
s
R

2
1
) y y (
SSE
i
Adjusted Coefficient of Determination
Adjusted

) n /( ) y y (
) k n /( SSE
R
i
1
1
1
2
2
Mean Square for Error
MSE = SSE/k
Mean Square for Regression
MSR = SSR/(n-k-1)
F-statistic
F = MSR/MSE
Durbin-Watson statistic

n
i
i
n
i
i i
e
) e e (
d
1
2
2
2
1
Time Series Analysis and Forecasting
Exponential smoothing
1
) 1 (

+
t t t
S w wy S
Nonparametric Statistical techniques
Wilcoxon rank sum test statistic
1
T T
E(T) =
2
1
2 1 1
) n n ( n + +
12
1
2 1 2 1
) n n ( n n
T
+ +

T
) T ( E T
z

Sign test statistic

x = number of positive differences
n .
n . x
z
5
5

Wilcoxon signed rank sum test statistic

+
T T
E(T) =
4
1) n ( n +
24
1 2 1 ) n )( n ( n
T
+ +

T
) T ( E T
z

Kruskal-Wallis Test
) n (
n
T
) n ( n
H
k
j
j
j
1 3
1
12
1
2
+
1
1
]
1

Friedman Test
) k ( b T
) k )( k ( b
F
k
j
j r
1 3
1
12
1
2
+
1
1
]
1

Spearman rank correlation coefficient

b a
ab
S
s s
s
r
Spearman Test statistic for n > 30
1 n r z
S
Statistical Process Control
Centerline and control limits for x chart using S
Centerline = x
Lower control limit =
n
S
x 3
Upper control limit =
n
S
x 3 +
Centerline and control limits for the p chart
Centerline =
p
Lower control limit =
n
) p ( p
p

1
3
Upper control limit =
n
) p ( p
p

+
1
3
Decision Analysis
Expected Value of perfect Information
EVPI = EPPI - EMV*
Expected Value of Sample Information
EVSI = EMV' - EMV*