HARARE INSTITUTE OF TECHNOLOGY

Applied Statistics: ISE1204/ICS2205

FORMULA SHEET
This formula sheet is ALLOWED in the ISE1204/ICS2205 Examination

1. Descriptive Statistics, measures of location and dispersion

Pn
2 i=1 x2i − nx̄2 s
Sample Variance: s = , CV =
n−1 |x̄|
1
 
q1 = x(⌊ n+1 ⌋) + x n+1 − x(⌊ n+1 ⌋)
4 4 (⌊ 4 ⌋+1) 4

3
 
q3 = x(⌊ 3(n+1) ⌋) + x 3(n+1) − x(⌊ 3(n+1) ⌋)
4 4 (⌊ 4 ⌋+1) 4

Measures of location and dispersion for grouped Data:

n
!
2
− CF
Median = L + h
f
!
f1 − f0
Mode = L + h
2f1 − f0 − f2
Pn
x2i fi − ( ni=1 xi fi )2
P
n i=1
Variance =
n(n − 1)
Pk
i=1 fi |xi − x̄|
Mean absolute deviation =
n
2. Probability Theory
P (A ∩ B)
P (A ∪ B) = P (A) + P (B) − P (A ∩ B) , P (A|B) =
P (B)
n
X
Law of Total Probability: P (B) = P (B|Ai )P (Ai ), i = 1, 2, 3, . . . , n
i=1

P (B|Aj )P (Aj )
Baye’s Theorem: P (Aj |B) = Pn , for any 1 ≤ j ≤ n.
i=1 P (B|Ai )P (Ai )

e−λ λx
Poisson pdf: fX (x, λ) = , x = 0, 1, 2, . . .
x!
!
n
Binomial pdf: fX (x, n, θ) = θx (1 − θ)n−x , x = 0, 1, 2, . . . n
x

Geometric pdf: P (X = k) = (1 − θ)k−1 θ, k = 1, 2, 3, . . .

Exponential pdf: fX (x, λ) = λe−λx , x > 0

1
3. Sampling Distributions
Inferences concerning µ (single population)
x̄ − µ0 x̄ − µ0
Test Statistics for H0 : µ = µ0 vs H1 : µ ̸= µ0 : Z= √ , t= √
σ/ n s/ n
Confidence Intervals for µ
σ
x̄ ± Zα/2 √ σ known
n
α s
x̄ ± tn−1 ( ) √ σ unknown, small samples
2 n
s
x̄ ± Zα/2 √ large samples, σ 2 unknown
n
Inferences Concerning two samples

Case Test Statistic Confidence Interval

s
(x̄1 − x̄2 ) − (µ1 − µ2 ) σ12 σ22
σ12 , σ22 known Z= r (x̄1 − x¯2 ) ± Zα/2 +
σ12 σ22 n1 n2
n1
+ n2

σ12 , σ22 unknown s

(x̄1 − x̄2 ) − (µ1 − µ2 ) 1 1
but assumed equal Z= q (x̄1 − x¯2 ) ± Zα/2 sp +
sp 1
+ 1 n1 n2
n1 n2
n1 ≥ 30, n2 ≥ 30

σ12 , σ22 unknown s

(x̄1 − x̄2 ) − (µ1 − µ2 ) 1 1
but assumed equal t= q (x̄1 − x¯2 ) ± tn1 +n2 −2 ( α2 )sp +
sp 1
+ 1 n1 n2
n1 n2
n1 < 30, n2 < 30

(n1 − 1)s21 + (n2 − 1)s22

s2p =
n1 + n2 − 2

x̄1 − x̄2 − (µ1 − µ2 )

r
s21 s22
σ12 , σ22 unknown Z= r x̄1 − x̄2 ± Zα/2 n1
+ n2
s21 s22
n1
+ n2
but assumed unequal
n1 ≥ 30, n2 ≥ 30

x̄1 − x̄2 − (µ1 − µ2 )

r
s21 s22
σ12 , σ22 unknown t= r x̄1 − x̄2 ±n1 +n2 −2 (α/2) n1
+ n2
s21 s22
n1
+ n2
but assumed unequal
n1 < 30, n2 < 30

2
 Inferences concerning Proportions

Case Test Statistic Confidences

Interval
p̂ − p0 p̂(1 − p̂)
Single proportion Z=q p̂ ± Zα/2
p0 (1−p0 ) n
n

s
p̂1 − p̂2 − (p1 − p2 ) p̂1 (1 − p̂1 ) pˆ2 (1 − pˆ2 )
Two proportions Z = q pˆ1 − pˆ2 ± Zα/2 +
p̂(1 − p̂)( n11 + 1
) n1 n2
n2

x 1 + x2 X1 X2
p̂ = p̂1 = n1
, p̂2 = n2
n1 + n2

4. Simple Linear Regression: Assumed model: Yi = β0 + β1 xi + ϵi

Pn n
xi yi − nx̄ȳ Sxy
β̂1 = Pi=1 yi2 − nȳ 2
X
β̂0 = ȳ − β̂1 x̄, n2 2
= , SSR = β̂ S
1 xy , SST =
i=1 xi − nx̄ s Sxx i=1
SSR 1 x̄ 2 s
R2 = , CI(β0 ) : β̂0 ± tn−2 (α/2)s + , β1 : β̂1 ± tn−2 (α/2) √
SST n Sxx Sxx

Hypothesis Testing

H0 : β1 = 0 vs H1 : β1 ̸= 0 H0 : β0 = 0 vs H1 : β0 ̸= 0
√
β̂1 Sxx β̂0
Test statistic : t = Test statistic : t = q
s s n1 + x̄2
Sxx

5. Regression through the Origin Yi = βxi + ϵi

Pn n n
xi y i
β̂ = Pi=1 2
X X
n 2
, SST = y i , SSR = β̂ xi y i
i=1 xi i=1 i=1

6. Correlation
Pn
cov(X, Y ) i=1 xi yi − nx̄ȳ
ρXY = q , rxy = q P
var(x)var(y) ( ni=1 x2i − nx̄2 )( ni=1 yi2 − nȳ 2 )
P

7. One-way ANOVA

ni
k X k
Y..2 Yi.2 Y..2
yij2 −
X X
SST O = , SST = −
i=1 j=1 n i=1 ni n

3
 (ȳi. − y¯j. )2
Scheffe Procedure Test Statistic: F =
M SE( n1i + n1j )

(ȳi. − ȳj. )
Fisher’s LSD Test Statistic: t = r  
1 1
M SE ni
+ nj

8. Non-parametric Statistics

k
(oi − ei )2
∼ χ2k−1
X
2
Expected frequency ei = npi , χ =
i=1 ei

Rx Ry − nR̄x R̄y
P
Spearman’s Rank correlation: Corr(Rx , Ry ) = q P
( Rx2 − nR̄x2 )( Ry2 − nR̄y2 )
P

6 ni=1 d2i
P
Spearman’s rank correlation: rs = 1 −
n(n2 − 1)
n
2
X(|oi − ei | − 0.5)2
Yates’ correction: χ =
i=1 ei

Kruskal- Wallis Test Statistic

k
#2 k
Ri2
"
12 X 1 ni (N + 1) 12 X
Test Statistic: H = Ri − = −3(N +1)
N (N + 1) i=1 ni 2 N (N + 1) i=1 ni

Correction for Ties

P
T
Adjustment factor is 1 − , Where T = t3 − t.
N3 −N

H
Adjusted test statistic is : Hc = Pg
(t3 −ti )
1− i=1 i
3
N −N

Wicoxon Rank Sum:

n1 (n1 + 1) n2 (n2 + 1)
u1 = n1 n2 + − R1 , u2 = n1 n2 + − R2 , U = min(u1 , u2 )
2 2

END OF FORMULA SHEET