MLS 054 FORMULA

Session 4

N
Systematic sampling interval k=
n
N
Sloven’s formula n=
(1+ Ne2)

Session 6

H o−Lo
Frequency distribution table =IS +¿
Ci

Mean of grouped data x́=

∑ fx
n

Session 7

no . of deaths
Annual crude death rate ×1000
midyear population
no . of deaths of age group
Age-specific death rate × 1000
population of age group
no . of deaths of specific cause
Cause-specific death rate or Case Fatality Rate ×1000
population of cases
no . of deaths of under 1 year
Infant mortality rate ×1000
number of live births
no . of deaths within1 st month
Neonatal mortality rate × 1000
number of live births
no . of deaths of ¿ 4−51 weeks ¿ ×1000
Postnatal mortality rate number of live births
no . of deaths due ¿ pregnancy ¿ ×1000
Maternal mortality ratio total no . of live births
no . of deaths of specific cause
Proportional mortality ratio ×100
total deaths
no . of deaths50 y . o up
Swaroop’s index ×100
total no . of deaths
no . of live births
Crude birth rate × 1000
total population
no . of live births
General fertility rate × 1000
no . of childbearing age
 no . of new cases
Incidence rate ×1000
total population
no . of existing cases
Prevalence proportion ×1000
total population
no . of deaths of given cause
Case fatality proportion ×1000
no . of cases

Session 10

X n +1 X n+ X n
+1
Finding the median M= or M + 2 2
2 2
Median of grouped data ~
x=Lcb + c × ¿

≜f1
Mode of grouped data mode=Lcb +c [ ≜ f 1+≜ f 2 ]

Session 11

2
∑ x2 ∑ x
Population variance (ungrouped) σ 2
= −
N ( ) N
2
Population standard deviation (ungrouped) σ=
√ ∑ x2 −
N (∑ )N
x

Sample variance (ungrouped) s=

1
2
n−1 [∑ x− 2
n]
(∑ x )

2
Sample standard deviation (ungrouped) s=
1
n−1 √ [∑ x−

∑ fx2 − ∑ fx
2 (∑ )
n ] x

Population variance (grouped) σ 2

=
N ( ) N
2
Population standard deviation (grouped) σ=
√ ∑ fx2 −
N (∑ )N
fx

Sample variance (grouped) 2

s=
1
n−1 [∑ fx− 2 ( ∑ fx )
n ]
 2
Sample standard deviation (grouped) s=
√1
n−1 [∑ fx− 2 ( ∑ fx )
n ]

Session 12

Mean absolute deviation (ungrouped) MD mean=

∑|x−μ|
N
Mean absolute deviation (grouped) MD mean=
∑ f |x −μ|
N
Median absolute deviation (ungrouped) MD median=
∑|x−median|
N
Median absolute deviation (grouped) MD median=
∑ f |x−median|
N
Coefficient of mean deviation (relative MD
coefficient of MD mean=
dispersion) mean
Coefficient of median deviation (relative MD
coefficient of MD median=
dispersion) median

Session 13

th
n+1
Percentile (ungrouped) Pi=i × [ ]
100
¿ −CFpre
Percentile (grouped) Pi=Lcb Pi +c × [
100
f ]
Exact percentile Pi=P1 +extra decimal ×( P2−P1)
th
n+1
Decile (ungrouped) Di=i × [ ]
10
¿ −CFpre
Decile (grouped) Di=Lcb Di +c ×
10
[
f ]
Exact decile Di=D1 +extra decimal × ( D 2−D1 )
 th
n+1
Quartile (ungrouped) Qi=i ×
4[ ]
¿ −CFpre
Quartile (grouped) Qi=Lcb Qi+ c ×
4
[f ]
Exact quartile Qi=Q1 +extra decimal × ( Q 2 −Q1 )

Session 14

Sample mean μ=
∑x
n
∑ ( x−μ )2
Sample standard deviation σ=
σ
√ n−1
Coefficient of variation (population) CV = × 100 %
μ
s
Coefficient of variation (sample) CV = ×100 %
x́

Inter-quartile range (IQR) IQR=Q3 −Q1

Q 3−Q 1
Quartile deviation (QD) QD=
2

Session 15

Conditional probability P(B|A)

Independent events P(B|A) = P(B)

Multiplication rule P(A and B) = P(A) × P(B|A)

Mutually exclusive events P (A and B) = 0

Non-mutually exclusive events P (A and B) ≠ 0

Addition rule P(A) + P(B) – P(A and B)

Session 16

n!
Binomial experiment b ( k ; n , p )=
( k ! ( n−k ) !)
k
× p ×q n−k

Binomial distribution δ =√ npq

Mean μ=np

Session 17

CONFIDENCE COEFFICIENT Α Zα α/2 Zα/2

90% 0.10 1.282 0.05 1.645
95% 0.05 1.645 0.025 1.960
99% 0.01 2.326 0.005 2.576

Point estimator X, interval estimator (1-α)

¿
100% confidence interval when δ is known
Point estimator X, interval estimator (1-α)
¿
100% confidence interval when δ is unknown
2
Sample standard deviation (ungrouped) s=
1
√
n−1 [∑ 2
x−
n ]
(∑ x )

Population proportion estimators

[ √
p−Z a
2
pq
n
, p+ Z a
2 √ ]
pq
n
x
Population proportion p= ∧q=1− p
n

Session 19
 ONE SAMPLE Z-TEST
Set Null Hypothesis Alternative Hypothesis Number of Tails
1 µ=M µ≠M 2
2 µ≥M µ<M 1
3 µ≤M µ>M 1

x−μ0
Z-score z=
σ √n
t-score degrees of freedom ν=n−1

Session 20

Ho Test statistic Ha Critical Region

Case 1: δ12 and δ22 are known
µ1 - µ 2 = d o ( x 1−x 2 )−d o µ1 - µ 2 < d o Z < -Za
z= µ1 - µ 2 > d o Z > Za
σ 21 σ 22

Case 2: δ12 = δ22 but unknown

√ +
n1 n2
µ1 - µ 2 ≠ d o |Z| > Za/2

µ1 - µ 2 = d o ( x 1−x 2) −d o µ1 - µ 2 < d o t < -ta

t= µ1 - µ 2 > d o t > ta
1 1
sp
√ +
n1 n 2
µ1 - µ 2 ≠ d o |t| > ta/2

2 ( n1 −1 )( s 1 ) ( 2 ) + ( n2−1 ) ( s2 ) ( 2 )
sp =
n1+ n2−2

ν=n1 +n2 −2
2 2
Case 3: δ ≠ δ and unknown
1 2
µ1 - µ 2 = d o ( x 1−x 2) −d o µ1 - µ 2 < d o t < -ta
t= µ1 - µ 2 > d o t > ta
s21 s 22
√ +
n1 n2
µ1 - µ 2 ≠ d o |t| > ta/2
 2
s 21 s22

ν=
( +
n1 n2 )
2 2
s21 s 22
( ) ( )
n1
+
n2
n1 −1 n2−1

Case 4: Two related or paired samples

µo = d o d−d o µ1 - µ 2 < d o t < -ta
t= µ1 - µ 2 > d o t > ta
Sd
µ1 - µ 2 ≠ d o |t| > ta/2
√n

d́=
∑d
n
2
(∑ d )
Sd =2 1
n−1 (∑ d−
n )
Session 21

between−group variance
F-distribution F=
within group variance

Critical value (Fa) df =( k −1, n−k )

2
T 2 T 2 T 2 (∑ x )
Sum of squares between groups
(
SSbetween = 1 + 2 + k −
n1 n2 nk n)
Sum of squares within groups SSwithin =SS total−SSbetween
2

Total sum of squares 2 (∑ x )

SSbetween =∑ x −
n
Degrees of freedom between groups df =k −1

Degrees of freedom within groups df =( n−1 )−( k −1 )

Total degrees of freedom df =n−1

 SS between
Mean square between groups MS between =
df between
SSwithin
Mean square within groups MS within=
df within
MSbetween
F-value F=
MS within

Session 22

( n ) ( ∑ xy ) −( ∑ x )( ∑ y )
Pearson’s coefficient of correlation r= 2 2
√ ( n ) (∑ x )−(∑ x ) √( n ) (∑ y )−(∑ y )
2 2

Simple linear regression analysis y=a+bx

( n ) ( ∑ xy )−( ∑ x )( ∑ y )
Regression coefficient b= 2
( n ) ( ∑ x 2 ) −( ∑ x )

Regression constant a= y−bx

Ho Test statistic Ha Critical Region

ρ=0 r √ n−2 ρ<0 t < -ta
t= ρ>0 t > ta
√1−r 2 ρ≠0 |t| > ta/2
ν=n−2

Session 23

Degrees of freedom df =k −1

( O−E )2
Chi square test statistic χ 2=∑
E
Expected frequency E=np
Session 24

( total ∈row )( total ∈column )

Expected values E=
total ∈sample
Degrees of freedom df =( number of rows−1 ) ( number of columns−1 )

2 (|O−E|−0.5 )2
Yate’s correction for continuity χ =∑
E
2
x
Phi coefficient φ=
√ n