UNIVERSITY OF GONDAR

COLLEGE OF HEALTH SCIENCE,

INSTITUTE OF PUBLIC HEALTH

DEPARTMENT OF HEALTH INFORMATICS

Biostatics Group Assignment

Submitted to mr Wullo.S(MPH)
 Group member

Full_Name Id.No

1. Dawit Tesfa------------------------------GUR/21846/2010

2. Kibruyisfa Telay ------------------------GUR/21858/2010

3. Mirtzer Tamene-------------------------GUR/21862/2010

4. Molla dessale----------------------------GUR/21864/2010

5. Seada Yesuf------------------------------GUR/21869/2010
 Answer

Question 1 answer

Steps to construct frequency distribution

A. Let find largest and smallest value

Largest=59 and smallest=21
B. Range= Largest-smallest: 59-21=38
C. Let select number of class(k)
K= 1+3.32log(n)

=1+3.32log(60)

=6.88 round up to 7

D. find width(W)
𝑅 38
w= =5.42 round up to 6
𝐾 7

E. Select starting point and construct upper and lower limit

Starting point is 21
 LCL= 21 27 33 39 45 51 57
Subtract U from the upper limit of first class to find to first class of upper
limit: 27-1=26
UCL= 26 32 38 44 50 56 62
F. Construct class limit by combined LCL and UCL.
21-26
27-32
33-38
39-44
45-50
51-56
57-62
G. Find class boundaries
20.5-26.5
26.5-32.5
32.5-38.5
38.5-44.5
44.5-50.5
50.5-56.5
56.5-62.5
VIII. find frequency distribution and relative frequency distribution (note this is
for question one A and B).

Class Class Class frequency CF < Relative. Cumulative.

Limit boundary mark type F RF
21-26 20.5-26.5 23.5 6 6 0.1 0.1
27-32 26.5-32.5 29.5 12 18 0.2 0.3
 33-38 32.5-38.5 35.5 4 22 0.7 0.36
39-44 38.5-44.5 41.5 6 28 0.1 0.46
45-50 44.5-50.5 47.5 9 37 0.15 0.61
51-56 50.5-56.5 53.5 15 52 0.25 0.86
57-62 56.5-62.5 59.5 8 60 1 1

C. line graph is appropriate

16
14
12
Frequency

10
8
6
4
2
0
20.5-26.5 26.5-32.5 32.5-38.5 38.5-44.5 44.5-50.5 50.5-56.5 56.5-62.5
Mid Point
Question 2 answer

Calculate mean

=141+354+249+427.5+802.5+476
6

= 2592/60

= 43.2

Calculate median

LCB+( n/2-FC) W
fc
LCB= lower class boundary of the median class

Fc= cumulative frequency just before the median class

fc=frequency of the median class

W =class width and n=number of observations

20.5+(30-28)6
9
=27.33
Calculate Mode

= L+ ( fm-fm-1 )W
(Fm-fm-1)+(fm-fm+1)

=50.5+ (15-9) 9
( 15-9)+(15-8)
= 53.2

Calculate variance for grouped data

 2   f ( M   )2 / N

= (23.3-43.2)26+(29.5-43.2) 212+(35.5-43.2) 24+(35.5-43) 26+(47.5-43.2) 29+(53.5-43.2)

2
15+(59.5-43.2) 28
60
= 8718.6

60

= 145.31

Standard deviation(SD) = √145.31

=12.05
 
Calculate coefficient of variation(CV)=
µ

𝟏𝟐.𝟎𝟓
=
43.2

= 0.279

Question 3 answer
city mean variance s. deviation c. variation
City one 23 10 3.16 13.74
City two 21.8 1.76 1.32 6.08
City three 29.2 14.96 3.86 13.24

Based our calculated data city two have small coefficient of variation (6.08)
compared to the rest city. A variable having small coefficient of variation is more
consistent so that city two has more consistent temperature than the other.

Question 4 answer

Given
P=25%=0.25 Require
N=5 a)=p(x=x)
q=1-p b)=p(x>=2)
q=1.025 c)=p(x<=3)
q=0.75. d)=expected value
e)=s2
 solution

a) p=x/n=x=np
x=5x0.25=1.25=1
p(x=x)=(nx) px(1-p)n-1
p(x=1)=(31)(0.25)1 (0.25)4
5!/(5-1) !1! (0.25)1(0.25)4=0.395=0.4
B) P(x>=2) =1-p(x<2)
=1-[(p(x=0)+P(x=1)]

=1-[(5)(0.25)0(0.75)5(51) (0.25)1(0.75)4]

=1-[5!/(5-0) !0! . (0.25)0(0.75)5+5! /(5-0,1) .(0.25)1(0.75)4]

=1-(1*0.23+5(0.25) (0.316)

=1.(0.625)

=1-0.625=0.375

c) p(x<=3) =(nx) px(1-p)n-x

p(x<=3)=p(x=0) + p(x=1) p(x=2) + p(x=3)

=(50) (0.25)0 (0.75)5 + (51) (0.25)1(0.75)4+(53) (0.25)3 * (0.75)2

=5! / (5-0) ! 0! (0.25)0 (0.75)5 + 5! / (5-1) ! 1! (0.25)1 (0.75)4 +5! / (5-3) ! 3! (0.25)3 (0.75)2
=0.23+0.39+0..84

=0.704=0.7

D) expected values = E(x)

E(x) = np = 5* 0.25 =1.25

E) s2 = s2 = npq

s2 =5*0.25*0.75=0.9375

9|Pag e
Question 5 answer

A.
Given

X=50

µ=60  = 𝟏𝟓

we should normalize to make standard normal distribution

𝒙−µ
Z= = 50-60/15 = -0.66


From the table 0.66=0.225

Pr(X>=x ) = Pr(Z>=zo) (upper tail)
Pr(x>50)=p(z>-0.66)=0.225

We can also find area under the curve as follow

=0.5_0.225

10 | P a g e
 B.

Given

X=30

µ=60  = 𝟏𝟓

we should normalize to make standard normal distribution

𝒙−µ
Z= = 30-60/15 = -2


Z value from the table -2=0.447

Pr(X<=x ) = Pr(Z<=zo) (Lower tail)
Pr(x<30)=p(z<-0.447)= 0.977

We can also find area under the curve as follow

By finding area under the curve we obtain 0.997

11 | P a g e
 C.

Given

X=30 and X=60

µ=60  = 𝟏𝟓

we should normalize to make standard normal distribution

𝒙−µ
Z= = 30-60/15 = -2 from the table -0.4772


𝒙−µ
Z= = 60-60/15 = 0 from the table 0.00


By finding under the curve we get

0-0.447= 0.477

D.

Given

X=90

µ=60  = 𝟏𝟓

12 | P a g e
we should normalize to make standard normal distribution
𝒙−µ
Z= = 90-60/15 = 2 from the table 0.4772


Question 6 answer

Given

CI=95%

Width (margin of error)=0.06

P=0.20 find n?

Q=0.8

N=(1.96)20.2(0.8)

=0.61/0.0036

=169.44 round to 170

13 | P a g e
Question 7 answer

Given

N=8

X=438 𝑺𝑫 = 𝟏𝟔

Find the 99%,95% and 90% of confidence interval

It is better to use T-test because sample size is small(n=8) and variance is unknown.

99% CI:

𝑺
=X±tα/2. tα/2= 6.31 from t=table
√n

𝟏𝟔
=438±6.31
√8

= 438±35.65

(402.3, 473.65 )

We conclude that 99% sure that patient treated in emergency room is average
402.3, 473.65

95% CI:

𝑺
=X±tα/2. tα/2= 2.571 from t=table
√n

𝟏𝟔
=438±2.571
√8

438±2.571×5.65

= 438±14.52

(424, 452.52 )
14 | P a g e
We conclude that 95% confident that average patient treated in emergency room is
between 424 and 452.52

90%CI:

𝑺
=X±tα/2. tα/2= 1.81 from t=table
√n

𝟏𝟔
=438±1.81
√8

438±1.81×5.65

= 438±10.22

(427.78, 448.22 )

We conclude that 90% confident that average patient treated in emergency room is
between 427.78 and 448.22

Question 8 answer

Given

N=25 SD=16 CI=95%

X=73 µ=70

Can we conclude the population mean is higher than 70?

Step 1 identify null and alternative hypothesis

H0:µ=70 H1:µ>70

Step two select test statics

Since variance is not known and n is small we use t-test

15 | P a g e
Step two identify significance level

α=0.05

Step three calculation

𝒙−µ
t-cal=
𝑺√ n

𝟕𝟑−70
=
𝟏𝟔√ 25

=0.0375

Step four decision

t-cal<t-tab; accept null hypothesis.

0.0375<2.571 accept null hypothesis

step five

conclusion:

we conclude that at 0.05 level of significance the population mean is equal to 70.

Question 9 answer

Given

N=950 p=611/950= 0.64

X=611

We use z test based on the above data

 p ± z ×(√ p(1 − p)/n

=0.64±1.96×0.095

=0.64±0.186

16 | P a g e
=(0.454,0.82)

Question 10 answer
The following table show 2×2 table

Disease status total

disease Not disease
Exposure Eat sandwich(E) 109 A 146 B 255
status Not eat 4C 34 D 38
sandwich(NE)
total 113 180 293

B. to check identical

Let calculate the expacted value of e11 and e12

𝒄𝟏𝒓𝟏
E11=
n

(𝟏𝟐𝟐𝟓)(𝟏𝟏𝟑)
263

= 𝟗𝟔. 𝟔𝟕
𝒄𝟏𝒓𝟏 (𝟑𝟖)(𝟏𝟏𝟑)
E12= =
n 263

=16.32

𝟏𝟎𝟗−𝟗𝟔.𝟔𝟕)𝟐
Sandwich eat and become ill= ( =1.57
96.67

𝟒−𝟏𝟔.𝟑𝟐)𝟐
Sandwich not eat and become ill= ( =9.3
16.32
We conclude that the two group become ill is not identical

17 | P a g e
C. to test association

Step one identify null and alternative hypothesis

H0: no association between gastroenteritis outbreak and given lunch

H1: there is association between gastro enteritis outbreak and given lunch

Step 2 identify level of significance α=0.05

Step 3 Find the critical value

Df=(r-1) (c-1) =( 2-1) (2-1)= 1

X2tab=3.84

Step 4 calculation

nad  bc
2
2 
(a  c)(b  d )(a  b)(c  d )

293(109×34−146×4)2
=
(109+4)(146+34)(109+146)(4+34)

293(3706−584)2
=
(113)(180)(255)(38)

2855837012
=
197094600

X2cal=14.49

Step 5 decision

If X2cal< X2tab accept null hypothesis but 14.49<3.84 is false

18 | P a g e
we reject the null hypothesis

step 6 conclusion

at 0.05 level of significance there is no association between lunch served and

gastroenteritis outbreak.

Question 11 answer

1.
A: correct

B: type II error

C: correct

D: type I error

2.

A: type II error

B: correct

C: type I error

D: correct

19 | P a g e