Engineering statistics l Asst.Prof.dr.sufian M.

salih / LEC 2

Probability and Population Distribution

4.1. Possibilities
Probability is the most basic issues in statistics . Because all the estimates, given every decision
is expressed by a certain probability level. That is, a certain error or confidence level is
concerned. Therefore, the possibility should be noted that some ofthe basic concepts and rules.
Here is a simple possibilities will be discussed. Because the subject matter is probably the single
head.
The probability of an event is related to the ratio of the number occurring in the total number of
votes. It is calculated by the following formula:
x
P( x) 
n
Here; n: total event number, x: represents the desired number of events.
A possibilities of an event is in the range of between 0≤P(x)≤1. P(x)=0 means that the event is
impossible and P(x)=1 means that the event is going to be.
Sample 1: In a region 75 newborns of 30 was calculated girl. The girl ratio in this populations is
30
found according to this; P( x)   0, 40 .
75
Sample 2: Think that in a health center patients are coming they are 20 of flu disease, 15 of
them are internal medicine patient and 10 of them are Infectious Diseases patients.

The possibility of internal medicine patient from all of those patients is.
15
P( x)   0,33 .
45
But sometimes the event may went complicated. Like some of the patient that is waiting
treatmen may can be internal medicine patient and Infectious Diseases patients in same time.
According to this the possibility is going to me calculated:
15 10
P( x)    0,33  0, 22  0,55 dir.
45 45
As shown, each block is the event in question. In such cases, the probability of occurrence of one
or the other is the sum of each separate occurrence probability is defined as the addition rule.
If the events are not affected by each other, if it is independent of or conjunction occurs, the
probability of both events occurring together is found by multiplying the probabilities of these
events that occurs separately and is defined as the multiplication rule.

[1]
Engineering statistics l Asst.Prof.dr.sufian M.salih / LEC 2
For example, a good pass rate of operation of the two patients operated on at the same time get
0.70 and 0.80. To the possibility of a successful operation Both of these patients;

P(x)=0,70*0,80=0,56

4.2. Permutations and Combinations

If in a n person is taken from x to create a series of an ordered system permutations is necessary,

if the ordered series system is not important tha combinations are going to be used. According to
that the n person which is taken with r value of each to for research would be find with these
formulas:

n!
Permutations; order system is important: nP x  And
(n-x)!

n!
Combinations; order system is not important: nCx 
(n  x)! x!

Here, n: represents the total event numbar, and x: the number of the events.
Sample: A, B, C is 3 student from the class president and vice, including the EU, the number of
permutations to be created by taking two at a time, AC, BC, BA, CA, as the CB was 6 because
there is a significant number of ranking combination AB, BA from; AC, BC, and CA is not
different from that of CB. Thus the number of combinations; AB, AC, BC in the form of three
types.

3!
Permutation Number; 3P2   6 and
(3 - 2)!

3!
Combination Number; 3C2  3
(3 - 2)!2!

4.3. Distribution of Population

Distributions according to different data structures has improved. The data that is obtained from
research generally conforms to one of these distributions. Although it is rarely found in the
unsuitable distribution for data. Such tests are used to analyze non-parametric data. The
discontinuous in distribution are as well as in data structures, and they fall into two groups,
namely constant. Naming and grading scale shows the distribution of discrete qualitative data
obtaining. Datas that are obtained by the quantitative feature interval and ratio scales shows a
solid dispersion.
Each Distribution has a function. This function is defined as the probability density function
which is usually continuous distributions f (x) and for discrete distributions P (x) as it is defined.

[2]
Engineering statistics l Asst.Prof.dr.sufian M.salih / LEC 2
There are a number of continuous and discrete distribution function, that is defined distribution.
In this section it will be evaluated to some discrete and continuous distributions and the
importance will be explained.

Discrete distributions from the binomial, Poisson distributions and continuous distributions will
be examined to normal distribution applications. All problems will related to normal
distributions which is converted to a standard normal distribution analysis and interpretation.

4.3.1. Discrete Distributions

4.3.1.1. Binomial Distribution
It is the data type as shown by discrete distribution. This is usually the result of two distribution
of data. Yes-No, Female-Male, on-off etc. Each distribution has a density function. The function
of the binomial distribution;
n n!
P (n, x, p)    p x q n- x  p x q n x
 x x !(n  x)!
Here, n: represents the total event number toplam olay sayısını, x: represents the event that is
wanted; p: represents the possibility of the event that may happens (succes possibility); q:
represents the impossibility of the event that may not happens (unsecces of the possibility).
Descriptive parameters of the binomial distribution,
Mean of binomial distribution; µ=np
Variance of the binomial distribution; 2=npq
Properties of the Binomial Distribution:
While the event is repeating the succes rate should not be change. For example: A coins heads or
tails possibility is 1/2 it never changes. Like a incoming of child gender will may be girl or boy
the possibility is again 1/2 it never changes.
In binomial distribution the repeat of the events are less and becouse of that the succes rate of the
events are usually high.
Example: In a 4 children family, what would be the minimal boy possibility of those children ?
Here; n=4; p=0.5 and q=0.5 and if x is {1, 2, 3 ve 4}. In short;
P(x≤1)= P(x=1) + P(x=2) + P(x=3) + P(x=4) or p + q = 1 it would be.
2. Option;
P(x≥1)=1 - P(x<1)=1 – P(x=0) has to be calculated like this. As seen the 2. Option has less
calculation process.
For the answer, the 2. Option would be a wisely choose.
4!
P(x≤1)=1 - P(x  1)=1 - (0.5)0 (0.5)40  1  (0.5)4  0.94
(4  0)!
In a 4 children family the possibility of a boy child is %94.

[3]
Engineering statistics l Asst.Prof.dr.sufian M.salih / LEC 2
4.3.1.2. Poisson Distribution
It is a discrete and distribution which rare events within a certain time intervalling. In other
words, rarely distribution shown by the incident. Rarely which used in the investigation of the
events that occurred.
For example; A patient that dies from narcosis, a good secretary which makes a letter mistake
when she/he was writing, the possibility the get a cold in summer. Poisson probability of
occurrence of the desired event is too small and the number of recurring events. Probability of
occurrence of the event is low. Does not exceed above 5%. Probability density function of the
Poisson distribution,
e-   x  x
P( x)  =  is like this.
x! e x!
Here x: represents the desired number of events, µ: represents the population mean and e:
represents the logarithm base value of approximately which is 2,718. The distribution parameters
are like µ=2=n*p this and the variance is equal to its mean.
Example: The death rate of narcosis is 0.001 in a hospital. What is the death ratio from 100
patients that takes a surgery in a year. ?
Those given µ = np =100*0,001= 0.1;
The disered is P(x≤1) = P(x=0) + P(x=1). As to the equvalent whent functions values are
e-0.1 (0.1)0 e-0.1 (0.1)1 2*0.1
replaced to the formula , P( x  1)    0.1  0.18
0! 1! e
the minimum death rate of narcois is %18.

4.3.2. Continuous Probability Distributions

The most commonly encountered in the distribution of the continuous distribution is normally
distributed. Distribution is a form of continuous data that is showed.
In general, the shape of the distribution of the data is obtained by measuring and weighting. The
probability density function of normal distribution;
1 ( x -  )2
1 - .
It is f ( x)  e 2 2
. However, it must necessarily be integral in each case to
2 2

make forecasting using this formula. This is not easy. Especially in terms of time would not be
possible at all during the test. Therefore, the value of the standard normal distribution is
symmetric (z) is utilized. Z standart normal values mean is 0 (µz= 0), the vriance is (  Z2 = 1)
x-
and it is summarized in z  N(0,1). In function of f ( x) , the z is equalized to  z and

x- 1  12 z 2
when it is writin in the in equation to z , normal distribution function f ( z )  e is
 2

[4]
Engineering statistics l Asst.Prof.dr.sufian M.salih / LEC 2
changed into a standart normal distribution function. The hole integrals are given in the
additional paper that includes the Z tables.
Example: 7-year-old children in one of an mean length of 130 cm and standard deviation was
determined to be 8 cm. Find different alternatives to these options and answer the questions
below according to this.
a) What is the ratio of these kids that are above from 130 cm ?
b) What is the ratio of these kids that are above from 135 cm ?
c) What is the ratio of these kids that are below from 125 cm ?
d) What is the ratio of these kids that are between 120 cm and 135 cm ?
Solution:
a) As it is a symetric distribution the ones that are smaller and
taller from the mean is find
%50 %50
P(z<0) = P(z>0) =0.50 and it is %50.
0

135  130
b) P(x>135) =P( z  )
8
 P( z  0.63)  0.2643; % 26.43 .
0.2643

0 0.63

120  130
c) P(x>120) =P( z  )
8
 P( z  1.25)  0.1056; % 10.56 . 0.1056

-1.25 0
It has to be awared that the negative values are not included in the
z table. According the symetry has to be used. Like:
P( z  1.25) = P( z  1.25) =0.1056 it is predicted.
120  130 135  130
d) P(120  x  130)  P( z )  P(1.25  z  0.63)  ?
8 8
These values can’t be find in the z table. Thus the areas wihch are
found will be subtracted from the sum.
= P(1.25  z  0.63) 0.63

 1  {P( z  1.25)  P(0.63  z)}

-1.25 0 0.63
 1  {0.1056  0.2643)  0.63 .
According to this the kids hight ratio that are between in a hight of 120 cm – 135 cm are %63.

[5]
Engineering statistics l Asst.Prof.dr.sufian M.salih / LEC 2

4.3.3. Central limit theorem

Some research will be based on the mean mean of these assessments are repeated examples
examined in the central limit theorem. For example, you have done a research on the
hemoglobin values of a community.Let us take 10 samples from this population and the mean of
these should be 12.5 and the standard deviation should be 2.5 . According to this in a any
sample’s , sample mean value possibility, that is bigger than 13.5 is used the central limit
x- x -
theorem. And becouse of the in the formula of z  will be a change like this z  .
 x
Values in the denominator is the standard error. When it comes x instead x to the denominator
 x2 will instead  . According to the disered possibility is;
13.5  12.5
P(13.5  x )  P(  z)
2.5 / 10
 P(1.26  z)  0.1038  0.10 . 0.1038

It will %10. 0 1.26

Asst.Prof.dr.sufian M.salih

[6]