Download as pptx, pdf, or txt
Download as pptx, pdf, or txt
You are on page 1of 30

dr.

Ayu Swandewi A, MPH


Dept. Biostatistics, Demography and Reproductive Health
School of Public Health, Fac. Medicine, Udayana University
 To describe the definition of population and
sample
 To define the population and sample for a study
 To explain the conditions required for a
representative sample
 To choose appropriate sampling methods
 To describe types of data and variable as well as its
importance on data analysis
GROUP OF OBJCETS
with certain characteristics

Human, animal, program, hospitals


To whom the result will be
generalised

Usually similar to population of study or


broader
TARGET POPULATION

SAMPLED
POPULATION
TARGET POPULATION

SAMPLED
POPULATION

SAMPLE
Researcher wish to explore the degree of pain among
cancer patients

Target Population

All cancer patients

Sampled population

Cancer patients at Sanglah Hospital from January- December 2011

Sample

Cancer patients that are selected from sampled population


A researcher want to study the effect
of hormonal contraception use to the
risk of breast cancer;
please determine:
◦ Target population
◦ Sampled population
◦ Sample
 Sample should represent the population
 Representativeness of sample influence by
◦ Sample size
◦ Sampling technique
◦ Subject that retain on the study
◦ Response rate
Representativeness

Sample Size Sampling technique

Proper method
The bigger the
gives more
more
representative
representative
sample
SAMPLE SIZE

1.VARIABILITY ()

2. RELIABILITY ()
and ()
/2 /2

3. PRECISION () POPULASI


SAMPEL
TEKNIKSAMPLING

Random Non Random

• Simple R.S. • Purposive S.


• Stratified S. • Convenient S.
• Systematic S. • Consecutive S.
• Multi Stage S. • Quota S.
• Cluster S. • Snow Balling t.
• PPS • etc.
• etc.
Technique Indication

Simple Random Probability of events in


Population is homogenous
Systematic Random Population is homogenous;
If we wish the sample
distributed systematically
Stratified Random Probability of events in
population is heterogeneous
Cluster Sampling Group of people with the
same characteristic
(profession, geographic)
Probability sampling
Sampling frame available or feasible to
construct
To get sampling frame
 Construct : Listing
 Use available list : Medical Registration,
students/employee records,
households records
SIMPLE RANDOM
 Random Number Table

 Digit:  1 digit: 1 digit population


 2 digit: 2 digit population
 3 digit: 3 digit population
 4 digit: 4 digit population
If we wish to select
13 sample from125
population
 Choose the random table (digit)
 Choose the page
 Choose the first sample
 Choose the next samples horizontal/diagonally
3958 0714 6102 5938 3298 4027
0591 8271 4058 3906 3014 1836
1682 0395 6241 9073 9125 7061
9014 2518 0395 4035 1056 3482
6015 1029 9021 4837 5638 0371
7298 5936 1253 0148 0391 6841
4619 2743 3012 9015 8514 6503
9013 9518 1538 4027 7425 9138
5137 4012 9506 6148 2957 2485
2501 7139 4920 1350 6294 1570
8320 3450 2864 7291 5827 8052
9361 3062 5893 2057 1064 3859
0572 9471 0572 9031 0475 8037
6851 8503 8501 0631 4792 5038
Sampling Technique
 Systematic Random  Tipe:  Liniar
 Circular

 Select 13 samples from125 population

1. Construct the sampling frame


2. Calculate theINTERVAL (k): N/n
3. Choose the 1st Sample (RS):
 Liniar: select from pop. no 1 to no  interval
 Circular: select from pop. no.1 up to the last no

4. Choose the next samples by following


certain rule
1. Construct the sampling frame
2. Calculate INTERVAL (k): N/n  125 / 13 = 9,615…...
1
2 3. Choose the 1st Sample:
3 (Random Start)
4
5  Liniar: select no. of the
6 populaton from 1 to  interval
7
8 4. Select the next samples by following
9
10 R (i) = RS + (i-1)(I)
11 16
2 = 6 + (1) x 9,615 = 15,615 ----->
.
. 3 = 6 + (2) x 9,615 = 25,230 -----> 25
.
. 4 = 6 + (3) x 9,615 = 34,845 -----> 35
. And so on . . . . . . . . . .
.
125
1. Construct the sampling frame
2. Calculate INTERVAL (k): N/n  125 / 13 = 9,615…...
1
2 3. Select the 1st sample (Random Start :)
3
 Circular: choose the pop no.
4
5 from 1 to the last
6
7
4. Choose the next samples as following
8 R (i) = RS + (i-1)(I)
9
10 2 = 92 + 1 x 9,615 = 101,615 -----> 102
11 111
3 = 92 + 2 x 9,615 = 111,230 ----->
.
. 4 = 92 + 3 x 9,615 = 120,845 -----> 121
.
92 Up to No. 125 . . . . We’ve not got 3 samples
.
----> Return to NO. 1
.
…… Equation
125
R (i) = RS + (i-1)(I) - N

5 = 92 + (5-1)(9,615) - 125 = 5,460 -----------> 5

6 = 92 + (6-1)(9,615) - 125 = 15,075 ----------> 15

7 = 92 + (7-1)(9,615) - 125 = 24,690 ----------> 25

And so on . . . . . . . . . . . . . . . . . . .

Note:
- RS = First Sample/random start
-i = Interval
- N = Total no of population
 Inclusion
◦ Represent the population
 Exclusion
◦ Contraindication
◦ Control confounding variable
◦ Assure data quality
 Drop out criteria
◦ Assure data quality
 Substitution criteria
◦ criteria to replace a particular individu
Study on the risk of PID among women who use
IUD and who don’t use IUD

Inclusion • Reproductive age(15-49 th)


Criteria • Married

• Contraindication to use IUD or


Exclusion laparoscopic surgery

Criteria • Imunodeficiency, multipartner


• Not cooperative
A clinical trial to study the effect of zinc on
the length of stay among infant patient acute
diarrhea with mild dehydration

Please determine the sampling


criteria!
• Pasien diare akut dengan
dehidrasi ringan
Kriteria • Umur 0-1 tahun
inklusi • Dirawat di RS Sanglah dari
Januari-Desember 2011

• Immune deficiency
Kriteria • Malnutrition
• Alergi probiotic
Eksklusi • Komunikasi buruk
VARIABLES and DATA

VARIABLE DATA STATISTIC


Aspects that is Result of Result of
observed/ observation / Analysis
measured Measuement

Sex L, L, F, L, F, 60% male

Birth weight 1,5 2,0 3,5 4,0 Mean BW= 2,8


3,0 kg

Nutritional good, poor, 40% good


status poor, good, 40% poor
moderate
Clasification of Variabel
QUANTIFIABILITY SCALE Function
CATEGORICAL NOMINAL •Dependent
• sex • sex,occupation
• occupation •Independent
ORDINAL
• economic
• Intermediate
status

NUMERIC DISCRETE • Confounding


• Birth weight • Parity
• Age • Control
INTERVAL
• temperature • Random
RATIO
• Weight,height
Relation Between Variables

Independent Intervene/Intermediate Dependent


Vaccination Immunity Dissease

Age
CONFOUNDING

Nutritional
status
CONTROL

Season

RANDOM
 Often some variable derived from other original
forms
 Method:
◦ Calculated from other variables
◦ Recoded from other variable
◦ Based on threshold value
◦ Transformed variable

You might also like