The words you are searching are inside this book. To get more targeted content, please make full-text search by clicking here.
Discover the best professional documents and content resources in AnyFlip Document Base.
Search
Published by balkish_mahadir, 2021-07-14 00:33:57

Draf Laporan PLB

Draf Laporan PLB

Nota Teknikal
Technical Notes

where; BP
=O

B = Census Population
P = PES Population
O = Matched persons common to Census and PES

Since: T  BP  B  ~x

O

where; ~x = estimate of number under enumerated

= T-B

So Under Enumeration Rate is given by:

Rˆ   ~x  100
 T 

Penilaian Liputan Banci 2020 182

Post Enumeration Survey 2020

Nota Teknikal
Technical Notes

8.2 Percentage Degree Of Agreement And Index Of Inconsistency

The formulae for the calculation of the Percentage Degree of Agreement3 and

Index of Inconsistency are as follows:

Census Classifications

(j=1,2.......,c)

Categories 1 2 .... j .... c Total

PES 1 y11 y12 …. y1j ….. y1c y1.
Classifications 2 y21 y22 …. y2j ….. y2c y2.
. . . …. . …. . .
(i=1,2...c) . . . …. . …. . .
. . . …. . …. . .
i yi1 yi2 …. yij …. yic yi.
. . . …. . …. . .
. . . …. . …. . .
. . . …. . …. . .
c yc1 Yc2 …. ycj ….. ycc yc.
Total y.c n
y.1 y.2 …. y.j …..

where;
n = Total cases reported by the Census and PES
c = Number of category by demographic characteristics
yij = Number of sample element for i category in PES and j category
in Census
y.j = Total number of sample element for j category in Census
yi. = Total number of sample element for i category in PES

Index of Inconsistency for i category,

y.i  yi.  2yii
yi.  yi. n 
  ˆIiny.in y.i x100 , i  1,.........c

3 For further information on the Percentage Degree of Agreement and the Index of Inconsistency, please refer to the U.S Department of
Commerce, Bureau of the Census, “Evaluating Census of Population and Housing”, September 1965 - Chapter 3 page 69,70,87 and 88.

Penilaian Liputan Banci 2020 183

Post Enumeration Survey 2020

Nota Teknikal
Technical Notes

Aggregate of Inconsistency Index,

n  c y ii 
ˆIAG 
 n i x100
1 c
y.i yi.
n i

Percentage Degree of Agreement for ith category,

Di  yii 100
yi.

Overall Percentage Degree of Agreement,

DAG  yii x100
n

9. PES APPROACH AND OPERATIONS
9.1 PES 2020 is using the ‘de jure’ approach which is everyone is counted on the

Census Day (7 July 2020) according to their usual place of residence.

9.2 To minimize problems of recall on the part of respondents, data collection for the
PES commenced immediately once fieldwork for the Census was completed.
The PES 2020 was carried out in four phases as follows:
(i) Coordinating data from Document 2 (Census) to PES System
CCES system will integrate the information from Document 2 (Census).
(ii) Listing and Enumeration
All living quarters within selected EBs were listed, household and persons
within the living quarters will re-enumeration. Demographic characteristics
such as name, age, sex, ethnic group citizenship, religion and marital
status were also collected.
(iii) Office Reconciliation
Living quarters, households and persons listed and enumerated by the
CCES were matched with the information obtained by the Census. The
demographic characteristics of matched persons were compared.

Penilaian Liputan Banci 2020 184

Post Enumeration Survey 2020

Nota Teknikal
Technical Notes

(iv) Field Reconciliation
Field reconciliation will be conducted due to any information which does
not match by re-visit that living quarter to make a confirmation.

9.3 A total of 586 enumerators and 109 supervisors were hired for listing,
enumeration, and field reconciliation work at the state level. Meanwhile, a total of
85 Daily Part-Time Workers were hired for data coordination from Document 2
and office reconciliation work at the Headquarters level. With one supervisor to
every seven enumerators, close supervision was possible to ensure work at the
each phase of the survey progressed smoothly. Every effort was made to
minimise the errors incurred by the enumerators. This included intensive training,
proper supervision and quality checks.

10. COVERAGE AND CONTENT ERRORS
10.1 Coverage Error

Coverage error occurs when part of living quarters, households or population
were omitted erroneously included or duplicated during census enumeration.
Normally, for large scale projects like the Population and Housing Census, this
type of errors is inevitable.
(i) Omissions

Living quarters, households or person were presumed to be left out if they
were not covered by the enumerator during the census time-frame.
Amongst the reasons were:
 The enumerator was unable to identify isolated living quarters or its

location was not accessible.
 Temporarily occupied living quarters which the enumerator assumed to

be unoccupied after failing to inform its occupants after three visits and
neighbours were not able to give the required information.

Penilaian Liputan Banci 2020 185

Post Enumeration Survey 2020

Nota Teknikal
Technical Notes

 The enumerator was unable to get details on the exact number of
households for the particular living quarters.

 Household members were away temporarily during the census time
frame.

(ii) Erroneous Inclusions
This type of error normally occurs when living quarters, households and
population supposed to be excluded, were however covered by the
enumerator. Amongst the reasons were:
 Visitors were assumed as usual members of that particular living
quarter during the Census Day.
 Persons died before the Census Day.
 Households and its members moved out before the Census Day

Table 3 shows the reasons of omissions and erroneous inclusions by the
enumerators for Census 2020. For the omissions errors category, population
were left out by the enumerators i.e. X per cent of the error, followed by
omissions of living quarters (X%). For erroneous inclusions, listing person errors
by the enumerator was contributed x per cent followed by mismatch of listing
living quarters by the enumerator (x%). These reasons showed that the tendency
for an enumerator to make mistake for living quarters and persons are quite
common.

Penilaian Liputan Banci 2020 186

Post Enumeration Survey 2020

Nota Teknikal
Technical Notes

Table 3: The reasons of omissions and erroneous during Census 2020

The reasons Per cent
(%)
Omissions
100.00
Members was left behind by enumerator (%)
Living quarters was left behind by enumerator
Households was left behind by enumerator 100.00
Move in before Census Day
Living quarter was wrongly defined
Visitor/relative live temporarily
Plan to move out
Unknown
New born before Census Day
Usual members but not stayed at living quarter on Census Day
Others

Total

Erroneous

Members was wrongly entered by enumerator
Living quarters was wrongly entered by enumerator
Households was wrongly entered by enumerator
Living quarter was wrongly defined
Border issues
Usual members but not stayed at living quarter on Census Day
Visitor/relative live temporarily
Moved out before Census Day
Passed away before Census Day
Plan to move in
Others

Total

Penilaian Liputan Banci 2020 187

Post Enumeration Survey 2020

Nota Teknikal
Technical Notes

10.2 Content Error
Content error is the error due to the difference in responses on specific
demographic characteristics of matched persons obtained from the Census and
PES.

Comparison and evaluation made are based on several demographic
characteristics such as sex, ethnic, age group and mode of enumeration.

10.3 Methods Analysis of Content Error
Two methods were used in evaluating the content error:
(i) Percentage of Degree of Agreement (D)
Measures the extent to which the responses from the Census coincide
with that of the PES. This measurement is based on the assumption that
PES data collection is better than the Census. Theoretically, this
measurement has a value between 0 to 100 per cent. A value of 0 per
cent shows there is no matching response for a characteristic. On the
other hand, a value of 100 per cent means that all the responses for
criteria obtained from the Census are identical with those of the PES.

(ii) Index of Inconsistency (l)
Measures the level of inconsistent in response for each characteristics or
class obtained from the Census compared to PES. This index has a value
between 0 to 100 per cent and high value corresponds to a high response
error level.

Penilaian Liputan Banci 2020 188

Post Enumeration Survey 2020

Nota Teknikal
Technical Notes

11. CONCEPTS AND DEFINITIONS
11.1 Urban areas

Urban areas are defined as gazette areas with adjoining built up areas with a
combined population of 10,000 or more.

Built-up areas were defined as areas contiguous to a gazetted area and had at
least 60 per cent of their population (aged 10 years and over) engaged in non-
agricultural activities as well as having modern toilet facilities in their housing
units.

Urbanisation is a dynamic process and keeps changing with development and
growth. Thus, the urban areas for the 1980, 1991, 2000, 2010 and 2020
Censuses do not necessarily refer to the same areas, as areas fulfilling the
above criteria of urban continue to expand and grow.

11.2 Enumeration Block
An enumeration (EB) block is a land area which is artificially created and consists
of specific boundaries. On average, one EB contains about 80 to 120 living
quarters with approximately 500 to 600 persons.

11.3 Living Quarter
(i) Living quarters (LQ) have been defined for census purpose as places of
abode, which are structurally separate and independent. The terms,
‘separate’ and ‘independent’, mean the following:
 Separate
A structure is considered separate if it is surrounded by walls,
fence, etc. and is covered by a roof.

 Independent
A structure is independent if it has direct access via public path,
communal passageway or space (that is occupants can come in or

Penilaian Liputan Banci 2020 189

Post Enumeration Survey 2020

Nota Teknikal
Technical Notes

go out of their living quarters without passing through others’
premises)

(ii) There are two categories of living quarters, that is housing unit and
collective living quarters

 The housing units are classified into detached, semi-detached,
terrace, link, cluster, townhouse, flat, apartment, condominium,
service apartment, Small Office Home Office (SOHO), Smaill Office
Flexible Office (SOFO), Small Office Versatile Office (SOVO),
shophouse/office, long house (Sabah and Sarawak only),
improvised/temporary hut, room (with direct access from outside)
and water village house.

 The collective living quarters are meant for living by a large group
of individuals and usually have some common facilities such as
kitchen, toilet, bathroom, lounge and bedrooms. Examples of
collective living quarters are hotel, homestay, rest house, medical
institutions, institutions of higher learning, charities/social welfare
institutions, religious places of worship, prisons, detention centers,
and temporary quarters for workers.

11.4 Household
A household consists of related and/or unrelated persons who usually live
together and make common provisions for food and other essentials of living.

11.5 Ethnic groups and citizenship
Classification of ethnic group used for Census 2010 ethnic group set by Inter-
Agency Technical Committee (IATC). IATC is a committee formed to coordinate
and monitor the implementation and use of standardised codes, classifications
and definitions used by the Department of Statistics, Malaysia and other
government agencies. For the purpose of tabulation and analysis, and taking into

Penilaian Liputan Banci 2020 190

Post Enumeration Survey 2020

Nota Teknikal
Technical Notes

account the diverse ethnic group in Peninsular Malaysia, Sabah, W.P. Labuan
and Sarawak, major ethnic groups according to region are as follows:

(a) Peninsular Malaysia (b) Sabah & W.P. Labuan (c) Sarawak

Malaysian Citizens (d) Malaysian Citizens (e) Malaysian Citizens
(f) Bumiputera Bumiputera (u) Bumiputera
(g) Malay Malay
Kadazan/Dusun Malay
Other Bumiputera Bajau Iban
Murut Bidayuh
(h) Other Bumiputera Melanau
(i) Other Bumiputera
(j)
(k) Chinese (p) Chinese (v) Chinese
(l) (q) (w)
(r) Indians
Indians (s) Indians
(m) (t) Others (x)
(n) Others (y) Others
(o)
Non-Malaysian citizens Non-Malaysian citizens
Non-Malaysian citizens

12. RELIABILITY OF STATISTICS

The statistics generated based on survey conducted with probability sampling
are subjected to two types of errors which are sampling and non-sampling errors.

12.1 Sampling error

Sampling error is a result of estimating data based on a probability sampling
survey compared to the population. Such error in statistics is termed as Relative
Standard Error (RSE) and is expressed in percentage. This error is an indication
to the precision of the parameter under study. In other words, it reflects the
extent of variation of sample-based estimates compared to the parameter of
population.

For instance, in PES 2020, the under unumeration rate of population in
Malaysia....

Penilaian Liputan Banci 2020 191

Post Enumeration Survey 2020

Nota Teknikal
Technical Notes

12.2 Non-sampling error
The error may arise through incomplete survey coverage, weaknesses in the
frame, response errors, non-response errors and also errors during processing
either through editing, coding or data capture.
To ensure that data is of high quality, several efforts had been made to minimize
non-sampling errors. Intensive training was conducted for supervisors and
enumerator. In addition, close supervision and random checks were carried out
on households covered by the enumerators to ensure the validity of the recorded
information.
At the data processing stage, consistency checking and validation process has
been systematically implemented for each variable in order to minimize the non-
sampling error. It is generally accepted view that the magnitude of non-sampling
errors in the PES is much less than in the Census.

13. NOTES AND SYMBOL
- Nil/no cases
W.P Wilayah Persekutuan

Penilaian Liputan Banci 2020 192

Post Enumeration Survey 2020


Click to View FlipBook Version