2080.5 - Information Paper: Australian Census Longitudinal Dataset, Methodology and Quality Assessment, 2006-2011  
ARCHIVED ISSUE Released at 11:30 AM (CANBERRA TIME) 18/12/2013   
   Page tools: Print Print Page Print all pages in this productPrint All  
Contents >> 2. Data linking methodology >> 2.4 Blocking and linking strategy used in the ACLD

2.4 BLOCKING AND LINKING STRATEGY USED IN THE ACLD

After examining the information available and the quality of the data, a blocking and linking strategy for the ACLD was developed. This was based on the quality study previously undertaken by the ABS (Assessing the Likely Quality of the Statistical Longitudinal Census Dataset (cat. no. 1351.0.55.026)). The final strategy employed for linking the 2006 Census sample to the 2011 Census followed on from this work, but contained tighter quality controls and a more comprehensive approach than the two pass approach used in the original quality study. A key feature of the enhanced strategy was the identification and specific targeting of key sub-populations and trying to maximise their opportunity to be linked.

The main features of the enhanced blocking strategy used for the ACLD were:

  • more linking runs/passes
  • more restrictive blocks in order to maximise linkage quality
  • combination of deterministic and probabilistic linkage techniques
  • combination of clerical review and decision rules for filtering
  • targeted sub-populations in linking and clerical review, such as children, and Aboriginal and Torres Strait Islander people, to improve the quality of these sub-populations
  • using information about other members of the household to assist in linkage and clerical review in order to improve the overall quality of the final linked file.


Another feature of the new approach was the ability to be responsive to any under-represented sub-populations in the current linkage file. Accordingly, at the end of each pass, the remaining unlinked records were analysed to determine the best approach to be undertaken for the next pass. This allowed for blocking fields to be customised in order to broaden the search for the remaining unlinked records and for tolerances to be relaxed for some linking fields in later passes of the linkage.

Table 1 displays the blocking and linking fields applied in this linking project for each pass.

TABLE 1 - BLOCKING AND LINKING FIELDS, By pass number and method

LINKING METHOD

Deterministic(a)
Probabilistic
PASS NUMBER
1
2
3
4
5
6
7
8
9
10(b)
11
12
CENSUS FIELDS
Personal information
Age
B
B
L
L
L
L
L
L
L
L
L
L
Sex
B
B
B
B
B
B
B
B
L
L
L
L
Day and Month of Birth
B
B
B
B
B
. .
L
L
L
L
L
L
Indigenous status
B
B
B
B
B
L
B
. .
L
L
L
L
Birthplace
. .
. .
L
L
L
L
. .
. .
L
L
L
. .
Year of Arrival
. .
. .
L
L
L
. .
. .
. .
L
L
L
. .
Marital status
. .
. .
L
L
L
. .
L
L
. .
. .
. .
. .
Level of Qualification
. .
. .
L
L
L
. .
. .
. .
L
L
. .
. .
Field of Qualification
. .
. .
L
L
L
. .
L
L
L
L
L
. .
Highest year of Schooling
. .
. .
L
L
L
. .
. .
. .
L
L
. .
. .
Occupation
. .
. .
. .
. .
. .
. .
L
L
. .
. .
. .
. .
Religion
. .
. .
L
L
L
. .
. .
. .
. .
. .
. .
. .
Language spoken
. .
. .
L
L
L
. .
. .
. .
L
L
. .
. .
Aged less than 15 block
. .
B
. .
. .
. .
B
. .
. .
. .
. .
. .
. .
Household information
Mothers Age
. .
L
. .
. .
. .
L
. .
. .
. .
. .
. .
L
Mothers Day and Month of Birth
. .
L
. .
. .
. .
L
. .
. .
. .
. .
. .
L
Fathers Age
. .
. .
. .
. .
. .
. .
. .
. .
. .
. .
. .
L
Fathers Day and Month of Birth
. .
. .
. .
. .
. .
. .
. .
. .
. .
. .
. .
L
Family ID block
. .
. .
. .
. .
. .
. .
. .
. .
. .
. .
B
. .
Geographic information
Mesh Block
B
. .
B
. .
. .
. .
B
. .
. .
B
. .
. .
SA1
. .
. .
. .
. .
. .
. .
. .
B
B
. .
. .
. .
SA2
. .
B
. .
B
. .
. .
. .
. .
. .
. .
. .
. .
SA4
. .
. .
. .
. .
B
B
. .
. .
. .
. .
. .
B

(a) The variables used in the deterministic linking have been classified as blocking fields, however, these could also be classified as linking fields.
(b) The results of Pass 10 were used to identify the blocking field to be used in Pass 11. As a result, there were no records output from Pass 10.




Previous PageNext Page