3302.0.55.004 - Information Paper: Death registrations to Census linkage project - Methodology and Quality Assessment, 2011-2012  
ARCHIVED ISSUE Released at 11:30 AM (CANBERRA TIME) 18/09/2013  First Issue
   Page tools: Print Print Page Print all pages in this productPrint All  
Contents >> Linkage accuracy

LINKAGE ACCURACY

Not all links are matches. That is, even where name and address are available, not all pairs assigned in a statistical linkage exercise result in a record pair belonging to the same individual. While the methodology is designed to ensure that the vast majority of links are true, some are nevertheless false.

In the 2006 cycle, statistical methods based on tolerances set in the linkage and clerical review process were used to estimate the proportion of false links. The false link rate for total deaths was estimated as 1.2%. Corresponding rates by Indigenous status were not calculated (see Linking Census Records to Death Registrations, cat. no. 1351.0.55.030).

As previously noted, greater resources were directed to clerical review in 2011 than in 2006 and higher clerical cut-offs were set. Cut-off weights were based on outcomes from clerically reviewing a random sample of record pairs. The upper cut-offs were chosen such that the number of estimated false links in each pass was zero. Thus the sampling estimate of link accuracy would be 100%. In reality there will be false links because:

  • sample sizes were not large enough to detect the very small number of false links present
  • some record pairs would have been wrongly assigned as matches in clerical review–while the clerical review staff made decisions in a highly consistent manner, even the small degree of inconsistency observed may have led to some false links being assigned.
While the number of false links is not able to be quantified precisely, the proportion is expected to be very small.



Previous PageNext Page