This page provides an overview of the wide variety of Labour Force Survey (LFS) quality measures which are available.
Additionally, the Office for National Statistics (ONS) has a programme of work which will be recommending how best to report quality measures in the future as part of its modernisation programme.
For further information see related links for detailed documentation and reports which are currently available.
Details of sampling errors are published each month in the Labour Market Statistics Integrated First Release.
Further information is also given in the LFS Quarterly Supplement and the LFS User Guide Volume 1.
ONS have expanded the range of sampling errors available. First results have been published on the National Statistics Website.
Response rates are available each quarter in the LFS 'Performance and Quality monitoring Report (PQM)'.
This report covers overall response, question specific, proxy and wave specific response and response by government office region.
The LFS has to complete fieldwork to a tight timetable and interview as many sampled households as possible.
Interviewers try to interview every adult in the sampled household, however this is not always possible to tight time scales and they accept information by proxy from another responsible adult in the household.
This naturally raises the issue of whether proxy responses are sufficiently accurate. Research carried out into the impact of proxy response error forms section 11 of the LFS User Guide volume 1.
Imputation of missing responses
The LFS rolls forward responses from certain types of non-responders who responded at the previous wave.
In particular, such imputation is done for non-contacts and 'circumstantial refusals' who it is felt could respond again in the future.
This procedure aims to reduce non-response bias in the cross-sectional estimates while improving precision by boosting the sample size.
Section 12 of the LFS User Guide volume 1 contains analyses carried out in this area.
The National Statistics LFS quality review recommended further investigation of the effect of imputation on estimates of change.
The biggest problem for non-response is to assess the bias it actually leads to.
A Census Non-Response study (CNR) was carried out which used the results of the 2001 Census and compared these with LFS responses of the same time period to assess the scope of non-response bias and investigate ways in which improved weighting systems can further reduce any such bias.
An initial report was completed out at the end of 2003. A report of the CNR conducted after the 1991 Census can be found in Section 9 of the LFS User Guide volume 1.
Information on a wide variety of LFS quality measures are contained in the LFS User Guide Volume 1.
This volume covers sample design, fieldwork and processing procedures and details of non-sampling errors; weighting; seasonal adjustment and discontinuity.
Labour Market Trends articles of relevance to LFS quality
The former ONS publication Labour Market Trends contained articles concerning different aspects of LFS quality, including:
'Methodology for the 2001/02 annual local area LFS data (301.2 Kb Pdf) ' Jan 2003
'The new ethnicity classification in the LFS (93.8 Kb Pdf) ' Dec 2002
'Review of the LFS (72.5 Kb Pdf) ' Oct 2002
'Effect of the introduction of SOC2000 on employment estimates (188 Kb Pdf) ' Sep 2002
'Hours worked: a comparison of estimates from the LFS and the NES (263.5 Kb Pdf) ' Aug 2002