Reliability

Estimates based on sample surveys are subject to error. The main factor that determines the width of the margin of error around a given estimate is the number of sample cases it is based on. Errors on estimates involving some form of measurement of individuals e.g. time taken off work are also affected by the variability of the measure from person to person. Aspects of the survey design also affect all errors, in particular whether the sample is stratified or clustered. The LFS is both stratified (by post code sector) and clustered (at household level).

The sampling errors in this report are expressed as 95% confidence intervals (denoted 95% C.I. in tables), which mean that each range has a 95% chance of containing the true value (i.e. the value that would have been found if the entire population had been surveyed) in the absence of bias. Confidence intervals should be quoted in preference to the prevalence or incidence central estimate or rate whenever there are less than 30 sample cases i.e. figures shown in italics within the tables. In order to reflect some of the variability in the days lost estimates (measure from person to person) as well as the sample numbers involved, confidence intervals should be quoted for days lost estimates and rates based on fewer than 40 cases, also shown in italics. Estimates are not included in the tables where sample numbers are too small (less than 20 sample cases) to provide reliable estimates.

Useful definitions

Prevalence:

The estimated number of people with a work-related illness at any time during the 12 month reference period. It includes the full range of illnesses from long standing to new cases. The rate is defined as the prevalence estimate divided by the population at risk of having a work-related illness. See section 2.7 and Appendix 3.2 in the 'Self-reported work-related illness in 2004/05' [PDF 2mb] (SWI04/05) report for further details.

Incidence:

The estimated number of new cases of work-related illness occurring in the 12 month reference period i.e. people first becoming aware of their illness in this 12 month period. The rate is defined as the incidence estimate (restricted to individuals working in the 12 month period) divided by the population at risk of experiencing a new case of work-related illness during the reference period. See section 2.8 and Appendix 3.2 in the 'Self-reported work-related illness in 2004/05' [PDF 2mb] (SWI04/05) report for further details.

Working days lost:

Days off work due to work-related illness. The figures are expressed as full-day equivalent days, to allow for variation in daily hours worked. See section 2.9 and Appendix 3.2 in the 'Self-reported work-related illness in 2004/05' [PDF 2mb] (SWI04/05) report for further details.