Area-based measurements can be useful for a number of reasons. For example, when the LS does not have the individual-level data we require, the area-based value for that person's place of residence might be used as a proxy instead. Area measures are also used to inform the distribution of health service resources. Furthermore, they can be used to analyse the effects of an area's characteristics on health independently of, or in conjunction with, any individual-level effects.

The LS contains the following area-based socio-economic indicators:

• Carstairs score at Census ward level, mapped to individuals, for 1981 and 1991 (CARSCO8, CARSCO9). However, you may need to use more recently-calculated scores which have not at present been adopted by ONS but are available for 1971, 1981, 1991 and 2001; your CeLSIUS User Support Officer can advise you.

It is also possible for a user to add external area-based measures to the LS, such as Townsend scores or Breadline Britain indices, through the use of a lookup table. The mapping of values from the lookup table to individuals in the LS can be done by CeLSIUS staff down to the county district level, or by the ONS Data Custodian down to ward level, subject to confidentiality constraints. For links to deprivation indices, see the Geography Training Module.

There are some problems with the use of such area-based measures. The ecological fallacy is an important potential limitation. It results from the false assumption that inferences can be made about individual phenomena based on observations of groups. For example, by no means all people living in deprived areas are themselves deprived.

Also, areas are not internally homogenous: a population containing a mixture of deprived and less deprived households is likely to have a middle ranking score. Such mixed populations are more likely to occur in rural areas. Therefore area-based scores are likely to provide a better indication of deprivation in urban than rural areas.

There are also problems for areas with large institutional populations; the location of an institution in a particular area may have little effect on the circumstances of its inhabitants, since they are effectively insulated from the surrounding area, particularly so in the case of prisons.