CONCLUSIONS AND RECOMMENDATIONS
Disability is an extremely complex and difficult concept to measure. It is especially difficult to capture the full complexity of disability experience and accurately assess the full range of disability severity in a small number of questions. This is precisely why the SDAC, which is the most comprehensive source of disability information in the ABS, has an extensive question set. Since 1992, the Short Disability Module has been used, in some form in other surveys, to provide insight into aspects of life that the SDAC does not examine. This manages the survey burden on respondents. For many areas of social interest, the Short Disability Module has been used to demonstrate how the characteristics of people with disability and their families are different from people and families without disability. Similarly, the questions on the Census, while not providing a complete picture of disability, are able to inform the provision of disability support services to subpopulations and smaller geographic areas than available from the surveys.
The question remains then - is the Short Disability Module measure useful for users of disability data? It is, as it increases the range of data available for the population with disability, as long as users are aware of the limitations of the measure. If the measure is cross-classified with other data items within the same survey, it will produce results that broadly describe the differences and similarities between people with disability and those without disability for that particular characteristic. For example, the inclusion of the Short Disability Module on the NHS enables a comparison of the health actions taken by those with and without a disability. It is likely that the relationship between disability and the characteristic of interest may be somewhat 'diluted' because of the inclusion of people who were not disabled in the overall measure of disability i.e. differences between the disabled and non-disabled populations will be less marked in surveys using the Short Disability Module than in the SDAC if using this measure.
In short, it should be noted that the SDAC produces the most conceptually accurate measure of disability and that the Short Disability Module includes some people who have a long-term health condition but who are not disabled in the measure of disability using the conceptual frame applied to the SDAC.
Disability rates from the Short Disability Module in different surveys have varied both between surveys and from the SDAC. It is probable that some of the variability relates to the method of collection, conceptual or wording differences, or to the rapport established between respondents and interviewers in longer interview times (Kalton & Schuman, 1982). In addition, some disabilities by their nature are variable in their impact on people. Also, respondents may be struggling personally at the time of interview which may affect their willingness to identify conditions or discuss issues.
In assessing the qualities of SDAC and the Short Disability Module in other surveys, the following key points are made:
The Short Disability Module is not recommended for measurement of prevalence of disability.
The analysis in this paper has shown that the total population with disability identified by the Short Disability Module is markedly different to the total disability population identified in the SDAC. The Short Disability Module identifies a larger population with disability by including some people with a long-term health condition without disability. However, when compared to the broader long term health condition identified by the SDAC, the Short Disability Module identifies neither the whole population with a long term health condition nor the population with a disability, but something between the two, with an emphasis on including people with long-term health conditions that impact on sensory and physical capacity. Therefore it is concluded that disability measured in surveys using the Short Disability Module should not be used to compare or update counts or prevalence rates.
When using the Short Disability Module, be aware that the social outcome differences between those with a disability and those without disability in the broadest measures may be underestimated.
Another implication for data users is that surveys using the Short Disability Module will probably show disability population characteristics that are somewhat more like the general population than the disability population as described in SDAC. That is, it is likely the surveys other than SDAC will show higher labour force participation, higher income, higher educational attainment and have a younger age distribution for people with disability than the SDAC. Thus, if disability is related to a particular characteristic, the broader disability group identified by the Short Disability Module will display a weaker relationship to that characteristic. There is still sufficient differentiation in the Short Disability Module between people with disability and those without disability, to be of value to users and to warrant the inclusion of the Module in future surveys, but the differences may be underestimated.
Measuring subpopulations with disability using the Short Disability Module.
There are some measures available from the Short Disability Module that are likely to be more comparable with SDAC than others. The measures from the surveys using the Short Disability Module that most consistently align with those from the SDAC, and thus are best for comparing estimates of people with disability and people without disability, are the 'Specific limitation or restriction' measure for younger people (<60 years) populations and the combined 'Profound/severe core activity limitation' measure. For users with an interest in broad disability populations, the 'Specific limitation or restriction' category will provide a broad and inclusive measure, yet still be comparable to the SDAC. It is preferable to the 'Disability or long term health condition' measure because it does not include large numbers of people with a non-restricting long-term health condition.
For users with an interest in the population that needs assistance, the combined 'Profound or severe core activity limitation' measure is recommended. The profound and severe categories are of particular value to users because they refer to those people with disabilities who need help with core activities. The most reliable way to analyse the characteristics of these people is to combine the categories into one: 'Profound or severe core-activity limitation'. Within this group, there are generally no major differences between the SDAC and the Short Disability Module for people under 70 years of age. For people in the Short Disability Module over 70 years of age, the pattern of higher prevalence with older age is the same even though the population identified is slightly smaller than the SDAC. It should be noted however, that the measure of profound/severe disability in the Personal Safety Survey is different to the other surveys using the Short Disability Module. Proxy interviews were not used in this survey and some people who could not answer for themselves will therefore be missing from the estimates.
Other categories from the ‘Severity of Disability’ measure should only be used with caution because of significant discrepancies with the comparable data from the SDAC. These are: 'No specific restriction or limitation', 'Education/employment restriction only' and 'Mild core-activity limitation'. The 'No specific restriction or limitation' and 'Education/employment restriction only' measures are significantly overstated compared with the SDAC, and the 'Mild core-activity limitation' is significantly understated compared with the SDAC.