Non-response and under-representation
in the Adult Social Care Survey

The project runs from June 2011 to March 2012, and has now been completed

Background

Transparency in outcomes: a framework for quality in adult social care was published in March 2011 and set out a new strategy for transparency, quality and outcomes in adult social care. An Outcomes Framework for social care, the Adult Social Care Outcomes Framework (ASCOF), was published within this report and is a key element of the Government’s approach to accountability. A number of measures included in ASCOF draw upon the national Adult Social Care Survey (ASCS) that local authorities conduct with a sample of their service users every year.

As the major source of social care-related outcomes information, the ASCS is an important resource but, as it currently stands, is not representative of all social care users. The ASCS does not include social care users (in residential or community‐based settings) who lack the capacity to consent to participate, and does not accommodate the use of proxy respondents (i.e. a person who provides responses on behalf of the survey recipient). Evidence from the user experience surveys, the predecessor to the ASCS, also indicates that the survey is likely to have relatively high levels of non-response, with significant variations in rates of non‐response across local authorities. To improve the representativeness of the dataset to ensure its continued usefulness as a resource for policy makers and other users (such as local authorities, providers and researchers), it is important to develop the ASCS so it is inclusive of people lacking the capacity to consent to participate. It is also important to better understand who is not responding to the survey, and whether the actions of councils have an effect on rates of response.

Aims and methods

The aim of this work was to provide useful analysis for the future development of the Adult Social Care Survey (ASCS) by

  • Understanding more about the factors affecting non‐response
  • Undertaking development work to address under‐representation
  • Reviewing the methods for gathering data from people who have difficulties in responding to the survey including those currently excluded, such as individuals who lack capacity, and investigate potential methods to address this, for example proxy questionnaires and observational methods.

There were two streams of work: the first was based on a literature review and explored methods for gathering data from people who have difficulties in responding to postal surveys, in particular those currently excluded from ASCS. The review included international literature which focused on methods that could be used to gather data from disabled people within the context of a survey.

The second stream used 2011 ASCS data to explore the individual‐ and authority‐level factors that influence non-response (e.g. variations in the way the survey is run by the different authorities, such as the use of incentives). The methods for overcoming the effects of bias due to non-response and their impact on the ASCS data and related ASCOF indicators were also investigated. Regression techniques, including multilevel models, were used to take account of the hierarchical nature of the data.

The results of the study should be valuable for the administration of future ASCSs by providing recommendations on how to minimise non-response, and the methods that can be used to minimise the impact of non-response by (a) providing recommendations on methods for minimising the impact of non-response in future ASCSs; and (b) providing ideas for methods to gather data from groups currently excluded from the ASCS.

Outputs

Strand 1: literature review

Nick Smith, Juliette Malley (2012) Understanding and addressing underrepresentation in a postal survey of social care users: Full Report, QORU working paper 2843, Canterbury / London.[Link]

Nick Smith, Juliette Malley (2014) Understanding and addressing underrepresentation in a postal survey of social care users: Summary, QORU working paper, Canterbury / London.

Strand 2: non-response analysis

Juliette Malley, José-Luis Fernández (2012) Patterns and effects of nonresponse in the English adult social care survey. Full report, QORU working paper 2841, Canterbury / London [Link]

Juliette Malley, José-Luis Fernández (2014) Patterns and effects of nonresponse in the English Adult Social Care Survey, journal submission in process.

Research Team

Juliette Malley, Nick Smith, Ann Netten, José‐Luis Fernández, Azusa Sato