Polls of representative public opinion provided the data for the analyses in this section of the example profiles. Two sources provided the polling data. Firstly, surveys commissioned for this study and secondly, responses to the Office of The Deputy Prime Minister's (ODPM) triennial best value household survey of 2004.
For the surveys commissioned for this study, the sampling frame comprised:
- 1,068 adults and 107 children living in the ten electoral wards; and
- 2,161 adults living in the region in which each CDRP area is located.
Polling took place between May and July 2005. Details of the polling data and a technical report describing the methodology are available from Ipsos MORI (external link).
In-depth research is frequently beyond the resources of many local agencies. Moreover, polls too often do not use representative samples and standard questions. Using standard questions enables comparison of local findings with those from national polls such as the British Crime Survey (BCS) (external link) or ODPM best value household survey. To provide value for money, surveys must be statistically valid, otherwise the findings are little more than anecdote and have doubtful value in properly guiding service planning.
GDI provides a basis for a local interpolation of findings from large-scale surveys such as the BCS. A GDI based analysis makes possible predictions about how local households might respond to BCS questions. Local agencies should consider using a predictive approach coupled with targeted testing to validate predictions. To plan and review service provision more effectively, local agencies would find this approach provides better value for money and greater statistical reliance.