Annex C1: Note on 'Delphi' methodology
The research methodology on which the two scenarios are based was developed in precise stages. It draws on the experience we have gained from doing similar studies over the years. We began by establishing a panel of experienced individuals, chosen for their broad understanding of the complexities and dynamics of Scotland's public service provision. This panel was made up of over 1000 people working across the public services - including representatives from central and local government, health, education and the police - as well as from representatives of the voluntary and private sectors.
Stage one: Establishing the panel
OPM approached approximately 450 senior individuals from a wide range of organisations all over Scotland. We explained the objectives of the survey as well as the proposed process: a modified Delphi survey combined with a 'cluster analysis' technique. The nominations were later added to by combining Care 21 nominations and a number of mailing lists. A 'panel' of over 1700 people was subsequently established.
Stage two: First-round questionnaire
At the beginning of December 2004, a first round questionnaire was sent to the panel members. They were asked to predict three to five external changes which they believed would have a significant impact on unpaid care over the next decade. We asked them to be as detailed as possible in their responses, particularly on the direction, level and impact of the changes. The questionnaire can be found in annex C2. Approximately 380 people responded to this first-round questionnaire.
Together, they gave us over 2000 individual predictions. These open-ended responses, or 'change statements', were the only source used to construct the second-round questionnaire. They were also used to provide specific detail in the three scenarios.
Stage three: Second-round questionnaire
The second-round questionnaire was sent to the panel in March 2005. This questionnaire aimed to measure similarities and differences in the panel member's opinions and to provide us with more information for the scenarios. Specific questions covered changes that might possibly occur in public services and the general environment over the next 10 years. Using a scale from one to five, the panel members were asked to indicate what they believed the likelihood was of each change occurring by the year 2014. The questionnaire can be found in Annex C3. The response rate to the round-two questionnaire was 22 % per cent, typically high for a survey of this nature. The responses were analysed using a statistical cluster analysis technique with the statistical IT tool, SPSS. This technique measured the similarities and differences in the responses and grouped the respondents according to their propensity to agree. The analysis identified three relatively evenly spread clusters of predictions amongst the panel. These are referred to as clusters one, two and three. We did not tell panel members which cluster they belonged to.
Stage four: Constructing the scenarios
The responses from the second-round questionnaire were used separately to construct the scenarios. In each case, the change statements derived from the questions were collated into several categories:
- Political overview
- Economy and workforce
- Public sector
- Health and social care
- Carers
- Society and lifestyles
We used the crude results from the questionnaire to develop a profile of each predicted
future and then used these profiles to write up the narrative accounts. The point of these narratives was to make each scenario more impressionistic than the crude survey would have read and so, to help the scenarios 'come alive', we asked Care 21 representatives to comment on them.
Finally, we named the scenarios: 'Prosperity for some', 'Inflexible cohesion' and 'Choice and division'.