Chapter Two: Research Methodology and Analysis
This chapter sets out the research methodology for both studies and provides details of data collection instruments used, sampling procedures employed and analysis undertaken.
The research comprised two main components, a nationally representative survey of 1,015 adults across Scotland and a series of 8 focus groups conducted in various locations across the country. This report provides details of the findings from both studies.
The National Survey
The survey comprised a section of questions on MORI's Social Policy Monitor. This is a quarterly research vehicle, designed for organisations to collect regular, robust data on the characteristics and opinions of the Scottish population.
The questionnaire was developed in close consultation with the CYPSC. It comprised 18 questions and was approximately 10 minutes long. Topics covered in the questionnaire included:
- Rights to social welfare and responsibility for welfare provision;
- Thresholds for social work intervention;
- Knowledge of the roles of social work services;
- Sources of information about social workers and social work services;
- Attitudes to social workers and social work services;
- Use and potential use of social work services;
- Satisfaction with social work services; and
- Future need and priorities for social work services
The first two topic areas included a number of scenarios which comprised hypothetical cases which social work services could be involved with. Similar scenarios have been used in other surveys, such as the Public Attitudes to Community Care in Scotland survey (Scottish Executive, 2001). These scenarios were used to gauge respondents' attitudes towards the level of involvement the state should have in its citizen's lives and at what stage social workers should intervene in a case. Early intervention was explored in greater detail in the qualitative research. A copy of the CYPSC questionnaire is appended to this report.
The survey was undertaken with a nationally representative sample of Scottish adults aged 16+ from 18 April - 19 th June 2005. The survey was conducted face-to-face in home, using CAPI (Computer Assisted Personal Interviewing). Households were sampled as follows:
- The sample was drawn from PAF (small user file);
- All output areas in Scotland were selected (including the Western Isles, Orkney and Shetland) and sorted into Local Authority area;
- These were then ranked by Scottish MOSAIC2 indicators;
- Output areas were then sampled proportional to the population in these areas;
- The final sample comprised of 70 output areas;
- In each output area, 25 addresses were selected at random from the whole list of addresses. This was to account for an average deadwood rate of 9%, a response rate of 65% and an anticipated 1,000 interviews;
- An advance letter was sent out to all sampled addresses before the survey fieldwork began to provide householders with brief details of the survey and how to contact MORI;
- Interviewers had to make at least six calls to sampled addresses, including at least one in the evening and one at the weekend;
- Interviews could only be conducted with an adult aged 16+ which was selected at random by the CAPI machine;
A total of 1,015 interviews were achieved from 1,750 allocated addresses. This gives an adjusted response rate of 64%.
All fieldwork was conducted using Computer Assisted Personal Interviewing ( CAPI) where data is inserted into hand held computers by MORI interviewers. The main strength of CAPI is that interviewers do not need to check questionnaire routing as the paper survey is pre-programmed using In2Quest software, which leads to improved data quality and avoids the need for a separate data entry process. It also facilitates versioning of the questionnaire and the rotation of question items.
Before the data could be analysed certain weights had to be applied so that the data would be representative of the population. The data was weighted by Local Authority, age and sex. The weights were derived from the 2001 Census. The effect of weighting can be found in the sample profile appended to this report.
Computer tables were prepared to facilitate reporting. Each question in the CYPSC module was analysed by a number of key variables, namely:
- Age (16-24, 25-34, 35-44, 45-54, 55-64 and 65+);
- Employment status (employed, unemployed, unable to work and retired);
- SIMD3 (Scottish Index of Multiple Deprivation) (20% least deprived areas and 20% most deprived areas);
- Urban/Rural classification;
- Tenure (owner/occupier, rent, other);
- Social Class - based on National Statistics Socio-Economic Classification 4 ( NS- SEC) (8 groups);
- Users/non-users of social work services;
- Perceived role of the Government; and
- Overall impression of social workers.
All survey results are subject to sampling variability which means that observed differences between sub-groups may not always be statistically significant i.e. they may have occurred by chance. A guide to statistical reliability is appended to this report.
Where percentages do not sum to 100%, this may be due to computer rounding, the exclusion of 'don't know' categories or multiple answers. Throughout the report, an asterisk (*) denotes any value of less that half a per cent.
Focus Group Research
CYPSC commissioned MORI Scotland to conduct a series of focus groups into public knowledge of, and, attitudes towards, social work in Scotland. These were designed to throw light on the survey findings, and enable the research team to explore issues that could not be probed in a quantitative survey because of the structured nature of the approach.
Recruitment of the focus groups
Eight focus groups in total were conducted between 16 th and 24 th May 2005. Six of the groups were conducted in urban areas and two in rural areas. The focus group research comprised adults 18+ from a range of socio-economic backgrounds. Quotas were set by class, age and ethnicity, in order to identify any variation in the views of different groups of people. Each focus group comprised a fairly homogenous set of people to help them to feel more comfortable with one another, in turn improving the quality of the information collected.
A recruitment questionnaire was designed so that only people who satisfied these quotas were invited to participate in the focus groups. Participants were recruited face to face from the general public between 4 th and 13 th May 2005. All but one of the groups was recruited by experienced MORI interviewers. The Chinese group was recruited on MORI's behalf by the Chinese Community Development Project. Table 2.1 overleaf sets out the composition of each group and provides details on attendance.Table 2.1: Focus group composition
Focus group composition
The focus group research was designed to supplement findings emerging from the quantitative survey. Interim data from MORI's Social Policy Monitor was used to inform the topic guide design. Interim data was used to provide additional insight into how the public's knowledge of, and attitudes towards social workers was taking shape.
The main themes discussed in the focus group sessions included:
- General perception of social work and social workers;
- Knowledge and understanding of social work and social work services;
- Current and anticipated use of social work services;
- Attitudes towards intervention;
- Desired and acceptable thresholds for risk and early intervention; and
- Priorities for social work services in the future
The main body of this report is structured around these key themes.
Interpretation of Qualitative Data
Two of the key strengths of qualitative research are that it allows issues to be explored in detail and enables researchers to test the strength of people's opinions and the underlying rationale for people's attitudes and behaviours.
However, it needs to be remembered that qualitative research does not allow conclusions to be drawn about either the extent to which something is happening among the wider public (although one might surmise that particular opinions appear to be widespread) or percentages of people that have certain attitudes or opinions. Qualitative research is designed to be illustrative rather than providing statistically representative data.