Sources of secondary analysis for social work research

Sources of secondary analysis for
social work research
How to do an empirical dissertation
without any access problems
When dissertation research goes well
• The topic really interests you
• There is plenty of rich data
• Perhaps the research is of particular interest
to your employers or placement providers as
• You learn something new and feed this back
to local practitioners and/or service users
When problems arise
• Data collection is very difficult to achieve
• The host organisation (gatekeeper) says no –
research access is refused
• The host organisation says yes in principle but
very few individuals agree to take part in your
– This can lead to poor quality research
– Poor quality research could be seen as wasting
your research participants’ time
Why does this happen?
• Social workers are busy with their core tasks
and do not see student (or any) research as a
priority – understandably enough
• Social services departments have little
research culture
• There are so many layers of bureaucracy to go
through, it is impossible to make direct
contact with potential research participants
There are other options
• Archived data sets
– 1. Quantitative
– 2. Qualitative
• Research with publicly-available documents
– 3. Research on enquiries and policy documents
– 4. Research on research
– 5. Research on media
• Anonymised data from social work
1. Archived quantitative data
• You need some statistical skills, and ideally be able to do
multi-variable analysis
• There are some excellent freely-available datasets
• Locate studies via the UK Data Service Variable and Question
Bank: (for
starters, try typing “child abuse”)
• Explore data via Nesstar: There is free
access unless you want to do cross-tabulations, in which
case you need to register using your university ID
• This registration should also let you download data sets. You
will have to fill in a form for this stating what your purpose
Which studies?
• Search / browse and have a look
• Bear in mind there are relatively few variables about
social workers. Still plenty of potential for SW
• See Maxwell et al. (2012) for a list of cohort/panel
studies which do have social work variables, e.g.
– Millennium Cohort Study
– British Household Panel Study
• Also cross-sectional studies, e.g.
– British Social Attitudes Survey (annual)
– SN 5280 -Mental Health of Young People Looked After by
Local Authorities in Great Britain, 2001-2003
• Example: Cheung and Buchanan (1997)
2. Archived qualitative data
• Qualitative and mixed methods studies via the UK Data
• e.g. click on link ‘discover qualitative and mixed
methods data’ and then type “social work” (with
inverted commas)
• The archive includes classic studies such as Dingwall et
al.’s The Protection of Children and Townsend’s The Last
Refuge, as well as more recent studies.
• Bear in mind age and usability of data. Historical
research is fine, but documents may be harder to read.
• Many will be very large datasets.
• Example: Evans and Thane (2006)
3. Research on enquiries and policy
• These days, public enquiries publish evidence on-line.
• e.g. Victoria Climbie Data Corpus Online:
• This is just one example. Also look at government
reports where they publish evidence submitted and/or
responses to consultations
• Discourse analysis of policy documents relating to an
aspect of social work (e.g. Humphries, 1997)
• See Gibbs and Hall (2007) on enquiries and references
4. Research on research
• What kind of social work research is going on?
Method, topic etc.
• Who is doing it and where?
• Who is citing whom?
• e.g. Slater et al. (2013)
5. Research on media
• Media coverage of social work or of a specific
issue (e.g. mental health).
• For newspapers, use Lexis Nexis or similar
• Study of publicly-accessible social media – e.g.
using Twitter searches to follow debate on a
hot topic in social work.
• e.g. Henderson and Franklin (2007) on TV
Another option – administrative data
• Neither data which you yourself generate or data
which are publicly available
• Data which are already collected by agencies and
which can be made available to you as an
anonymised data set, so no permissions needed.
• Care needed to ensure real anonymity – e.g. DoB
(but you might need some kind of data on age)
• e.g. Winter and Connolly (2005)
Cheung, S-Y. and Buchanan, A. (1997) High malaise scores in adulthood of children
and young people who have been in care. Journal of Child Psychology and
Psychiatry 38: 575-580.
Evans, T. and Thane, P. (2006) Secondary Analysis of Dennis Marsden Mothers
Alone. Methodological Innovations Online 1(2) 78-82.
Gibbs, G.R. and Hall, C. (2007) The research potential of testimony from public
inquiry websites. Children and Society 21, 69–79.
Henderson L. and Franklin, B. (2007) Sad not bad: Images of social care
professionals in popular UK television drama. Journal of Social Work, 7(2), 133-153.
Humphries, B. (1997) Reading social work competing discourses in the Rules and
Requirements for the Diploma in Social Work British Journal of Social Work 27(5):
Maxwell, N., Scourfield, J., Gould, N. and Huxley, P. (2012) UK panel data on social
work service users. British Journal of Social Work 42 (1): 165-184.
Slater, T., Scourfield, J. and Sloan, L. (2012) Who is citing whom in social work? A
response to Hodge, Lacasse and Benson. British Journal of Social Work. 42 (8): 16261633.
Winter, K. and Connolly, P. (2005) A small-scale study of the relationship between
measures of deprivation and child-care referrals. British Journal of Social Work
35(6): 937-952

similar documents