“Big data” initiatives that aim to bring together and mine data from multiple databases across government and non-government agencies promise new insights into human service delivery. Specifically they aim to provide information about what services are being used, how, by whom and with what outcome. However, the process of achieving such insights poses both practical problems and ethical dilemmas. In this presentation, drawing from an extensive literature review and research with government and non-government human service organisations focussing on the design and redevelopment of electronic information systems, the most significant problems and dilemmas will be explored. It will be argued that current frameworks for ethical social work and human service practice will need to be expanded to accommodate developments in technology which have made ‘Big data’ projects possible.
“Big data” in human services organisations: Practical problems and ethical dilemmas.
1. “Big data” in human services
organisations: practical
problems and ethical dilemmas
Dr Philip Gillingham and Mr Timothy Graham
University of Queensland
2. Defining Big Data
Size of datasets.
Complexity or structure of datasets.
Tools used to analyse them.
Aim is to discover new patterns within datasets and new associations
between variables – targeting services, evidence informed practice.
3. Purpose of presentation
• Hot topic
• Longer history in health (health
informatics)
• Starting to be used in social work
(social informatics)
• Those involved dealing with the
challenges
• Should be debated more widely
• Limited selection
Google Trends – popularity of search term “big data” over time.
4. Big Data – Practical problems (1)
Sampling
• Big Data means bigger samples……n=all?
• Data integrity. Incomplete case files.
• Subjective interpretations of the social world.
5. Big Data – Practical problems (2)
Data extraction
• ‘Wrangling’ – preparation of data for
analysis.
• Series of decisions about what to include
and exclude.
6. Big Data – Practical problems (3)
Data analysis
• algorithms constructed by people.
• Political decisions about what a 'social
problem’ is.
• Big Data, despite claims of objectivity,
not immune.
Wizard of Oz – “the man behind the curtain”.
7. Big Data – Practical problems (4)
Data interpretation
• Apophenia – patterns or associations where none exist
• De-individualisation – group traits assigned to individuals to predict
behaviour
8. Big Data – Practical problems (4)
Data interpretation
• Lack of explanation – patterns don't explain why a social problem
exists
• Numbers speak for themselves?
9. Big Data – ethical problems (1)
• Consent
• Secondary use of data
10. Big Data – ethical problems (2)
• Privacy – individuals identifiable through
level of detail
• Who owns the data? Agency,
researchers, service user?
11. Finally
• A full article is under review, stay in touch if interested.
p.gillingham@uq.edu.au
12. Selected References
Bollier, D. (2010) The Promise and Peril of Big Data. Accessed 17 Feb 2014 at:
http://www.aspeninstitute.org/sites/default/files/content/docs/pubs/The_Promise_and_Peril_of_
Big_Data.pdf
boyd, d. and Crawford, K. (2012) Critical questions for Big Data: Provocations for a cultural,
technological and scholarly phenomenon. Information, Communication and Society, 15(5), pp. 662-
679.
Crawford, K. (2013) The hidden biases in Big Data. Harvard Business Review (HBR) Blog Network (1
April). Accessed 29 November 2013 at http://blogs.hbr.org/2013/04/the-hidden-biases-in-big-
data/
Desrosières, A. (2002) The politics of large numbers. Translation by C. Naish. Harvard University
Press.
Foster, V. and Young, A. (2012) The use of routinely collected patient data for research: a critical
review. Health, 16, pp. 448-463.
Gutierrez, D. (2014) The unsexy part of data science: data munging. Accessed 17 Feb 2014 at
http://inside-bigdata.com/2013/11/15/unsexy-part-data-science-data-munging/
13. Selected References - continued
Henman, P. (2005). E-government, targeting and data profiling: policy and ethical issues of
differential treatment. Journal of E-Government, 2(1), pp.79-98.
Keen, J. , Calinescu, R. , Paige, R. and Rooksby, J. (2013) Big Data + Politics = Open Data: the case of
health care data in England. Policy and Internet, 5(2), pp. 228-242.
Naccarato, T. (2010) Child welfare informatics: a proposed subspeciality for social work. Children and
Youth Services Review, 32, pp. 1729-1734.
Nguyen, L. H. (2007) Child welfare informatics: a new definition for an established practice. Social
Work, 52(4), pp. 361-363.
Parker-Oliver, D. and Demiris, G. (2006) Social work informatics: a new speciality. Social Work, 51(2),
pp. 127-134.
Varian, H. R. (2013) Big Data: new tricks for econometrics.
http://people.ischool.berkeley.edu/~hal/Papers/2013/ml.pdf
Ward, J. S., and Barker, A. (2013). Undefined by data: A survey of big data definitions. The Computer
Research Repository. http://dblp.uni-trier.de/db/journals/corr/corr1309.html#WardB13a
Philip Gillingham and Tim Graham, University of Queensland 13