Tracking Social Practices with Big(ish) data
1 like674 views
The document discusses the tracking of social practices using big data, highlighting the importance of understanding habits and routines in relation to energy consumption and other social behaviors. It explores various methods such as time-use surveys and technotraces, while addressing challenges like data quality and representativeness. The author presents empirical insights and methodologies for analyzing social practices to better inform sustainable practices.
Tracking Social Practices with Big(ish) data
- 1. Tracking Social Practices with Big(ish) data Dr Ben Anderson Sustainable Energy Research Centre, Faculty of Engineering & Environment www.energy.soton.ac.uk 26th June 2014 @dataknut
- 2. @dataknut: Tracking Social Practices with Big(ish) data #pbes Contents § Background – Practices – the view from here § Tracking them down – TimeTraces – TechnoTraces § Challenges 2
- 3. @dataknut: Tracking Social Practices with Big(ish) data #pbes Contents § Background – Practices – the view from here § Tracking them down – TimeTraces – TechnoTraces § Challenges 3
- 4. @dataknut: Tracking Social Practices with Big(ish) data #pbes So what are practices? a temporally unfolding and spatially dispersed nexus of doings and sayings Schatzki, 1996 ‘habits’, ‘bodily and mental routines’ ‘permanent dispositions’ Reckwitz, 2002; Entities Performance habituation, routine, practical consciousness, tacit knowledge, tradition Performance often neither fully conscious nor reflective Warde, 2005 Why people don’t do what they ‘should’ - Jim Skea, 2011 Embodied habits & competencies (skills), Meanings/ conventions (image) Material artefacts (stuff) Shove & Pantzar, 2005
- 5. @dataknut: Tracking Social Practices with Big(ish) data #pbes Can we observe them? (an empiricist’s response) Image: Anthony B. Wooldridge Image: Eric Shipton “The recurrent enactment of specific practices leaves all sorts of “marks” – diet shows up in statistics on obesity; heating and cooling practices have effect on energy demand, and habits of laundry matter for water consumption. Identifying relevant “proxies” represents one way to go.” ESRC Sustainable Practices Working Group (SPRG) Discussion Paper, 2011
- 6. @dataknut: Tracking Social Practices with Big(ish) data #pbes § Tried: • Shadowing/tracking/observation – Small n, can ask why, investigator effects (?) – Historical? • Time use surveys (diaries, e.g. UK ONS 2000, MTUS) – Big n, non response issues, can’t ask why, complex data – Rarely longitudinal, sometimes historical (MTUS) § Relatively Untried: • Expenditure Surveys – Big n, proxies for practices, can’t ask why, complex data – e.g. http://link.springer.com/article/10.1007/s11269-012-0117-y • TechnoTraces (Savage & Burrows, 2007; 2009; 2014) – Transactions/meters/bills, proxies for practices, complex data, difficult to process How to detect ‘marks’ & proxies?
- 7. @dataknut: Tracking Social Practices with Big(ish) data #pbes Contents § Background – Practices – the view from here § Tracking them down – TimeTraces – TechnoTraces § Challenges 7
- 8. @dataknut: Tracking Social Practices with Big(ish) data #pbes Time Traces § Large sample time-use surveys 8 0 10 20 30 40 50 60 70 80 90 100 06:00 08:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 0:00:00 Time % Phone/email friends Travel Computer Hobbies/other Going out Friends/Family at home Sport/exercise Reading TV/radio shopping adult care child care civic acts education work housework eating/drinking washing sleeping Data: % of sample reporting activity Source: ONS 2005 UK Time Use Survey, all 16+
- 9. @dataknut: Tracking Social Practices with Big(ish) data #pbes Time Traces § Large sample time-use surveys 9 Credit: Mathieu Durand-Daubin (EDF R&D) drawing on INSEE (2012) “Le temps de l’alimentation en France”
- 10. @dataknut: Tracking Social Practices with Big(ish) data #pbes Time Traces § Large sample time-use surveys § Over time – E.g. Laundry 10 0.00% 0.20% 0.40% 0.60% 0.80% 1.00% 1.20% 1.40% 04:00 05:30 07:00 08:30 10:00 11:30 13:00 14:30 16:00 17:30 19:00 20:30 22:00 23:30 01:00 02:30 Sunday 1974 2005 0.00% 0.20% 0.40% 0.60% 0.80% 1.00% 1.20% 1.40% 04:00 05:30 07:00 08:30 10:00 11:30 13:00 14:30 16:00 17:30 19:00 20:30 22:00 23:30 01:00 02:30 Monday 1974 2005 Data: % of reported laundry being done at given time Source: Multinational Time Use Survey Dataset (UK, 1974-2005, all 18+)
- 11. @dataknut: Tracking Social Practices with Big(ish) data #pbes Cooking 0% 2% 4% 6% 8% 10% 12% 14% 16% 00:00 01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:00 09:00 10:00 11:00 12:00 13:00 14:00 15:00 16:00 17:00 18:00 19:00 20:00 21:00 22:00 23:00 %respondents(weighted) UK Italy Germany Norway Bulgaria Time Traces § Large sample time-use surveys § Over time – E.g. Laundry § Internationally 11 § Source: E-living Survey (2002) n ~= 1100 per country (Norway, UK , Bulgaria, Germany, Italy)
- 12. @dataknut: Tracking Social Practices with Big(ish) data #pbes Contents § Background – Practices – the view from here § Tracking them down – TimeTraces – TechnoTraces § Challenges 12
- 13. @dataknut: Tracking Social Practices with Big(ish) data #pbes TechnoTraces: Practice Hunting § Inspiration: – Qualitative study of telephone calling – Lacohee & Anderson (2000) Interacting with the telephone, doi:10.1006/ijhcs. 2000.0439 § Call types – Duty calls: generally to family members and were made because the caller felt a sense of duty to keep in touch – Maintenance calls: real motivation was to maintain a friendship – Grapevine calls: a series of calls often prompted by a call e.g. passing on news – Batch calls: making a series of outgoing calls e.g. cheap rate, bored or lonely § Question: can we identify them in a call records dataset? – c. 1.5 million incoming/outgoing phone call records (time, duration) linked to surveys of c 1000 GB households 1999-2001 13
- 14. @dataknut: Tracking Social Practices with Big(ish) data #pbes TechnoTraces: Practice Hunting § Algorithm: – Sequence identifier – Flexible ‘gap’ parameter § Batch calls (Out, Out, Out…) – 20:00 -> late – Not Thursdays or Fridays – Sunday evenings § Grapevine calls (In, Out, Out…) – 18:00 – 19:30 – Sunday evenings 14 Source: BT HomeOnline Survey (2000), n calls ~= 1.5 million from c. 310 households http://repository.essex.ac.uk/2294/ Data processing by Dr David Hunter (ECS, University of Essex)
- 15. @dataknut: Tracking Social Practices with Big(ish) data #pbes TechnoTraces: Practice Hunting § Contrasts § Requires – ‘Labeled’ data 15 Source: BT HomeOnline Survey (2000), n calls ~= 1.5 million from c. 310 households http://repository.essex.ac.uk/2294/ Data processing by Dr David Hunter (ECS, University of Essex)
- 16. @dataknut: Tracking Social Practices with Big(ish) data #pbes TechnoTraces: Applied to energy? § Contrasting gas consumption 16 Source: EPSRC DANCER Project baseline gas consumption monitoring - http://www.dancer-project.co.uk/ § Gas consumption per 5 minutes, identical dwellings in South East UK, same street, both couples with 3 children, male partner working, female partner not § December 2012 – February 2013
- 17. @dataknut: Tracking Social Practices with Big(ish) data #pbes Using linked mixed methods? § E.g. TechnoTraces & TimeTraces! 17 Electricity Source: Small scale energy diary and consumption monitoring study lead by Kathryn Buchanan, University of Essex http://www.dancer-project.co.uk/ GasElectricity
- 18. @dataknut: Tracking Social Practices with Big(ish) data #pbes Contents § Background – Practices – the view from here § Tracking them down – TimeTraces – TechnoTraces § Challenges 18
- 19. @dataknut: Tracking Social Practices with Big(ish) data #pbes ‘Big’ Data Challenges § Provenance: – Who did what to ‘my’ data? § Quality: – It’s never clean § Samples – What (or who) does it represent? § Sampling – Do we really need it all? § Linkage – Multiple methods & multiple views 19 It might be big but is it clever? Are people the only agents? And the bigger it is the harder to clean What & why?
- 20. @dataknut: Tracking Social Practices with Big(ish) data #pbes Thank you § Questions? – b.anderson@soton.ac.uk – @dataknut § http://www.energy.soton.ac.uk/ 20