Oxford eXperimental Laboratory is undertaking laboratory-based experiments (eg information-seeking tasks) on networked computers in two disciplines: Economics (interactive decision making) and Political Science (evaluating government information online).

  • Overview
  • People
  • Publications
  • News

The Oxford eXperimental Laboratory (OxLab) is an experimental laboratory for the social sciences, set up by Nir Vulkan (Said Business School, Oxford University) and Helen Margetts (OII) in 2006 with funding from the University of Oxford Fell Fund. A range of laboratory and field experiments are run from OxLab, from both economics and political science perspectives. Experiments include investigation into the effect of different information environments on collective action, the impact of different types of e-government provision on citizens’ information seeking behaviour, and deadline effects in auction design.

The massive growth in Internet-mediated interactions between societies, governments and commercial organisations of all kinds creates a need for innovative methods to research online activity. Laboratory-based experiments where subjects are brought in and incentivized (via cash payments) to participate in games or information-seeking tasks on networked computers are an excellent way to develop such methodologies. Such laboratories have been used by experimental economists for some time, but the great expansion in online social and commercial activity means that as well as being more central to Economics research they have growing utility across other academic disciplines, particularly sociology, computer science and political science.

What Contributes to the Success of e-Petitions?

Graphs of OxLab data

FIGURE EXPLANATION. LEFT: Successful petitions (green) are those which receive 500 or more signatures, therefore receiving an official response from the government: only 5% ever reach this point (95% are unsuccessful, shown in pink). We can see that successful petitions tend to collect the required 500 signatures quickly and early: often within the first day (green lines concentrated to the left). Data collected from http://www.number10.gov.uk. RIGHT: Laboratory experiments (OxLab, 2008) have shown that the number of existing signatories affects the probability that an individual will sign a petition: red bars indicate a high number of existing signatures and show a higher than average likelihood that someone will sign each of the six petitions; yellow bars indicate low numbers of existing signatories and show a lower that average likelihood for signatures. The petitions were presented in a laboratory setting. These graphs show that to be successful, a petition must collect many signatures relatively quickly (LEFT); the data also hint at a positive feedback loop and tipping point that may contribute to a petition’s ultimate success (RIGHT): once a petition reaches a certain threshold of signatures within a critical early period, it will tend to continue to collect them. More details: http://www.governmentontheweb.org/

Take part in our experiments!

We constantly seek volunteers for our experiments and you might be interested to know that we pay our participants! Experiments are simple and completely anonymous. They usually involve some form of decision making in an economic context or navigating the World Wide Web. To find out more, visit OXlab’s Online Recruitment System.

Support

This project has been funded by Oxford University Press’s Fell Fund.

John Fell OUP Research Fund

Conference papers

Reports

  • Dunleavy, P., Margetts, H., Bastow, S., Escher, T., Pearce, O. and Tinkler, J. (2007) Government on the Internet. Report by the Comptroller and Auditor General, HC 529 Session 2006-2007, 13 July 2007. London: The Stationery Office.
  • Dunleavy, P., Margetts, H., Goldchluk, S., Khan, M.K., Tinkler, J., Towers, E. and Escher, T. (2009) Department for Work and Pensions. Communicating with Customers. Report by the Comptroller and Auditor General, HC 421 Session 2008-2009. London: The Stationery Office.