Key Check List for Setting Up an Experimental Lab in a University Without A Tradition in This Area

Sunday, July 10, 2011

Experimental economics is a growing area but there are still plenty of universities without a laboratory. What are some key things that one needs to look out for if one is considering setting up an experimental laboratory at a university without one and without a tradition of one?

Infrastructure

Space and hardware: A networked computer laboratory with a sufficient number of desks (min 20) with the possibility of partitions between terminals, ideally such that subjects cannot see each other. The laboratory should be of exclusive or primary use for experimental research purposes.  'Primary' means that any request for non experimental use should go through the experimental lab manager.

Software: no special costly programming software is required. The standard experimental software (ZTree) is free and works well with Windows 7, though problems may occur for at least some versions of Windows NT. A standard recruitment software (ORSEE) is also free.

Personnel

A research officer runs the laboratory. There are aspects of his/her role that could be done by a technician. Administering and supervising the implementation of experiments does not fall in this category.

HR procedures should enable casual research assistance (e.g., at UEA, for up to 3 months contracts, there is no need to advertise etc.). There should also be a culture enabling appropriate individuals (usually though not necessarily research students) to take advantage of casual research assistance opportunities.

Training will clearly be an important part of creating an environment where experiments can be conducted by appropriate research officers/assistants and by research students.

Procedures

Smooth procedures are crucial for the success of an experimental laboratory, and the failure of any one of them is likely to lead to the straight unfeasibility, or significantly inefficiency relative to the competition, of engaging in experimental research:

Ethical approval: There are two aspects to this. First, an ethical scrutiny which is (a) compliant with the appropriate regulations (in the UK the ESRC ones being the most relevant ones), (b) aware of the appropriate ethical conventions in experimental economics (as opposed to, say, social psychology) and (c) capable of identifying problem cases is obviously desirable and necessary. Second, routine professional experimental economics research easily meets the standard ESRC requirements and unnecessary delays need to be avoided when this is the case; while there are different ways of approaching this, the one I would recommend is to devise is a two stage approach in which any piece of research needs to check a standard check list and assuming there is nothing that requires attention from the light touch approach, this is signed off by the appropriate ssREC (subject specific Research Ethics Committee, using UK terminology) Chair or her/his deputy – otherwise a full ethical review will take place -.

Recruitment: a professional online recruitment system (such as ORSEE) would be set up and managed primarily by the research officer. You would need to be allowed to send _a minimum_ of one email to all the student population to invite them to join the relevant mailing list, but ideally more than one as this would be highly restrictive. Appropriate university generic email accounts would also need to be set up to facilitate recruitment.

Cash in advance: it should be possible to collect cash in advance (as at UEA) for experimental payments. Otherwise one is asking researchers, or research students, to front up from their own pocket what may be in the hundreds or even thousands of pounds. After the experiment, cash in advance is either accounted for with receipts (for money spent) or returned (for unspent balances).

Payments to subjects: a simple receipts system (e.g. as the one used to pay small amounts to participants to surveys) should be feasible.