Experiments

Experiment design

320px480px640px786px1024px1280px1440px

In addition to the laboratory, the experiments are currently being conducted increasingly online. With a network consisting of 68 computers, the Cologne laboratory is one of the largest and most modern of its kind in Germany. Through the use of world-leading software, it offers the possibility to conduct interactive experiments completely digitally and to observe precisely stimulated situations.

The "design" of an experiment means the specification of the underlying "game", its parameters and the monetary incentives. Preferences, technologies and other economically relevant parameters can be perfectly induced and controlled. Often, only one parameter is changed to compare two different situation constellations. Thus, changes in the behaviour of the test subjects can be traced back to exactly this one parameter.

This allows important conclusions to be drawn about the nature of economic behaviour and the performance of markets and other institutions. The research goal is to develop modern, practical algorithms to improve marketplaces and other economic or social institutions.

A typical procedure of an experiment

1. The idea

2. The design

In the "design" of the experiment, the underlying "game", its parameters and the monetary incentives are specified. Preferences, technologies and other economically relevant parameters can be perfectly induced and controlled. Often, only one parameter is changed between two arrangements to be compared. Thus, changes in the behaviour of the test subjects can be traced back to exactly this one parameter.

3. The organisation

Participation in economics experiments is voluntary. Interested individuals can register in the laboratory's online database. When a new experiment is conducted, those registered receive an invitation by e-mail and can register for a session of the experiment. Experimenters can also select the participants to be invited (mostly, but not exclusively, students of the University of Cologne) on the basis of various characteristics such as field of study and profession, participation in previous experiments and much more.

Before the experiment, the instructions have to be designed and the software has to be programmed and tested.

4. The implementation

Participants are welcomed to the laboratory by the experimenter and assigned to laboratory places. Each laboratory place is equipped with a computer, monitor, input devices and visual protection. The computers are connected by a network to an experiment server on which the experiment programme is run, which processes the decision data centrally and controls the output at the laboratory places.

After the participants have taken their seats, they receive instructions on how to conduct the experiment. Questions are answered, but communication between participants is often not allowed until the end of the experiment.

In the experiment, depending on the experiment, different templates appear on the screens of the lab seats, informing about the decisions of the other participants, random moves or payoffs, and on which the participants enter their own decisions.

The participants' rewards from the rules and decisions according to the respective instructions of the experiment. At the end of the experiment, all participants are paid separately and dismissed.

UNIVERSITY OF COLOGNE

Faculty of Management, Economics and Social SciencesCologne Laboratory for Economic Research (CLER)

Experiment design