Item

Abstract

Conducting psychophysical experiments on the internet might constitute a useful addition to the existing spectrum of experimental methods in behavioural and cognitive sciences. Web-experiments allow access to a huge number of anonymous subjects that spend a short amount of time on an experiment. The technique is therefore suitable for experiments that (i) need few data per subject (eg, one-shot recognition experiments), (ii) need data from many subjects to get a significant result, (iii) cover a large parameter space, (iv) are designed to catalogue possible behaviours (eg, identification of exploration strategies in navigation), or (v) have a demand for subject diversity.

We have implemented five web-experiments that investigate gender perception, perception of face orientation, visual encoding of scenes, canonical views, and memory for faces (http://exp.kyb.tuebingen.mpg.de/web-experiment/). In the first few months we had up to 200 subjects for the shortest and most attractive experiments, whereas experiments that took more time or required additional browser plug-ins (VRML) attracted only 20 - 50 subjects.

The performance of a control group (n=20) that ran the same experiments on a computer in our laboratory confirmed the validity of our web-experiment data. Feedback from this group allowed us to improve the design of the experiments.

We conclude that web-experiments can form a valuable method for accessing large groups of subjects, provided careful thought is given to the limitations of using anonymous subjects on remote locations. The technique is especially well suited for performing quick pilot studies and for validating laboratory experiments on larger numbers of subjects. Experiments that require precise control of timing, colour, display characteristics, etc (low-level psychophysics) should not be considered for internet implementation.