International Journal of Internet Science
A peer reviewed open access journal for empirical findings, methodology, and theory of social and behavioral science concerning the Internet and its implications for individuals, social groups, organizations, and society.
Physical or Virtual Presence of the Experimenter: Psychological Online-Experiments in Different Settings
Heike Ollesch, Edgar Heineken, Frank P. Schulte
Abstract: Internet-based experiments are characterized by the physical absence of an experimenter and the difficulty to provide a standardized experimental arrangement. How does this affect internal validity and precision? Three experiments on different psychological questions, following studies by Heider and Simmel (1944), Morris and Stevens (1974), and Moyer (1973), were conducted in the virtual laboratory Lab.OR. To different degrees each of the selected experiments demanded commitment of participants and features measurements, which are usually considered susceptible to external disturbances. Each experiment was conducted as an Internet-based experiment and, using WLAN-notebooks, additionally in two other settings (laboratory rooms / public rest areas on campus), so that the degree of standardization of the experimental arrangement was varied. While the experimenter was virtually present in the Internet-based experiments, the kind of physical presence of the experimenter was varied in the laboratory and campus settings: either he was actively involved and gave oral instructions, or he just attended the experiment, but did not play an active part while all information was provided on the computer's screen.
Results show that all experiments provide equally valid and precise results, independent of setting or mode of presence of the experimenter. Taking into account the different psychological variables studied in the different experiments, some interesting effects were found: Participants' performance was affected by the kind of presence of the experimenter, when they had to write an elaborate report, but not when response latencies or recall rates were obtained. Lack of standardization of the environment did only affect response rates: performance in laboratory and Internet settings was superior to campus setting.
Implications for experimental online research and the potential of using the virtual laboratory Lab.OR in teaching experimental psychology to point out methodological specifics of online-experiments conducted in different settings are discussed.
Keywords: Internet-based experiments, laboratory experiments, web research methods, comparability, social presence
Download full paper