Experimental Tests of Normative Group Influence and Representation Effects in Computer-Mediated Communication.

Two experiments addressed the questions of if and how normative social influence operates in anonymous computer-mediated communication (CMC) and human-computer interaction (HCI). In Experiment 1, a 2 (public response vs. private response) × 2 (one interactant vs. four interactants) × 3 (textbox vs. stick figure vs. animated character) mixed-design experiment (N = 72), we investigated how conformity pressure operates in a simulated CMC setting. Each participant was asked to make a decision in hypothetical social dilemmas after being presented with a unanimous opinion by other (ostensible) participants. The experiment examined how the visual representation of interaction partners on the screen moderates this social influence process. Group conformity effects were shown to be more salient when the participant's responses were allegedly seen by others, compared to when the responses were given in private. In addition, participants attributed greater competence, social attractiveness, and trustworthiness to partners represented by anthropomorphic characters than those represented by textboxes or stick figures. Experiment 2 replicated Experiment 1, replacing interaction with a computer(s) rather than (ostensible) people, to create an interaction setting in which no normative pressure was expected to occur. The perception of interaction partner (human vs. computer) moderated the group conformity effect such that people expressed greater public agreement with human partners than with computers. No such difference was found for the private expression of opinion. As expected, the number of computer agents did not affect participants' opinions whether the responses were given in private or in public, while visual representation had a significant impact on both conformity measures and source perception variables.