Reeves and Nass (1996) made experiments to study if people are polite to computers in the same was as they are polite to humans interviewers, i.e., if they give more positive answers when a computer asks questions about itself than if another computer is used for questioning. Their hypothesis was that computers “are close enough to human that they encourage social responses” and therefore, “will get human treatment, even though people know it’s foolish and even though they likely deny it afterward“. In the first experiment, participants studied some general facts with the computer, assessed their learning, and received feedback on their learning from the computer. At the end of the session, the computer also gave an assessment of its own performance telling that it had done a great job.
After this session, half of the participants assessed the performance of the computer and interactions with the same computer, and the rest with another computer in the same room. With 22 participants, the results showed significantly more positive responses from the participants answering with the same computer than from the others using another computer. Also another rule of politeness was conformed: the responses with the same computer were more homogeneous than the other responses, as is the case if a person asks a question about him- or herself. Thereby, people were polite to computers, too.