Already in the 1970’s, numerous experiments were proving that a brief touch of one’s arm may influence our behaviour in many ways. We may become more altruistic, compliant, or generous. This phenomenon in psychology is called the Midas touch effect. The name was inspired by the ancient Greek myth about the king Midas, who was granted a power of turning anything he touched into gold. Similarly, a waitress touching the client’s arm while asking whether everything is ok with the order can hope for a higher tip. This effect does not occur always in this way though. Research has proven that in Poland when a man touches another man while making a request, his touch may, in fact, decrease his chances for a positive reaction. Thus, this research project aimed to explain what modifies the efficiency of the Midas touch technique. We focused on two main hypotheses:

1. Poland is a country, in which male homophobia is very common. A male touch may bring homophobic associations and therefore, discourage the touched to interact in a prosocial way.

2. Polish culture belongs to the circle of cultures of honour, in which men care and compete for the social status. Crossing personal boundaries through touch may evoke a feeling of threat, fear, or anger.

To verify these hypotheses we used the well-known decision-making game ultimatum in an immersive virtual reality version. The goal of the game is to divide how to split a pie between myself and the opponent. In a series of experiments, Polish men played with virtual humans (autonomous embodied agents) of various genders, social statuses and masculinity levels. We registered their psychophysiological reaction to the agent’s touch. In four experiments, we proved that men bring their prejudice and attitudes to the virtual environments, offering to the most feminine-looking male agent less than to the rest. Moreover, the touch was the most efficiently increasing compliance in the case of the low-status agent. Interestingly, the high-status agent was evoking high compliance in participants regardless of touching or not before making an offer. These results match with the current knowledge both within the psychology of social influence and virtual reality. They prove that people play with embodied agents “mindlessly” as if they were humans. Our participants played in a quite altruistic way, sharing with the opponents on average 40% of the coins and rejecting the extreme offers. This means they wanted to be fair instead of selfish, even though they were aware they played with a computer algorithm. Even though, they remained influenced by the authority of the high-status agent. They were also compliant to the agent of the low status but only in a situation when he used the Midas touch technique. On a physiological level, participants were reacting in a specific way to the touch of the embodied agents, it felt real to them.

Our studies offer answers to several research questions, which remained thus far unanswered, and enriched the discussion of the mechanisms underlying the efficiency of the Midas touch effect. They shed also new light on the role of social status and homophobia and their relationships with men’s compliance to the interpersonal touch. The most important conclusion for the society is that people comply to the virtual, embodied agents in the exact same way as they do to other people in the real world, even when the manipulation takes a shape of a subtle influence technique like the Midas touch. These results may have a significant impact on many disciplines, in which the virtual reality technology wins more and more applications, from education and gaming to therapeutic programs and marketing.

A brief touching of one’s arm or shoulder can increase our proneness to social influence techniques. Interestingly, a virtual touch also influences behaviour. This phenomenon, called the Midas touch effect, is not always equally efficient. Previous studies indicate that in Poland, men who were touched before they heard a request were not only resistant to the manipulation, but even less compliant than those who were not touched in similar circumstances. This result was explained by a strong homophobia among Polish males. Nevertheless, there is an alternative explanation based on social status, which indicates who, whom, and in which circumstances can touch. Cultural differences play a significant role too: in Poland interpersonal touch is much less common than, for instance, in Spain. Our study aimed to understand how cultural background, male homophobia, and social status increase the efficiency of the Midas touch effect. We ran the experiment in Warsaw and Barcelona in order to compare the behavior or men in a noncontact (Polish) and contact (Catalan) cultures. To obtain that goal, we collaborated with scientists from Event Lab at the University of Barcelona. We have built an immersive virtual environment UltimatumVR, which was based on a classic decision making game ultimatum. The goal of the game is to divide a certain amount of money between two players. One player makes an offer of the split and the other one responds to it either by accepting the offer or by rejecting it. If he/she accepts it, both players receive as much as they agreed to. In case of a rejection, none of the players receives anything. We designed an immersive version of that game where users played in a randomized order with two embodied agents which looked and behaved like humans, but were driven by a computer algorithm. Agents represented a man of a stereotypically high and low social status which were signalized by their profession, name, and clothing. Players used the HTC Vive head-mounted display to get immersed in the game’s world. In the game, they were going through a training to get familiar with the interface and then played 18 trials of ultimatum game, 9 with each of the agents. Additionally, before making the offer, the proposer could have touched the responder, which was transmitted by a vibration. The results suggest that in identical conditions, people from different cultural backgrounds react differently to simple social interactions in virtual reality. Moreover, a touch both in Poland and in Spain increased compliance towards the agent, which stands in opposition to the literature. Simultaneously, the more prejudiced toward homosexual men were Polish players, the more compliant they were towards the agent. In case of Catalans, the result was opposite – stronger homophobia predicted less accepted offers from the agent. It is important also to note that Polish men were generally much more homophobic than Catalan. Interestingly, it did not matter whether the agent represented a high or a low social status. At the same time, player’s generosity, measured as the offer’s value proposed to the agent, did not depend on homophobia, but on the extent to which he valued social status as an important element of life. Also here were observed cultural differences: in Poland, giving more importance to social status predicted less generous offers for the agent. In Catalonia though, the result was opposite: the more valued social status, the better offers received the agents. These results contribute to the discussion of human-computer interaction and of the masculinity image in different cultures. They indicate that the immersive virtual environments have to be designed with consideration of cultural differences to assure that users will read certain verbal and non-verbal messages in the expected way. The described study indicates also that the efficiency of social influence techniques may be modified by cultural and individual differences, and that virtual reality opens new paths to study all these aspects simultaneously.