An investigation on the social and cognitive consequences of virtual communication

posted 9 Nov 2015, 22:04 by Franco Bagnoli   [ updated 9 Nov 2015, 22:22 ]
By Mirko Donadoni and Andrea Guazzini



Humans are social animals. Most of our actions are determined by the neighbouring environment, or, better, by what we perceive as our neighbouring environment. So, we act differently if we are aware of being observed with respect to the feeling of loneliness, and clearly this depends on our perception of others. In particular, we are subject to almost unconscious reactions to eyes: it is sufficient a photograph or synthetic eyes on a computer screen to alter one’s behaviour. On the other hand, we tend not to intimately appreciate the possibility of being spied by a camera ocular, even if the “recording” light is on.

We are concerned not only by the possibility of being viewed directly, but also by the possibility that other people can access our documents, especially if they are embarrassing or not socially acceptable. This jealousy for personal items tends to diminish with time.

On the other hand, we are quite curious of knowing personal details of neighbouring individuals, and like to share/comment/discuss them with friends (gossips).

The respect of social norms is clearly affected by the possibility that some misconduct can be directly or indirectly observed. For instance, a fine is much more effective if the infraction is made public.

A final point concerns how our cognitive engagement depends on the group size. It is well known that we switch to different “modalities” if we are involved in a pair discussion, a chat group of 3-5 people, a small group of 5-12 members, or a crowd of more people. We are also able to dominate hierarchic relationship, say a pair discussion while interacting with a small group surrounded by a crowd.

In the virtual world, most of the “automatic” signals that we are accustomed to in real life do not work. We are literally unaware of how many people can view or access to what we feel like “personal” pieces of information on the web. People tend to share personal information on social networks, without considering how much time they can persist and how easily they can be found.

Humans need to understand their social world, as pointed out by scholars of social cognition.

However it is not obvious that in a virtual environment the processes that regulate the actions and the perception of reality are the same as the real world. The processes by which one acquires information, its interpretation and storage, as well as the cognitive performances of the subject are influenced by the environment. To confirm this, researchers from the Social Identity model of de-individuation have shown that despite the condition of anonymity offered by ICT, people tend to rely even more on social norms than the real world.

In order to investigate these topics, we are planning a long-term set of studies, both theoretically and experimentally. In particular, we are going to describe a set of experiments carried out in the fall 2014 by students of the course “applications in social, work and organizations psychology”.

The purpose of the following studies is to investigate how, in a context of virtual interaction and in a short amount of time, the social dynamics become established and which mechanisms are involved.

Procedures, Methods and Sample Description

The 34 participants of the experiment were recruited through voluntary census.

Participants had at their disposal a console with two text windows, one to communicate with others in the experiment in public and another to do it in private. There were also two “radar” interfaces within which were placed the symbols representing the subjects. The first, public and visible from other subjects, allowed the participants to move only their own symbol in order to facilitate communication in the public chat. Indeed, messages were darker the closer was the sender to receiver in the public radar, and vice versa they tended to fade when the sender was farther from the receiver. The second, private and invisible to others, allowed them to move other participants’ symbols, bringing them closer or moving them away from their own symbol fixed at the centre, indicating in this manner the perception of closeness of opinion with others.

In this way it was offered to the participants an equivalent of the non-verbal communications similar to change place so to be closer to a given person (the public radar), and a mnemonic aid for the representation of others’ identities and their perceived social proximity (the private radar).



Guazzini, A., Vilone, D., Bagnoli, F., Carletti, T., & Grotto, R. L. (2012). Cognitive network structure: an experimental study. Advances in Complex Systems15(06), 1250084.


In a preliminary stage, the participants were asked to provide anonymously a shortlist of topics perceived as polarizing. Among these were chosen two topics: use of technology (iPhone, iPad) with children (A); adoptions to homosexual couples (B). Participants were then randomly assigned to the four conditions listed below.


1- A – B [F – NF]

2 – B – A [F – NF]

3 – A – B [NF – F]

4 – B – A [NF – F]


In the frustrated game condition (F), lasting 11 minutes, the subjects had to face with others about one of the two polarizing topics, but at the same time they received the instruction to get the greater betweenness to win. In the unfrustrated game condition (NF), of the same duration, participants were directed to discuss freely on the proposed topic.


It is expected that the use of the text windows and radar interfaces could be different in the two conditions. If in the unfrustrated condition subjects can interact, and then use chats and radars, in the most spontaneous manner, in the frustrated condition the use of these will be influenced by a strategic component. For example in order to win, getting the greater betweenness, participants could modulate their way of communicating and relating to others, more than in the unfrustrated condition. It is also expected that the greater openness, strategic or implied that it is, may also affect the perceptions of the other participants and therefore the activity on radars.

The experiments took place in the computer room of the Psychology Faculty of the University of Florence (Italy).

At the end of the session the five factor adjective short test (5-fast) and the State-Trait Anxiety Inventory were administered and data about the socio-demographic variables were collected.

Participants in the experiment obtained a STAI test score between 11 and 29 (M = 17.68; SD = 4.8). Regarding the measure of personality traits the sample had the following ratings: Neuroticism (M = 13:22; SD = 3.78), Surgency (M = 15:22; SD = 2.88), Agreebleness (M = 15:55; SD = 3.69), Closeness (M = 9.84; SD = 3.61), Conscientiousness (M = 13:45; SD = 4.83).