Skip to content

H809 – New vs. old theme – Readings 11 and 12 (A11.6)

May 2, 2011

Bos et al., (2002) Effects of four computer-mediated communications channels on trust development


Examine trust development in a set of  four media conditions (face-to-face, video, audio and text chat) playing a social dilemma game.


Experimental setting – 198 subjects (105 male, 93 female, mostly students, average age =23), 66 x 3 person groups (nine all-male groups, seven all-female groups, 36 mixed gender groups)


Collaboration and communication to build up trust.


  • pre- and post-questionnaires measuring their general level of trust
  • playing 30 rounds of the social dilemma game called Daytrader
  • level of cooperation is measured by the groups total payoff  – assumption that high level of cooperation, must have achieved a degree of trust – corroborated by results from post-questionnaire

‘Social dilemma games are defined as situations where the best interest of the group as a whole conflicts with the best interest of each individual, so that if each looks out only for themselves, all lose’.


  • Communication media did affect how much groups were able to make in the social dilemma.
  • ANOVA analysis showed that communication conditions had significant effect on investment, F(3, 58)=6,4, p<.01 – all mediated communications had some advantages (so with varying degrees) compared to f-t-f communication
  • Tukeys test showed that text chat made significantly (p<.05) lower payoff than each of the other conditions (Figure 2).
  • Post-surveys of trust confirmed that the differences observed are related to trust.
  • Round-by-round analysis of group performances give evidence that all three of the mediated conditions delayed trust to some degree, with text chat having the greatest delay. Video and audio groups performed almost as well as f-t-f, but it took longer to reach high levels of cooperation (Figure 3).
  • Figure 3 shows that mediated conditions are more vulnerable to defections, seen in the pronounced drop offs (fragile trust).
  • possible reasons: lack of visual and verbal cues, CMC decreases the other-awareness and the inhibitions the person feels about their own behaviour


  • relative short duration of the experiment – no follow-up so far
  • use of student-age subjects – using older subjects would strengthen findings
  • experimental setting no natural setting
  • more research is needed in laboratory and field settings about how trust-based relationships develop and erode over time and multiple interactions
  • emerging tools, media types and collaboration settings might change over time – comparability




Trust is important and required in business to work together effectively. The research was done to help managers to make decisions whether they might invest in certain computer-mediated communication tools or better travel to see business partner face-to-face.

Table 11.1

Interventionist – Experimental (Behaviour is actually influenced or controlled. People are allocated randomly to experimental Outcomes are established by measuring behaviour or by using psychometric tests.

Table 11.2


Words or concept I don’t understand

  • so far so good 🙂

Statistical terms or methods that are new to me

How convinced I am by the research

I am not really convinced. We had the topic social cues in H807 a couple of weeks ago and other researchers like Walther (1994) or Utz (2002) argued that with adequate training and enough time computer-mediated communication can be greatly improved and is similar to f-t-f communication capable to carry socio-emotional information. Spears and Lea (2002) claim that CMC technologies can increase a stronger feeling of belonging, contrary to face-to-face settings. The group seems more homogenous without visual cues, leading to e. g. greater group identification. That all confirms Walter’s ‘hyperpersonal interaction’ and shows that less (visual and verbal cues) can be indeed more. (More detailed information can be found in the blog post When more is less from 15 March 2011). Controlled experimental settings cannot be really compared with natural settings and I doubt if the results can be so easily translated into a business setting. In a business setting trust depends on more factors, not only if somebody is willing to spend tokens for the group good.

Qualitative methods (statistical tests like ANOVA, p-values, etc) have the disadvantage that they are to crude and artificial to generate useful insights in the study of human behaviour. Qualitative methods are not sensitive to context and content.


Bos, N., Olson, J., Gergle, D., Olson, G. and Wright, Z. (2002) ‘Effects of four computer-mediated communications channels on trust development’ in Proceedings of the SIGCHI Conference on Human Factors in Computing Systems: Changing Our World, Changing Ourselves, Minneapolis, Minnesota, 2002, New York, NY, ACM; also available online at (Last accessed 02 May 2011).

From → H808

Leave a Comment

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: