Behaviour Change

PROPAGANDA FOR CHANGE is a project created by the students of Behaviour Change (ps359) and Professor Thomas Hills at the Psychology Department of the University of Warwick. This work was supported by funding from Warwick's Institute for Advanced Teaching and Learning.

Thursday, February 4, 2016

"I'll do yours if you'll do mine": The role of trust in reciprocity.

It’s dissertation time at Warwick University, and all those doing experiments are scrambling for participants wherever and whenever our timetables allow. Thankfully, there’s an easy way to get other experimenters to take part in yours: just offer to participate for them in return!

Although somewhat questionable when it comes to ethics (did they really volunteer to take part if they’re getting something in return…?), this is a prime example of the norm of reciprocity: the reason you feel you must do something nice for someone who’s helped you out in the past (Cialdini, 2009).

One key part of reciprocity is the trust that you must place in the other person – how can you be sure that they’ll do your experiment if you do theirs first? Berg, Dickhaut and McCabe (1995) used an investment setting to investigate the role of trust in reciprocity. In this study, participants each received a $10 show-up fee, and were assigned to one of two conditions in a set of 32 pairs: they would either be a trustee, immediately pocketing the full amount; or they would become a trustor, deciding how much of their $10 to send to their trustee. This investment was then tripled and given to the trustee, who then had to decide how much of their new, larger total to send back to their trustor. Got that? Phew.

Essentially, Berg et al.’s (1995) participants were being tested to see how much trust they placed in the norm of reciprocity – should we all invest our entire earnings into an anonymous stranger in the hope that our generosity may be returned?

So, what did they find? Well, the figure above (Fig. 2; Berg et al., 1995) shows the amount paid and received by each participant within their pairs. The initial investment fluctuated considerably, suggesting a wide variation in the amount of trust felt by the trustors, affecting their generosity. However, the trustee actions are more interesting: just over a third of those who received money reciprocated that trust with an act of greater generosity, returning a larger amount to their trustor.

You might initially think we’d be placing too much trust by investing in a 33% reciprocity rate. However, Berg et al. (1995) considered this by introducing a “social history” element: a further 28 pairs of participants were involved in a replication experiment, but this time all participants knew the rates of reciprocation from the previous experiment. As such, you might expect fewer trustors to invest. The figure below (Fig. 4; Berg et al., 1995) suggests you should consider otherwise.

When previous reciprocity rates were known, trustees were more aware that a larger investment from their trustor was related to a higher level of trust, and therefore deserved more reciprocation, resulting in a correlation between reciprocity and investment (Berg et al., 1995). So, knowledge about previous rates of the benefits of trusting increased the likelihood for participants to believe in its effects, and therefore use it as a basis for investment and follow-up reciprocation.

How does all this investment nonsense relate to gathering participants, then? In the case of other experimenter, let them know that they can trust you, and in return you’d be placing a lot of trust in them. After all, it’s your dissertation – the most important project of your uni career! And if them investing time in something so crucial for you means that you can reciprocate the same for them, all the better.


Cialdini, R. B. (2009). Influence: Science and practice. Boston: Pearson Education.

Berg, J., Dickhaut, J., & McCabe, K. (1995). Trust, reciprocity, and social history. Games and Economic Behaviour, 10, 122-142.

No comments:

Post a Comment

Note: Only a member of this blog may post a comment.