This research investigates the use of pro-social emotions to improve cooperation. In particular, it tries to reconcile the results from Noussair and Tucker (2007) and Lopez et al. (2010). To reach this goal the experiment considers different degrees of revelation: no revelation, partial and full disclosure of information. Additionally, I use different microeconometric specifications to accommodate different hypothesis about the motivation of the subjects. My results diverge from those of Lopez et al. because I find that revealing the decision of a single subject at random does not significantly increase cooperation, which is the main result of these authors. Also, my findings indicate that cooperation is triggered only when I reveal information of either 3 or all the subjects in the group, the last case being similar to the public observability of Noussair and Tucker. These authors find a non-permanent increase in contributions, so I do but using a positive framed-experiment with disclosure of additional information, the group’s earning loss. Therefore, random revelation together with the disclosure of information about subjects’ decisions appears to be a good alternative to promote cooperation in a sample pool of undergraduate students. Also, I observe a reduction in contributions over time, but in the random revelation treatment this decay by less than 40%. The most interesting result that I obtain is the evidence of altruism and positive reciprocity in the specification of Ashley et al. (2003, 2010), instead of the matching in contributions reported by these authors.