Skip to main content

Prosocial behaviors in urban context

Cite this dataset

Gavreliuc, Alin; Gavreliuc, Dana (2021). Prosocial behaviors in urban context [Dataset]. Dryad.


We analyzed prosocial behaviors in a field experiment (N=307) conducted in an urban context (Timisoara, Banat region, Romania), starting from a classical Cross-Cultural Psychology research organized in UK and Iran by Collet & O'Shea in 1976. If the evoked study is focused on comparing prosocial behaviors in two very different national cultures (UK vs. Iran), we compared helping strangers strategies within the same national culture in relation to the regional identities of the help-seeking subjects. A behavioral scenario was created by asking naïve participants to offer support and give directions to a place even if they did not know its whereabouts. Drawing on social identity theory, it was tested whether regional belonging of the help-seeker (in-group vs. out-group) predicts the availability of help-givers for offering help, their availability for giving wrong directions, as well as their emotional expressiveness. Results are interpreted within the perspective of social distance between groups and show that the more distant regional identities are perceived to be, the less generous help-givers are, both in terms of their decision to help and to give wrong directions, as well as in their expressed emotions.


The data have been collected in a field experiment, in urban settings. The participants have been involved in a spontaneous interaction with a confederate, who has declared his regional identity in a interactional scenario. For an adequate calibration of our sample size, we performed a power analyses (PA), using G*Power, version Because our analyses imply chi-square tests and one-way ANOVA, we performed PA for both. Thus, for an average effect of Cohen's w = .30, an a priori PA for a chi-square test with df = 4, α = .05 and a power of .80, revealed that the required sample size was N = 133, while if the effect was lower (w = .20), the required sample size was N = 299. Regarding the PA for ANOVA test, for an average effect size f = .25, with five groups, α = .05 and a desired power of .80, the total required sample size was N = 200, while for a smaller effect size f = .20, the required sample size was N = 305. In the end, we decided to select the most conservative option for being assured that our sample size is sufficient to identify even a smaller effect. Based on this rationale, our global sample was established at N > 305. Our concrete sample was formed by N= 307 subjects. SPPS v 21.0 was used to conduct all analyses.

Usage notes

There are any missing values. 

We have uploaded on this repository both versions (Romanian and English) of the databse (as *.sav files). 

In order to understand the content of our database, the variables used are the following ones:

- the independent variable (IV) was the regional belonging of the help-seeker (manipulated, with five conditions).

- the dependent variables were: acceptance of helping others operationalized by engaging or not in a conversation with the help-seeker and indicating the availability to help, (DV1); sending other to nowhere, operationalized by giving directions to a fictional place or not (DV2); emotional expressiveness of the pedestrians, which was measured by a group of experts situated in the location of this spontaneous interaction (DV3), who observed all the communication between the confederate and the naïve subject and assessed his/her emotional disclosure by using behavioral descriptors.