How Many Dimensions in the Prosocial Behavior Scale? Psychometric Investigation in French-Speaking Adolescents


Alexia Carrizales, Bordeaux University, France
Cyrille Perchec, Bordeaux University, France
Lyda Lannegrand-Willems, Bordeaux University, France


Prosocial behaviors refer to behaviors intended to benefit others and have been theoretically and empirically linked to a variety of positive psychological outcomes for personal and social adjustment especially during adolescence. Although the importance of understanding behaviors that benefit society has been highlighted, few measures and none in French are available. Among the few mesasures assessing the multidimensional nature of prosocial behaviors, Caprara, Steca, Zelli and Capanna (2005) developed a scale for measuring adults’prosocialness (PBS). However, no factorial validity of the multifactorial structure of PBS has been published. The current study investigated whether this scale would be the same in a French sample of early to late adolescents. Specifically, we evaluated its factor structure, internal consistency, convergent validity and measurement invariance across gender and age.


Three independent samples were used for the analyses, Sample 1 (N = 1141, Mage = 14.35, SD = 1.69) used in a first Confirmatory factor analysis (CFA) for testing the original factor; Sample 2 (N = 1071, Mage = 15.19, SD = 2.27) used in the Exploratory factor analyses (EFA) and Sample 3 (N = 1640, Mage = 14.58, SD = 1.88) used in the CFA analyses (to assess both the factorial validity and the measurement/structural invariance of the PBS). They were divided into two groups based on educational grade. The first one included 867 junior high school adolescents (grade 6–9) and the second one included 773 high school adolescents (grade 10–12). All participants were administered the PBS, a 16-item questionnaire rated on a 5-point Likert scale that assesses four types of prosocial behaviors (sharing, helping, taking care of, and feeling empathic).


Although the original four-factor structure yielded an acceptable fit, we decided not to retain this structure due to low discriminant validity. Replication analyses in EFA suggested a two factor solution. CFA showed that the two-factor model that we labeled Helping and Caring had a good fit. They also had good internal consistency. Multigroup CFAs revealed configural and metric invariance across gender and partial scalar invariance.


We showed that a two-factor model of the PBS, comprising Helping and Caring, was the best from early and late adolescents. Further investigations with other samples (e.g., emerging adults and adults) compared to adolescents are needed in order to test whether the two-factor model is the best one across age, or whether this structure in two dimensions is specific to the period of adolescence. Indeed, several theories on the development of social cognition posit that during adolescence important steps forward are made in social perspective-taking (Hoffman, 2001), thereby fostering prosocial behaviors (Eisenberg, Spinrad, & Knafo-Noam, 2015). Meanwhile, the capacities of internalized / self-reflective other–oriented modes of reasoning observed across the 20s and into the early 30s might lead to a finer distinction of different dimensions of prosocial behaviors (Eisenberg, Hofer, Sulik, & Liew, 2014). These considerations support our findings and other investigations across age from a developmental perspective.

Link to paper > Contact > Website >
Back to Spotlights