Assessing Empathy in Salvadoran High-Risk Youth and Youth Gang Members: A Spanish Validation of the “Basic Empathy Scale”

Purpose: The concept of empathy is central for understanding a broad range of behaviors, particularly violent behavior and victimization. The Basic Empathy Scale (BES)(Jolliffe & Farrington, 2006) is a promising 20-item measure for both affective (AE) and cognitive empathy (CE). The BES has shown good psychometric properties and, in addition to English, has already been validated in Italian and Turkish. It has been suggested that low empathy is related to criminal offending based on the assumption that youth who can experience a victim's feelings are less like to victimize others. Thus, this concept plays a critical role in violence prevention programs. The objectives of the paper are to examine the reliability and validity of the BES and assess its factor structure using the Spanish translation of the scale in high risk youth sample in San Salvador, El Salvador.

Methods: This study examines survey data from a nonprobablity sample of 208 youth in 10 urban and semi-urban neighborhoods in the Greater San Salvador Metropolitan Area collected between August 2009 and February 2010. All data were collected as part of the “Youth Study about High-risk Youth and Youth Gang Members in the cities of Boston, Lynn and San Salvador”. Subjects were selected in collaboration with a Salvadoran high-risk youth development organization that works closely with active gang members. Survey respondents were predominantly male (81.7%), the mean age was 20 years (with 94.4% under 26), and 62.3% reported having participated in a gang. The BES was translated to Spanish and pilot tested twice with Latino youth in the US and youth in El Salvador. After testing for normality, a confirmatory factor analysis was conducted to verify the factorial structure of BES scores; reliability estimates were determined, and construct validity was estimated. ANOVAs were conducted to examine gender and gang membership differences on both subscales.

Results: The indices in the estimated model revealed an acceptable fit for the observed data (NFI=.93; CFI=.94; RMSEA=.67; SRMR=.05; GFI=.92; AGFI=.88). Factor variances were fixed at 1.0 in the resulting model, to standardize the latent variables; all factor saturations were significant at p < .01 (ranging from .28 to .80). The model showed reasonable evidence for a two-factor solution. Scale reliability (α=.80 for the global scale, α=.82 for AE, and α=.73 for CE) was also satisfactory. Females and non gang members scored higher on both affective and cognitive empathy. Results from the analyses showed that the Salvadoran study replicated the same factorial structure as the original study by Jolliffe and Farrington (2006).

Implications for policy and practice: Youth violence and gangs represent a serious problem for Latino youth in the US and El Salvador. This study validates an empathy measure in Spanish that can strengthen the development of effective violence prevention programs at the individual and community level. By creating interventions which focus on developing and/or increasing empathy in youth, social workers involved in both countries are well-positioned to lead the way in preventing violence. This measurement would further strengthen evidence-based strategies.