Development and Preliminary Validation of Scales to Assess Puerto Rican Adolescents' Intentions to Engage in Dating Violence

Background and Purpose: Many of the scales used to assess dating violence (DV) in Puerto Rico have been developed elsewhere and have not been developed or adapted and validated for Puerto Rican (PR) adolescents. Scales used to assess DV do not always consider the impact of culture and assume invariance and equivalence of the measures. This two-phase study aimed to develop and validate a scale which measures diverse factors influencing PR adolescents’ intentions to engage in DV (e.g. behavioral beliefs, social norms, emotions). The scale under investigation is the Behavioral Intentions Scale (BIS), which was developed by the authors based on a qualitative study of 48 PR adolescents.

Method: In the first stage, we sought to develop and conduct a preliminary validation of the BIS using elements from the ITC Guidelines for Translating and Adapting Tests. An Expert Committee was formed to develop the BIS scale and assess for face and content validity, as well as to assess for cultural, linguistic, and development stage aspects. Then, scales were field tested via an online survey in a sample of adolescents (n = 30; 13-17 years old).

Based on the initial field test, specific items were further refined in a second stage with four focus groups (n = 6 per group) divided by development stage and gender. Questions regarding the survey were asked (e.g. readability of items, clarity of instructions, perceptions of how peers and parents would respond to survey). Finally, a sample of n =120 adolescents completed all revised scales via an online survey. Preliminary psychometric properties were established for all scales.

Results: In the first stage, Cronbach's alpha for the BIS by sub scales and comparable questions were: .62, .80, .93,.95 .94,.96, .91, .94, .84, .74 (N =20-24). In addition, a few items received only a few responses. In the second stage, four themes emerged in the focus groups. Developmental group differences were found, suggesting the need to develop two scales. Gender-based views of DV and concern over peer and parents perceptions about the scales were noted, indicating the need to incorporate a social desirability measure. Adolescents across all groups suggested adaptations to the instructions and items of some of the scales and sub-scales. Cronbach's alpha for BIS by sub scales and comparable questions were: .88, .91, .92, .97,.87,.93, .88, .94, .95, .97, .85, .82, .91 (N=82-90).

Conclusions and Implications: Results from this study showcase how each step of a mixed-methods study can contribute to a final product in which the development of a new measurement is completed. In addition, the study validation flow describes how scale development is critical to building knowledge in human and social sciences. The use of a rigorous methodology informed by evidence-based guidelines allows for the rigorous and reliable measurement of a serious problem. Finally, validation of adequate and culturally sensitive measures enable researchers and practitioners to acquire knowledge on the determinants influencing a critical problem, PR adolescents' intentions to engage in DV.