The aim of this study was to explore possible parent and peer influences on adolescents’ valuing of academics and intrinsic academic motivation in cultures varying in traditional emphasis on the family unit (Cuba, Canada, and Spain). Perceived parent and peer support, parents’ expectations, and valuing of academics significantly predicted adolescents’ valuing of education and motivation. Spanish adolescents were less motivated than members of the other ethnic groups. The Spanish participants also reported lower perceived parental expectations than Chinese Canadians and less perceived peer support than did Cubans and Chinese Canadians. Perceived social support from same- and opposite-sex friends predicted adolescents’ valuing of academics and intrinsic motivation most strongly in the Canadian sample. Cuban parents’ perceived valuing of academics predicted adolescents’ intrinsic academic motivation more strongly than in the other cultures sampled.

Introduction: The term "generativity" was used for the first time by Erikson (1950). It refers to the state of adulthood in the life cycle, which implies procreativity, productivity and creativity, it also impulses the development of one's own identity. There are several ways to measure generativity such as interviews, direct observation, case studies, or self-report questionnaires. The most widely used method is self-report, and among the available scales, the Loyola Generativity Scale (lgs) is one of the salient scales. lgs is a self-report measure composed of twenty items that measure a general factor of generativity. Objective: This study examined the factorial validity and internal consistency of the adaptation of the Loyola Generativity Scale to Spanish for its use with Dominican teachers. Method: Two competitive factor models were tested based on previous existing literature, and a third model with method effects associated to negatively worded items was added, in order to find the best fitting solution for this sample. In order to examine the factorial validity of the three models, three Confirmatory Factor Analysis (cfa) were performed, and internal consistency and composite reliability indices were also calculated. Results: In general, the three models showed good psychometric properties. However, the third model that considered a general factor of generativity together with a method effect factor showed the best fit for this sample. Conclusion: Implications for an adequate measurement of generativity are discussed. More research is needed to examine if these properties are stable on different samples from different populations.