Adaptation of the Positive Youth Development (PYD) Model: Evaluating the Psychometric Properties of the Adelante PYD Measure Among a Sample of Latino Adolescents in the U.S.
Open AccessAbstract of DissertationEvaluating the Psychometric Properties of the Adelante PYD MeasureBackgroundDespite comprising a significant and growing proportion of the U.S. population, Latino adolescents are considered underserved and greatly understudied. Based on what we do know, many Latino adolescents face unique challenges that are likely to influence their development. Although positive youth development (PYD) has been studied extensively, U.S. studies have included predominantly middle to upper class white adolescents. Despite the existence of several PYD models and measures, to our knowledge, none have been developed or used among Latino adolescents in the U.S. Of those studies that have successfully engaged Latino adolescents in research, the majority have focused on participation in risky behaviors. Although these studies clarify the negative impact adversity has on Latino adolescents, it is important to also explore areas where they may be thriving or have become resilient. To better understand Latino adolescents’ experiences and behaviors, the Adelante PYD measure was developed to address this disparity among a large sample of Latino adolescents residing in two marginalized communities in the Washington, DC metropolitan region. The purpose of this dissertation is to evaluate its psychometric properties through a series of factor analyses, thus determining the degree to which it is reliable, valid, and generalizable to youth beyond the Adelante study sample. Methods A theoretical PYD framework and associated measure containing 89 items assessing PYD constructs was developed using four subscales and administered to Latino immigrant adolescents residing in the DC metropolitan area suburbs of Langley Park, MD and Culmore, VA. The Adelante PYD measure includes four subscales: confidence, competence, connection, and contribution; all items were assessed using a 5-point Likert scale. Using baseline survey data (n=720) collected as part of the larger Adelante study, a series of exploratory and confirmatory factor analyses were conducted to examine the psychometric properties of each subscale. Tests of measurement invariance were used to determine whether gender contributes to significant differences in the interpretation of and reporting on the measure’s items for each subscale. The baseline data was randomly split and used to conduct both the exploratory (EFA) and confirmatory (CFA) factor analyses. Initial assessment of each subscale’s reliability and the EFAs utilized 40% (n=288) of the baseline study sample while the other 60% was used to conduct the CFAs. EFAs were conducted using principal axis factoring. Two alternative models for each subscale were also produced to validate the final factor model. CFAs were generated to confirm the factor structure of each measurement model and acceptable fit with the data. Multigroup CFAs using a hierarchical approach were used to estimate measure invariance by gender. For all four final measurement models, each type of measure invariance was tested for; fit parameters were used to assess invariance of each model. Results The main findings of this dissertation provide support for assessment of the 4 Cs among Latino adolescents in the U.S. There were 720 respondents who completed the Adelante PYD measure at baseline. Overall, participants had been living in the U.S. for nearly 10 years on average, and had resided in their respective community (i.e., Langley Park or Culmore) for an average of 6 years. Participants were offered the survey in Spanish or English; 81% opted to complete it in English. Preliminary analyses of the initial subscales suggest good internal consistency. Findings from the series of EFAs using 40 percent of the baseline survey data support hypotheses for four of the five measurement models’ factor structure and each model fit well with the baseline data. Although some items were dropped, no more than 20 percent of the items were removed from each subscale, resulting in at least three items loading on to each factor, as hypothesized. Results from the series of CFAs that followed confirmed the factor structure of all four subscales using the remaining 60 percent baseline survey data. Results from the measure invariance analyses imply adolescent boys and girls in this sample did, to some degree, understand the conceptualization of the factors and items included in each subscale. The final confidence measurement model performed best, achieving partial invariance which supports the notion that boys and girls in this sample shared the same conceptual understanding of confidence and employed that shared understanding when responding to the confidence items. The competence and contribution measurement models performed similarly; results supported metric invariance but did not support configural or scalar invariance. Scalar invariance was achieved for the connection subscale. Overall, the results are promising and imply not only are the 4 Cs relevant to boys and girls in this sample but that the measure, following some improvements, could be used to accurately capture true differences in the 4 Cs between adolescent boys and girls. Mean scores were relatively high across factors and across boys and girls, indicating general agreement with items assessing overall confidence; competence in workforce preparation, academics, and athletics; and connection to social and cultural identity, school, family, friends, and community. Civic action and contribution subscales had the lowest overall mean scores. Conclusion Results assessing the psychometric properties of the Adelante PYD measure administered to a sample of Latino adolescents serve as an initial step to better understanding relationships between developmental characteristics and risk behaviors among this population. This is the first measure known to include culturally relevant domains and items to assess PYD, with versions available in English and in Spanish. Although revisions are recommended, the results provide early evidence for the measure’s reliability and validity. Analyses of each of the four subscales resulted in clean, simple factor structures that fit the baseline Adelante PYD data well. Future use of this measure will allow researchers to assess assets rather than deficits among this population and can be administered to measure the impact of programs and policies that may directly or indirectly influence the 4 Cs and adolescents’ PYD.
- All rights reserved
Notice to Authors
If you are the author of this work and you have any questions about the information on this page, please use the Contact form to get in touch with us.