4.3 Aids prevention among latino teens: the role of family  (Page 8/9)

Methods reviewed in studies

The sample size of the etiological studies are usually modest and are not always samples of probability as can be seen, for example, in the studies of Brook (1990), who uses convenience samples. A greater variety of sizes occurs in samples of evaluative investigations like those of Kirby (1994), 7.753 teens in the ENABL program, 3.058 teens in the Safer Choices program (Coyle 1996-2001).

The preferable design is experimental, but there always are quasi-experimental designs. The randomization unit is normally the school, since the programs try to improve the preventive formation and not only to the individual student. However, the randomization of a small number of schools makes the comparison between the experimental and control groups difficult. With the new multi-level software packages as BMPD5, one can analyze individual data as well as aggregates by school or classroom.

The methods of statistical analysis range from the simplest to the most sophisticated. Thus, for example, whereas in the evaluation program ENABL (Kirby, Korpi, Barth,&Cagampang, 1995), the authors resorted to Student “t” tests. Others, like Basen-Enquist et al. (1997), preferred a design of cohorts in their evaluation of the Safer Choices programs. Kirby et al. (1999) used an analysis of variance of two factors keeping “school” as the random effect to consider the intra-class correlation in the evaluation of a program applied in schools in Seattle. In any case, the collection of data must allow matching questionnaires in the base line (pretest) and, at least, six months after the intervention.

Strategic conclusions and limitations

HIV/AIDS Research Synthesis Project: Compendium of HIV Preventioum Interventions with Evidence of Effectiveness (1999) of the CDC of the North American Ministry of Health and Human Services represented a step ahead and, simultaneously, a proof of little sensitivity toward ethnic and gender factors. None of the included programs were in Spanish, so there is no way of knowing whether they are effective or not with Latino teens. Public health policies would have to promote preventive strategies that recognize the contextual factors, such as socioeconomic conditions; cultural factors like norms regarding roles, race, and ethnicity; and, mainly, the inequalities in sanitary attention, instead of prioritizing the models of prevention that focus on the individual oriented toward the general population. As Barbara Marín (2003) said,

programs for teens and their parents are needed; in particular those oriented toward those teens that date older people. These programs must include an ample understanding of the Hispanic culture and the forms in which it can protect and expose people to sexual risk situations. There is a lack of investigation about programs that approach the roots of the lack of power (disempowerment). (p. 191)

Thus, to help give power to Hispanics they must overcome the obstacles that racism, sexism, and homophobia represent in society as a whole.