Taking the assumptions of a medical approach to primary prevention, traditional strategies for preventing alcohol intake have focused mainly on equipping adolescents with information (Ballesteros et al., 2003). From the bio-psycho-social approach, prior evidence has pointed out the limitations of such an approach, as well as the need to use comprehensive training programs that include self-improvement skills and social skills (Marlatt and Witkiewitz, 2002; De Ridder and de Wit, 2006; Lemstra et al., 2010). Some recent studies from experimental and clinical approaches have reported improvement effects in the adolescents' psychological variables, such as self-esteem and self-efficacy (Alexander and Anderson, 2017), internalizing personality variables (anxiety) and externalizing variables (behavior problems) (Scalco et al., 2014; O'Leary-Barrett et al., 2016; Wills et al., 2016). The role of social identity has also been analyzed, through online contexts (Pegg et al., 2017), and universal programs for preventing alcohol intake in adolescents have been applied (Teesson et al., 2017). Some improvement is seen in all cases, but without conclusive results.

An interactive approach from Educational Psychology, the recent Theory of Self- vs. Externally-Regulated Learning (SRL vs ERL) (de la Fuente, 2015, 2017), has focused on the nature of the adolescent (Oliva et al., 2002; Oliva, 2007), and his/her environment–whether regulatory, a-regulatory or dysregulatory. Regulation is assumed to be an essential procedural component of the competency for managing alcohol intake and other chronic health concerns (Clark et al., 2001; Hull and Slone, 2004; Duell et al., 2016). Thus, the worst scenario for primary prevention would consist of an adolescent with a low level of self-regulation, and a profile of dysregulatory behavior (negative proactivity toward alcohol intake) in interaction with a dysregulatory context (actively encouraging consumption), thereby establishing a high probability of consumption and nonadherence to preventive treatments. In accordance with this theory, the competency model incorporates different behavioral levels in a wholistic fashion, and serves as a model of an Educational Psychology intervention, being distinct from other approaches, as has been illustrated in the case of academic stress (de la Fuente, 2015; de la Fuente et al., 2017).

An individual may be considered competent when, in a given context, he or she is able to effectively and efficiently solve the problems that arise (Roe, 2002, 2003). From an Educational Psychology point of view, other authors have established that being competent means having the knowledge, skills and attitudes that allow a person to make an appropriate response to a particular, real, problematic situation (de la Fuente et al., 2005). This gives rise to a three-fold division into subcompetencies that are necessary in order to be competent in any field of knowledge: conceptual (knowing), procedural (being able to, knowing how) and attitudinal (wanting to, being). In the specific case of educational prevention of alcohol intake, this approach means delimiting the level of subcompetencies that adolescents possess at each level.

The concept of adjustment in interacting with alcohol is understood to be modulated by age, context and culture. In adolescence, this adjustment can be conceived as the behavior of delaying contact with alcohol in a proactive regulatory manner (de la Fuente, 2017), that is, voluntary nonconsumption of alcohol and substituting it with other more adaptive behaviors, such as sports, leisure activities or having fun without recurring to harmful substances. In other words, adopting the absence of substance use in order to gain a state of bio-psycho-social wellbeing (Becoña, 2007a,b).

This research study has several motivations. On one hand, prior research on alcohol prevention has addressed many variables from a partial standpoint, seeking to establish probabilistic relationships between isolated variables and alcohol use in adolescents. However, an Educational Psychology approach, adopting the conceptions of competency and subcompetency, would seek to integrate the different variables into a more powerful schema. Based on the foregoing, then, the objectives and hypotheses of this study were as follows:

The population under study were students from public secondary schools in a southern province of Andalusia (Spain). In order to include different types of schools in this investigation, schools were selected from three types of urban areas: (1) center of town, from a medium-high social stratum (n = 106), (2) surrounding neighborhoods, from a medium-low social stratum (n = 101) and (3) outlying, marginalized population areas with compensatory education, from a low social stratum (n = 121). The homeroom teachers from every group participated voluntarily in the experiment, having been invited by the School Psychology adviser of the local Teacher Development Center. Participating students were between the ages of 12 and 17 years [12 (n = 51), 13 (n = 80), 14 (n = 97), 15 (n = 55), 16 (n = 35), and 17 (n = 10)] and were enrolled in compulsory secondary education (grades 7–10) at one of three public secondary schools. This age range was selected for methodological purposes, making it possible to form two groups, the 12- to 14-year-olds (n = 207), corresponding to the first stage of adolescence (puberty, or early adolescence), and the 15- to 17-year-olds (n = 121), corresponding to the second stage of adolescence (adolescence per se). The final sample size from which all measurements were taken contained 328 subjects. Of these, 178 were male (54.3% of the sample) and 150 were female (45.7% of the sample). The mean age of the sample was 13.82 years, with a standard deviation de 1.19.

The Alado Program is an online program (de la Fuente et al., 2012), designed for adolescents, to work on the three levels of subcompetencies previously established in the model. The program provides assessment and intervention in different matters of learning. It may be used by students, teachers and parents, but in this case it was applied only to students. The Teacher's Guide and Instructions for Use have been published.

Data collection instruments were applied over the course of the school year 2009-10, within the framework of the Alado Project of Excellence (2007–2010), through an online utility created for this purpose (www.alado.es). Specifically, these data correspond to the project baseline, having been collected in September-October, 2008. Previously, cooperation had been requested from the Teacher Development Center, from the students' parents and from the School Board, for their participation. The project was approved by the University Bioethics Commission (University of Almería). The students participated voluntarily. The parents were informed in writing. As the participants in the Project were minors, both the parents and school administrators gave written informed consent for the study. All data was collected in accordance with the principles of the Psychologist's Deontological Code and the Spanish Data Protection Act.

For this set of data, obtained at the program baseline, an ex-post facto methodological design was used. Relationships were measured just as they occurred in the natural environment, without any treatment, using inferential and structural analyses based on the Presage-Process-Product Model (Biggs, 2001), and applied to educational problem areas in students in compulsory secondary education. The variables selected were as follows:

Inferential statistical analyses (multivariate analysis, ANOVAs) were carried out using SPSS (v. 23.0) for Windows. For levels of the independent variable self-regulation, cluster analysis was used, obtaining three levels: low, medium and high. AMOS (v. 23.0) for Windows was used for the structural validity analysis of each inventory and for constructing the structural prediction model. To interpret the CFA and SEM model fit, we focused on the comparative fit index (CFI) and the root mean square error of approximation (RMSEA). CFI values equal to or more than 0.90 and 0.95 respectively were taken to indicate acceptable and close fit to the data (McDonald and Marsh, 1990). RMSEA values equal to or below 0.05 and 0.08 were taken to indicate close and acceptable levels of fit, respectively (Jöreskog and Sörbom, 1993). Keith (2006) proposed the following educational research benchmarks for direct effects in the form of beta coefficients: less than 0.05 is considered too small to be meaningful, above 0.05 is small but meaningful, above 0.10 is moderate, and above 0.25 is large. For indirect effects, we use Kenny's (2012) definition of an indirect effect as the product of two effects; using Keith's benchmarks above, we propose an educationally meaningful, small indirect effect = 0.003, moderate = 0.01, and large = 0.06.

This empirical evidence suggests different implications for educators who wish to make students competent, by working in primary prevention: (1) The concept of competence is multidimensional (Gagné, 2013); therefore, teaching and learning should work on all the levels of learning involved. (2) While information about alcohol is important–especially information that helps students construct an integrated, explanatory model of the effects of alcohol–excess information and inadequate modeling at a young age can be counterpreventive, with a dysregulatory effect. One notable effect was the inverse predictive relationship between adolescents' level of self-regulation and their degree of knowledge about alcohol. From the assumptions of the SRL vs. ERL Theory, this could be interpreted as an effect of a regulatory vs. dysregulatory context. In other words, adolescents who are growing up in a regulatory educational context (urban school and parents with secondary education) are trained to a higher level of self-regulation and they are protected from environments with excessive information about alcohol. The opposite would occur in dysregulatory educational contexts (marginalized school with compensatory education), where adolescents are allowed to have more direct contact with alcohol consumption, and early alcohol intake, or dysregulation, is encouraged. These results are consistent with the effect of the educational context at school, as seen in a previous dissertation report (Marcos, 2013), but research should continue to investigate this in more depth. (3) In addition, this evidence reaffirms the role and importance of planning for adolescents to learn self-regulation and self-discipline (Duckworth et al., 2011) – in association with proactive attitudes of substance rejection, since such attitudes are decisive to whether self-regulation and self-discipline will be practiced. Therefore, from the educational point of view of primary prevention, the experiences, information, modeling and attitudes offered by the educational context (family, school, peers) are determinants in encouraging regulation, nonregulation or dysregulation of the adolescent's alcohol intake (de la Fuente, 2017).

These results corroborate the idea that educational programs focusing on the primary prevention of alcohol intake should rest on two basic pillars: (1) an eminently informational thrust, for learning facts and concepts about the problem (also containing a developmental component); and (2) the more developmental thrust, referring to learning behavioral principles, behavior management skills or self-regulation, and the motivational or attitudinal aspect. Conceptual, procedural and competence would be impregnated with a holistic model of education, approaching each level of competency as different and needing to be approached differently. This question therefore deserves some profound analysis. In the first place, one must reinforce the importance of principles in alcohol use—norms, standards and goals for behavior (Pons et al., 1995). Without behavioral principles, we cannot determine the direction of self-regulation. It is therefore unadvisable to move toward preventive programs that have a large amount of merely conceptual information; rather, principles of behavior should be emphasized, as representations of reality and of oneself, especially in the stage of adolescence and pre-adolescence. In second place, training in self-regulatory behaviors—as a procedural meta-skill—is a must in educational programs for primary prevention of alcohol intake, given that this behavioral level has a clear effect on adjustment, as has been demonstrated in prior evidence from secondary and tertiary therapeutic programs (Fernández et al., 2009). Finally, there is no substitute for educational treatment of the attitudinal level. Attitudes must involve a desire for healthful behavior (Fergus and Zimmerman, 2005; García del Castillo and Días, 2007).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.