The Generalizability and Specificity of Theory
Article Outline
See Related Article p. 224
Adolescents' transition from childhood to adulthood is plagued by risk practices for injury, disease, and death. This critical period involves a complex web of genetic, biologically programmed growth, and behavior derived from social learning processes. As such, investigators with expertise in biology, medicine, psychology, and other social sciences are required to combine a wide range of empirical findings that define a coherent theory of development and that would guide prevention of risk practices that lead to morbidity.
The Brook et al article in this issue of the Journal of Adolescent Health [1] is part of a larger effort to predict and ultimately prevent sexual risk practices among adolescents. This study contributes to the published data concerning adolescent drug use, sexual practices, and consequential morbidity by focusing on a special sample: adolescents whose fathers were drug-abusing or drug-dependent, and about half were HIV positive. This sample offered the investigators an opportunity to explore the sexual risk practices of adolescents in the context of rarely studied paternal addiction and HIV status.
This study was based on the Family Interaction Model (FIT), which is based on developmental processes, learning theory, and personality traits. This model focuses mostly on the intra-family dynamics for predicting adolescent risk behavior. As such, this model serves as a subset of larger models that include influences from family members, peers, media, and other environmental factors [2].
Brook and Brook [3] contributed to the published data by providing a more fully specified model of adolescent development and risk practices compared to many that preceded it. The “developmental” concepts of growth and their genetic foundations imply links with biology. Such links should be hierarchically related to biology, chemistry, and their many sub-specializations. Thus, the Brook et al conceptual model is moving toward an integrated theory with “natural sciences.”
However, as is always true of empirical studies, the operational procedures are curtailed with respect to the theoretical model to be tested. In this case, the etiological factors investigated were limited to paternal factors: the father's negative attributes and relationship with the adolescent. This operational model was underspecified by omitting theoretical “axioms” generally accepted as critical to adolescent's behavior. These include developmental levels, maternal practices, peer influences, as well as larger societal cultural effects. Although this study is not unique in testing an “underspecified” operational model, such under-specification can be viewed as limiting the “theoretical fidelity” of the study [4]. Thus, this study (along with most studies) sets the stage for a more complete test of theory, and ultimately a test of more complete theory. Such tests should strive to match operations to the specific theoretical model so as to achieve theoretical fidelity.
One factor (discrimination/victimization) included in the Brook et al analysis represented possible ecological influences on sexual risk behavior and again linked the study (and theory) to ecological concepts from sociology and biology. This factor probably represents discriminatory cultural practices toward adolescents which sometimes cause harms and/or influences risk practices.
Limited conceptual work has taken place to integrate learning theory with ecological factors. Current endeavors include research on the built environment and obesity [5], [6], and the Behavioral Ecological Model (BEM) [7]. The BEM emphasizes the role of contingencies of reinforcement as the unit of function at both the intrapersonal and the ecological level of influence (e.g., metacontingencies) for groups and whole populations. The BEM sets the stage for integrating biological/ecological concepts with principles of behavior. However, it does so without clear links to concepts such as personality traits (e.g., aggressive personality) which summarize behavioral patterns that evoke reactions (e.g., discrimination) from members of social networks, such as teachers and other adult supervisors. Such reactions probably contribute to changes in adolescents' aggressive behavior as well as other practices.
Thus, the FIT, BEM, and other models offer overlapping concepts of child development specific to sexual practices, but do not offer coherent language to describe common concepts. The common and discriminating features and the degree of consistency in language warrant clarification to advance the science of adolescent development. This is true for the myriad of psychobiological and psychosocial models in the field of child development and behavioral medicine.
The readers and investigators of Journal of Adolescent Health might provide a foundation from which greater cross-study consistency in language could be forthcoming; a foundation from which agreement might be reached about common concepts for which similar names could be assigned; and for differential concepts that distinguish one model from another. A consensus on these features would enable the research community to clarify common empirical findings that may look incoherent because of different languages. More importantly, consensus will enable future research to focus attention on the discriminating features of multiple conceptual models to identify the valid subsets and ultimately to create a common theoretical model that includes newly validated components derived from multiple models. This would enable the Brook et al study, as well as other investigators' research, to create a system that achieves consensus on validated axioms and produce more effective and efficient research that yields a more fully specified and generalizable theory of adolescent development and behavior.
Integrating behavioral and social sciences with basic biological sciences, pediatrics, and with ecological research, is daunting. Investigators in the field of adolescent health and medicine will agree that these sciences are required to achieve a fully specified theory of development. If our view of the interplay between empirical research and theory and the variations among investigators is correct, Journal of Adolescent Health might serve as a catalyst for codifying current theoretical models and determining their common and discriminating features. Consensus might lead to a composite model with recommendations for future studies to test the discriminating features among models. The exact mechanism by which the research community represented by Journal of Adolescent Health helps bring about a coherent language and more comprehensive theory to guide our research is an important task. We urge the editors, investigators in the field, and readers to enter into a dialogue by which we can clarify theories and enhance our science to increase our ability to prevent or minimize adolescents' risk practices and consequential morbidity.
References
- A longitudinal study of sexual risk behavior among the adolescent children of HIV-positive and HIV-negative drug-abusing fathers. J Adolesc Health. 2010;46:224–231
- A behavioral-ecological model of adolescent sexual development: A template for AIDS prevention. J Sex Res. 1994;31:267–281
- . The psychosocial etiology of adolescent drug use: A family interactional approach. Genet Soc Gen Psychol Monogr. 1990;116:111–267
- Enhancing theoretical fidelity: An e-mail-based walking program demonstration. Am J Health Promot. 2005;20:85–95
- . Physical activity and food environments: Solutions to the obesity epidemic. Milbank Q. 2009;87:123–154
- . Ecological models of health behavior. In: Glanz K, Rimer BK, Viswanath K editor. Health Behavior and Health Education: Theory, Research, and Practice. 4th edition. San Francisco, CA: Jossey-Bass; 2008;
- . The logical and empirical basis for the behavioral ecological model. In: DiClemente RJ, Crosby R, Kegler M editor. Emerging Theories and Models in Health Promotion Research and Practice: Strategies for Enhancing Public Health. 2nd edition. San Francisco, CA: Jossey-Bass Inc; 2009;
PII: S1054-139X(09)00702-2
doi:10.1016/j.jadohealth.2009.12.018
© 2010 Society for Adolescent Medicine. Published by Elsevier Inc. All rights reserved.
Refers to article:
- A Longitudinal Study of Sexual Risk Behavior Among the Adolescent Children of HIV-Positive and HIV-Negative Drug-Abusing Fathers , 21 August 2009
