Lewin's  statement that "there is nothing more practical than a good theory" might be used to characterize the top-down approach. It is assumed that the development of an internally coherent theory is important, not just for its theoretical elegance, but also because it allows for hypothesis driven research to test mechanisms underlying behavior change. Once these pathways and processes are identified, interventions can be developed. A key advantage in the top-down approach is that it lends itself to understanding how and why an intervention works, not only if.
The starting point of top-down development often involves a clear articulation of a meta-theoretical foundation, that is, a view on human kind that serves as a philosophical underpinning of one's theorizing. To illustrate, within SDT , an organismic viewpoint is embraced. It is assumed that individuals have a natural inclination to be active and to move their lives in desired and specific directions rather than being passive and completely subjected to environmental forces that push them around (e.g., through the use of punishments and rewards). Through this activity, people would move towards increased integration, growth and wellness given sufficient supportive circumstances. As growth-oriented organisms, individuals stand in a continual interface with the social environment, which can either support and facilitate the unfolding of this natural growth process or block it by frustrating people's basic psychological needs  for autonomy (i.e., experiencing a sense of psychological freedom), competence (i.e., experiencing a sense of effectance) and relatedness (i.e., experiencing a sense of intimacy and connectedness).
The articulation of this meta-theory provides the foundation for developing a formal theory in a systematic, research-driven way within the top-down approach. New ideas get naturally and steadily integrated into the theory following sufficient empirical support, which helps the theory to maintain internal consistency. To use a metaphor, theory development with the top-down approach resembles the construction of a puzzle . Over the years, new pieces of the puzzle get only added once their fit is determined. Such evolution seems to typify the development of SDT, whose five mini-theories (i.e., cognitive evaluation theory, organismic integration theory, causality orientation theory, basic psychological need theory, and goal content theory) have been gradually developed over the past four decades . As the theoretical puzzle is growing, over time, the theory might be perceived as highly complex. However, when the internal logic of the theory is fully understood, SDT is characterized by a sense of parsimony and elegance, as a minimum set of concepts is used to explain a wide variety of phenomena across age groups, life domains, and cultures. This is because new pieces are added to the theory when (a) the new piece represents a "compelling theoretical necessity", that is, when a clear and meaningful argument can be provided for the enlargement of the theory, and (b) when sufficient empirical support is gathered for the theoretical claim. To illustrate, whereas the satisfaction of the basic needs for autonomy and competence was considered critical for the maintenance and development of intrinsic motivation , it is only when the process of internalization - that is, the personal endorsement of externally offered norms, values and regulations - was studied that the basic need for relatedness received a more prominent place . Indeed, people often engage in enjoyable and intrinsically motivating activities by themselves (e.g., reading), that is, without significant others being present to support them, while the development of satisfying relations is critical for individuals to fully endorse and willingly subscribe (i.e., internalize) to norms and behaviors they don't particularly find interesting, and, as a result would not spontaneously engage in [1, 12]. Because most health related behaviors (e.g., stopping smoking, getting a mammogram to screen for cancer) fall into this category, SDT-based interventions involve the facilitation of internalization through the support of relatedness in addition to autonomy and competence. Several studies have confirmed the importance of relatedness satisfaction for the internalization process .
Due to the focus on the explanatory processes underlying change in the top-down approach, it becomes possible to a priori predict specific social-contextual variables that promote change. Within SDT, for instance, counselors are encouraged to support clients' basic psychological needs as much as possible as the experience of need satisfaction leads to a greater integration and anchoring of change into one's life style and values, resulting in maintained change . More specifically, the concept of basic need satisfaction allows delineating specific motivating counseling techniques (e.g., shared agenda setting, choice etc.) and orientations (e.g., empathic style, unconditional regard) (see Patrick & Williams this issue for a more elaborated discussion). Dozens of self-report and experimental studies within SDT have studied such specific need-supportive practices. To illustrate, choice provision is considered as a key strategy to support participants' volition. Such need-supportive strategies were first studied in the laboratory where typically university students would serve as study participants [15, 16]. Based on these basic lab-based findings, studies were set up to test the role of choice across several real world domains where it was found to yield manifold advantages, including the facilitation of heart disease management among senior women (> 60 years; ), the reduction of drop-out rates among eating disorder clients , the promotion of well-being among elderly who stay in nursing homes , and the increase of vitality among students in PE classes [20, 21]. Thus, real-life experiments complemented the initial laboratory studies, thereby increasing the ecological validity of the theory.
Based on the examination of different experimentally isolated need-supportive components (e.g., choice, rationale provision, etc.) and the development of reliable measures to assess theoretical constructs, SDT scholars more recently proceeded to set up controlled clinical trials in which a theory-driven intervention is tested against a control group. An SDT driven intervention was composed by combining previously isolated need-supportive components. For instance, while rationale provision [22, 23], empathy , and autonomy-supportive vs. controlling language [25, 26] had been examined in isolation in experimental studies, these and other need-supportive facets were taken together to create a need-supportive intervention. Over the past several years, an increasing number of SDT based controlled clinical trials in a wide variety of domains, including physical activity (, Fortier, Duda, Guerin, & Teixeira, this issue), tobacco dependence , dental care  and weight loss  have been conducted . The advantage of such controlled clinical trials is that they further confirmed the ecological (or external) validity of SDT in the real world, and established the efficacy of the interventions.
However, this intervention research is still in its infancy phase and, hence, various issues deserve further attention. First, because various need-supportive components are simultaneously manipulated in controlled clinical trials, it remains unclear which of the components are driving the observed effects, an observation that also applies to MI. Future studies may want to isolate these various need-supportive components, as done by Deci, Eghrari, Patrick, and Leone  and Sheldon and Filak  in laboratory studies as to examine main and interaction effects between those isolated components. This would help to answer the question whether the whole is more than the sum of various need-supportive components; for instance, within SDT, it is maintained that for competence support to yield a truly vitalizing effect such competence support is best provided in an autonomy-supportive rather than controlling way , thus suggesting that positive feedback and autonomy-supportive language can best be simultaneously provided. Second, the efficacy and effectiveness of a SDT intervention relative to other psychotherapeutic approaches has not extensively been examined yet, as most of the available controlled clinical trials contrast the experimental group with a no-treatment control group.
Third, although various key ingredients of a need-supportive intervention have now been identified (see Patrick & Williams, this issue), this list remains open for additions and refinements. This is because within a top-down approach as SDT cross-fertilization between basic laboratory studies and (experimental) field studies is pursued. Through laboratory studies, the effects of isolated need-supportive components can be examined, which then might provide the basis for refining the intervention approach. To illustrate, although choice has on average a motivating and vitalizing impact, laboratory studies  have demonstrated that choice can be provided in a controlling rather than an informational way such that participants do not experience a subjective sense of choice, willingness, and psychological freedom in spite of being offered an objective array of options to choose from. Also, an overload of options to choose from has been found to undermine the motivating impact of choice as people are less likely to feel effective to make an accurate choice in such a case .
Such research is likely to have important counseling implications: although a clinician might at the surface offer a client who hesitates to stop therapy the choice to do so, the client might feel pressured to continue the counseling if this choice is provided in a controlling and anxiety-provoking way (e.g., "it is up to you to choose to leave treatment or to stay but I can tell you that many clients who leave treatment fail to make it on their own"). Regarding the issue of option overload, although a doctor might provide a number of options to a client how to deal with his situation, the client, already anxious about his health condition, might feel overwhelmed and not have the energy to process all of the information independently. In such a situation, the autonomy-enhancing effect of choice would be diminished by its anxiety-inducing and competence-thwarting effects. What is important in such a situation is not to take away the choice, but to structure the situation such that the provided options become manageable for the client so that he ultimately feels competent to choose and fully endorse one option.
Such basic experimental research is considered fundamental within SDT as it helps to understand how social-contextual factors can simultaneously affect the satisfaction of multiple basic needs. This basic research provides more detailed insight in the need-supportive practices that are characteristic for a SDT intervention and the way how they need to be introduced during counseling to be optimally motivating.