What is a healthy child? We have concluded that the positive perspectives provides a consensual answer to this question, but it is only a snapshot. We know relatively little about who these young people are except that they can be found in all walks of life. Urgently needed is a broader characterization of youth who are doing well—where do they come from, where do they go, and what are the choices made and routes taken in between? A good first step has already been taken by studies already underway that use epidemiological samples followed over many years (e.g., Hawkins, Catalano, & Miller, 1992).

We propose further studies of this sort that use the full array of positive measures and indicators now at our disposal. These studies should be patterned on the Terman (1925) study of adolescent geniuses and the Grant Study of the best and brightest of Harvard University undergraduates (Vaillant, 1977) in the sense that they be large scale—i.e., have big samples, longitudinal designs, and multiwave assessments—but not start with the most fortunate or the most privileged in our society. Dissemination of information about youth who are thriving might help combat negative stereotypes about teenagers. Realistic portrayals of young people, including their flaws and problems and how they cope with them, might inspire other teenagers to focus on what they do well and to eschew a victim mentality (Shih, 2004).

We propose that these studies of the natural history of positive youth development include, obviously, measures of positive characteristics (positive emotions, flow, life satisfaction, character strengths, skills, talents, and callings), mea

sures of risk, and measures of problems (negative emotions, risky behaviors, symptoms, and psychological disorders). It would be a shame if the positive psychology perspective leads researchers to repeat the error of business-as-usual psychology by ruling out a balanced view of youth and the adults they become.

Inclusion of both positive and negative measures over time allows the critical questions we have posed to be answered with hard data (cf. deVries, 1992). Do positive characteristics preclude recurrence of problems? Do they limit them? Do they allow youth to learn lessons from crises, episodes of disorder, and misfortunes? Which positive characteristics provide the best buffers against depression, substance abuse, or anxiety disorders?

A retrospective study we have done with several thousand adults asked respondents if they had ever experienced a severe psychological disorder and, if so, how well they had recovered from it (Park et al., 2003). We also measured their life satisfaction and various strengths of character. Individuals who had fully recovered from a disorder were just as satisfied with their lives as those who had never experienced a disorder. At least for some, there is light at the end of the psychopathology tunnel: “Tis an ill wind that blows no good.” And individuals who had fully recovered from a disorder also reported higher levels of specific strengths of character—i.e., appreciation of beauty, bravery, creativity, curiosity, forgiveness, gratitude, love of learning, and spirituality—compared to those who had never experienced a psychological disorder. Whether these character strengths were in place before the disorder and helped in recovery or whether they represent lessons learned during dark days is unclear from the research design; the need for a richer prospective study is implied (Linley & Joseph, 2004).

Prospective studies of psychological problems need to be informed by varying base rates of different disorders. Depression and substance abuse are so common in the contemporary United States that their eventual onset can arguably be investigated in unselected samples of several hundred youth. In contrast, other sorts of problems—e.g., schizophrenia, bipolar disorder, anorexia, and bulimia—are less common, which means that studies would need to oversample at-risk youth, but we stress that we are not calling for studies of only at-risk adolescents. That strategy would deny the premises of the positive perspective and preclude the lessons to be learned from charting the positive development of youth per se.

We are interested in an approach to psychological disorder that we dub “dealing with it,” or keeping on with life despite problems. Our interest was stimulated by conversations with those in the military about how they train personnel to perform optimally under the most extreme circumstances. How does a sniper learn to shoot accurately after crawling into a position and staying there for 48+ hours? How does a pilot learn to fly skillfully in a dizzying free fall? How does a submariner learn to live and work with others in extremely cramped quarters? Business-as-usual psychologists would probably target and then relieve the negative emotional states that accompany these circumstances—boredom, fear, anxiety, fatigue, and discomfort. But that is not how the military proceeds. They teach their personnel how to perform in spite of these circumstances. They teach personnel to deal with aversive states, to do what needs to be done regardless of how they feel at the moment.

If these examples are too militaristic to be compelling, then what of the identical lessons we learned from interviewing firefighters and paramedics who perform well—heroically, in our view—under frightening circumstances (Peterson & Seligman, 2004). In no case did anyone we interviewed say that they had eradicated their fear. Rather, they learned to do their job so well that their fear did not get in the way (cf. Rachman, 1990). One of our firefighters told us of rescuing an infant from a smoke-filled building: He lost control of his bladder, but never his grip on the baby. Extraordinary? Yes and no. Deserving of study are the more mundane among us who go to school or show up at work or raise our families even when we are depressed or anxious.

Crucial in studying youth from a positive perspective is taking into account the institutions that influence them—programs, organizations, and communities; friends and families; mental health professionals; and the media. Needed

here is an elaboration of institutions and their features that can be applied throughout the lifespan, and not just to youth (cf. Cameron, Dutton, & Quinn, 2003).

The data from such studies can be productively examined with the techniques of causal modeling (e.g., Connell, Gambone, & Smith, 2000; Gambone, 1997; Halpern, Barker, & Mollard, 2000; Walker, 2001). Sample sizes must be large enough, especially to discern interactions between and among variables. But with adequately powered designs, these models allow inferences about what might prevent what and why. As already emphasized, explicit theory is imperative to specify hypothesized links prior to causal modeling.

Some practitioners in the youth development field have called for extremely ambitious community-level interventions, in which all of the institutions that influence youth development would be explicitly programmed and linked. Interventions would target all children and adolescents and presumably last for years. In the abstract, we understand the sentiment behind this recommendation and agree that the links among different institutions and socializing agents deserve study in terms of their effect on youth development.

But in the real world, there are many objections to this research agenda (cf. Wandersman & Florin, 2003). On scientific grounds, community interventions cannot be easily manualized (i.e., explicitly described in detail and thus generalized), and if all youth in a given community are to be included, then what sorts of comparisons are possible? It is difficult to think of a meaningful control group to isolate the active ingredients of such global and enduring interventions. These problems can be surmounted, but it is still unlikely that a society with dwindling resources would be willing or able to initiate such grand interventions on a routine basis, which makes those already under way all the more worthy of attention. Thus with respect to urgently needed intervention studies, we believe that there are two promising research avenues to pursue that are somewhat more modest but infinitely more feasible.