Journal of Rural Community Psychology Volume E7 Number 2 Fall 2004
Convergence Between Psychology
and Public Health in a Rural Community
Terrence J. Schwartz
This paper describes several areas of convergence between public health and psychology through description of activities in a community health board. Community, organizational, and clinical psychology provide examples of the convergence. A brief historical review is provided of the board’s efforts to identify and address four problem behaviors. Psychological theory and research offer several implications for advancing conceptualization and intervention for the problem behaviors. The paper concludes with an examination of the utility of psychological theory and research for augmenting community-based outcome assessment.
The many changes proposed and implemented in health care policy over the past decade have served to widen the range of disciplines relevant to public health. Beyond the traditional roles played by medical providers and health departments, economists and sociologists have added their voices to the expanding debate on public health issues. Psychologists, too, have come to realize that many of these issues may be addressed by theories and research in such diverse areas as community, clinical, and organizational psychology (Cavaliere, 1995; Leviton, 1996).
A substantial body of recent literature has illustrated the relevance of psychological topics to public health issues. Comprehensive reviews include those by Lorion, Iscoe, DeLeon, and VandenBos (1996), Coreil, Bryant, and Henderson (2001) and Schneiderman, Speers, Silva, Tomes, and Gentry (2001). Specific areas of overlap between psychology and public health include topics as diverse as injury prevention (Durlak, 1997), breast cancer screening (Banks, Salovey, Greener, Rothman, Moyer, Beauvais, and Epel, 1995), violence (Cooley-Quille and Lorion, 1999), and religion (Becker, 2001). The purpose of the present article is to add to this growing body of literature by documenting several connections between psychological issues and a program addressing public health issues in a northwestern state. The author participated in the activities of a citizens’ board charged with the task of identifying and addressing public health issues in a rural community. This paper will first outline the history of the statewide effort to address these issues. Particular activities of the board are then examined with the purpose of discerning issues with important psychological implications. At the conclusion of this article, the reader should gain a specific perspective on the interaction between psychology and public health through analysis of this case study. Additionally, the article will advance certain refinements and extensions of public health efforts suggested by recent trends in psychological theory and research.
A Brief History
The Washington State Legislature passed the Youth Violence Reduction Act (1994) with the primary intent of reducing the rate of youth violence in the state. A secondary intent was to reduce the rate of seven identified problem behaviors: child abuse and neglect, domestic violence, school dropouts, teen substance abuse, teen suicide attempts, violent delinquent acts, and teen pregnancy and male parentage. A primary sentiment guiding this legislation was the establishment of “grass roots” or local efforts to address the primary issue of youth violence and the associated problem behaviors. The legislature created the “Community Public Health and Safety Networks” to address these behaviors. Citizen boards were created to represent the population within the boundaries of these Networks, often but not exclusively identified as counties within the state. Legislative mandates to these boards were multiple and complex. Local boards were required to identify several of the most prevalent problem behaviors within their boundaries. They were then expected to develop a comprehensive plan to address the behaviors. Part of this plan was a description of risk and protective factors affecting the selected problem behaviors. This was combined with an effort to develop “data-based” descriptions of the behaviors. An allied requirement included the designation of measurable outcome criteria for social service programs whose efforts might reduce the problem behaviors. Finally, the local boards were required to monitor the results of any such program efforts.
There are many areas of overlap between the activities of the Networks and psychological issues. Even a cursory examination of the list of problem behaviors should prompt psychologists of many different persuasions to consider their potential contribution to public health issues. Principles of community and organizational psychology underlie the formation and activities of the boards. The process of the identification of prevalent problem behaviors involves principles from the psychology of judgment and decision-making. Plans to remedy these behaviors can be linked to efforts within counseling and clinical psychology to reduce behavior disorders. Allied with this effort are important presumptions about the origin and maintenance of social problems. This may intersect with research efforts that attempt to formulate multivariate models of complex behavioral sequences. Altogether, these areas of convergence between psychology and the Networks’ activities provide the outline for this paper. Along with noting the convergence, recommendations for enhancing and refining the public health approach will be advanced.
Convergence With Organizational and Community Psychology
Considerations Related to Organizational Psychology
The structure and function of the citizens’ boards represent subjects of study within organizational psychology. Legislative efforts in establishing the Network follow important principles within this discipline. As with any effective intervention, social policy should be formulated upon a comprehensive needs assessment. The legislature’s charge to communities to identify problem behaviors within their boundaries is clearly illustrative of this principle. Such identification is best accomplished by stakeholders with access to appropriate data. An important distinguishing feature of these boards was a preference for local control and self-determination. Recruiting efforts for board membership have been very successful, resulting in at least 10,000 volunteers statewide (Kinney, Haapala, and Happy, 2002). The boards reviewed extensive data on the incidence of the seven identified problem behaviors within their boundaries. The initial selection of four behaviors by the Network in this case study (child abuse/neglect, domestic violence, teen substance abuse, and youth violence) was based upon several considerations. These included a higher incidence of child abuse/neglect and teen substance abuse within the county than within the state, increased arrest rates for domestic violence within several municipalities of the county, and increased incidence of violent acts by youth within the county.
The structure of the community boards provides a second opportunity to observe convergence with organizational psychology. Each “Network” is actually all communities, school districts, parents, youth, government agencies, and social service organizations within certain geographic boundaries. A board of 23 members represents this extensive Network. One of the first difficulties faced by the boards was the coordination of the activities of so many members. In an effort to address this issue, the board enlisted as a consultant a social and organizational psychologist from a local university. He observed that the board was too large to conduct its business successfully. He recommended that the board subdivide itself into several committees, each charged with executing functions associated with initial tasks mandated by the legislature. The board was receptive to this idea and created two task forces, one involved with data gathering and another with community relations.
A third observation related to the structure of the Network boards relates to important principles of group dynamics. Following legislative direction, the 23-member board was divided into 13 non-fiduciary and 10 fiduciary members. A fiduciary member was defined as one who represented local agencies including law enforcement, social services, health care services, or local Indian tribes. Non-fiduciary members were local residents not affiliated with government agencies. The legislature clearly intended that the views of parents and citizens prevailed in the deliberations of the Network boards (Youth Violence Reduction Act of 1994). Apart from the political sentiment for local control and “grassroots” organizing, this distribution of membership also illustrates principles of group dynamics related to groupthink and persuasion. Groupthink is prominent in highly cohesive groups of individuals who share a common background (Janis, 1982). This description might well have characterized a Network board whose membership was restricted to individuals commonly associated with public health concerns (i.e. the “fiduciary” members). By requiring a majority of non-fiduciary members, the legislature has invoked at least one useful strategy to avoid groupthink by including individuals who are outside the health service professions in the decision-making process. Such inclusion may avoid the insulation from diverse viewpoints and resultant constriction of discussion that often characterizes cohesive, homogeneous groups.
It is not clear, however, that the heterogeneity of group membership required by the original legislation has been effective. Important issues of power and persuasion potentially accompany this allocation of membership status. Despite the numerical superiority of non-fiduciary board members, the fiduciary members may well carry greater power by reason of their status as professionals in health-related issues. Their influence upon the decision processes of the non-fiduciary members may serve to obviate the legislative intent. Without specific training in public health issues, the non-fiduciary members may yield to the specialized, professional training of the fiduciary members especially when ambiguous or controversial issues arise. No such training was specifically mandated for non-fiduciary members. The Family Policy Council, which oversees the Networks at the state level, has promoted efforts to provide new member orientations that promise to address board member roles, as well as issues of inclusion and respect. This certainly appears to be a step in the direction of acknowledging influence issues within the Network Boards. The distribution of membership status presents a wonderful opportunity for a qualitative study of group processes in decision-making.
Interaction Between Organizational and Community Psychology
The psychology of judgment and decision-making has relevance to the process of problem behavior identification. This process takes place within a political and social environment of competing interests and incomplete data. It is precisely this type of environment that produces an atmosphere of uncertainty, potentially eliciting the operation of biases and heuristics so well characterized by Kahneman, Slovic, and Tversky (1982), Bell, Raiffa, and Tversky (1988), and Piattelli-Palmarini (1994). With respect to the availability heuristic, for example, the focus is upon the nature of the decision criteria for selection of the problem behaviors. Readily available decision criteria may contribute disproportionately to the selection of a problem behavior, with such bias potentially diminishing the validity of the selection process. The primary data presented to the board for evaluation of problem behaviors was frequency or incidence data. While relative frequency or increased incidence is one possible criterion for selection, it is not the only one. The simple fact that the incidence of a behavior is greater within one county does not in itself describe the impact of that behavior upon public health. Frequency data do not necessarily reflect the magnitude of the specific health problem, particularly one as broad as “youth violence”. More notably, frequency data do not represent the costs of addressing or ameliorating the problem behavior. Costs may be diminished or increased depending upon factors such as the availability of services or the propensity of affected individuals to seek available services. Service access is a particularly important issue in rural areas. Substance abuse prevention and treatment services, for example, are typically located near large municipalities with a sufficient number of clientele to maintain their operation. If teen substance abuse is targeted for reduction as a problem behavior and treatment facilities are inadequate, the likelihood of successfully addressing it is seriously lowered. At minimum, frequency data should be combined with cost analyses and resource availability to more fully represent the public health impact and potential for successful amelioration. This approach also fosters greater clarity in problem identification and reduction. The incidence of a problem behavior is as much a problem as the scarcity of appropriate interventions. Additionally, reliance upon incidence may promote the treatment of the selected behaviors as independent factors. Psychological research, however, clearly points to a number of interactions among the problem behaviors. Adolescent suicide and substance abuse, for example, are known to have significant association (Berman and Schwartz, 1990). Although the involvement of stakeholders in selecting problem behaviors is congruent with sound principles of community and organizational psychology, the nature of the decision process for such selection may benefit from a more detailed analysis. One contribution of the psychology of judgment and decision making to public health, therefore, lies in the generation of additional decision criteria affecting choice of the problem behaviors.
Accurate identification of problem behaviors is only the first step, however, in effective intervention. After identification, a large array of potential interventions is potentially available. Choice among these interventions requires consideration of their appropriateness for specific populations. An important characteristic of local populations is their receptivity to interventions. Help-seeking behavior may differ greatly between urban and rural populations. Individuals in the rural environment often express a traditional belief in self-sufficiency that is a significant barrier to seeking appropriate treatment (Kreitlow, 1988). However, most research on intervention methods is conducted on urban populations. The results of interventions for such populations may not provide appropriate indicators for the effectiveness of the intervention in a rural population. The interventions may simply not generalize. An additional issue concerns the distinction between efficacy and effectiveness research (Street, Niederehe, and Lebowitz, 2000). Efficacious interventions have shown treatment gains in controlled laboratory or clinical studies. Effective interventions, however, are those exhibiting treatment gains in community or practice environments. Neither receptivity nor effectiveness was explicitly acknowledged in the selection of interventions.
Assuming accurate problem identification and appropriate selection of an intervention, a third issue arises around policy implementation. Effective implementation requires at least four elements: access to the target population, sufficient resources, careful monitoring, and timely revision in response to monitoring. The second of these elements is often the most crucial, particularly with respect to monetary resources. The legislature created the Networks as part of an effort to reform social services for children and families. A critical issue emerged a few years ago that illustrates the crucial role of such resources. Initially, two funding sources supported the Networks. One proceeded from a federal program, the Family Preservation and Support Services Act (Title IV-B). Another originated from a state program, the Violence Reduction and Drug Enforcement account. During the 1997 legislative session, the state support was withdrawn, leaving the Networks dependent upon federal funding. Allocation of the federal funds, however, was intended solely for programs addressing child abuse and neglect. Many Networks had chosen to focus on other problem behaviors. They found it necessary to reformulate their comprehensive plans in an effort to qualify for this funding. Finally, some Networks found that limitations on administrative funding imposed by federal spending guidelines necessitated substantial budgetary reorganization, resulting in staff layoffs and deferred expenditures. At the time of this event, several Networks considered dissolution as a result of this funding alteration. In effect, this funding shift created an “unfunded mandate” to the Networks with respect to any problem behaviors other than child abuse and neglect. There have been subsequent, successful efforts to redress this resource issue, yet this example illustrates a significant barrier to effective policy implementation, one anticipated by models of effective organizational behavior from psychology.
Convergence with Clinical Psychology
Once the Network had selected a set of problem behaviors, it was necessary to identify associated risk and protective factors as required by the original legislation. This important and complex task bears a strong correspondence to efforts within psychology to discover the origin and maintenance of various disorders. The attempt to discern causal variables associated with personal or social maladjustment is an essential part of the history of psychology and the social sciences. It has promoted the development of numerous theories of personality and psychopathology, theories which are frequently merged in clinical psychology. A brief review of the Network’s efforts to model the problem behaviors is described below, followed by a discussion of relevant psychological literature.
Network Efforts to Identify Risk and Protective Factors
The legislature mandated that the Network assume a public health perspective in problem assessment and intervention. More specifically, that perspective employs the concepts of “risk” and “protective” factors. Risk factors are conditions that increase the likelihood of problem behaviors; protective factors are conditions that act to prevent those behaviors. For example, risk factors for youth violence may include a family history of abuse or crime; protective factors may include community norms proscribing violent behavior. While it is not possible in this article to survey all the risk and protective factors identified by the Network for each problem behavior, the examination of the factors impacting one behavior in particular (child abuse and neglect) will illustrate the possibilities and limits of this approach. It is also one of the original four problem behaviors that was a consistent focus of the Network’s efforts.
The following information regarding the risk and protective factors surrounding child abuse and neglect was extracted from the Network Plan (Kittitas County Community Public Health and Safety Network, 1997). Risk factors were (1) lack of parenting and family management skills and (2) lack of clear behavior standards. Protective factors were (1) parent (family) training and/or mentoring and (2) opportunities for enhanced participation and ownership in the community. Although these factors were chosen as the basis of the Network’s initial response to child abuse and neglect, they are only an approximation of the complex causal environment surrounding this behavior. Knowledge of the research literature in fields such as psychology and social work considerably augments this chain and enhances the likelihood that intervention efforts may successfully address this problem. It is particularly important to realize that the “risk factor chain” for abuse may differ from that for neglect, especially with regard to parental cognitions (Sattler, 1998). Additionally, different “chains” probably exist for different subtypes of abuse (physical, emotional, sexual). Even if we were to limit our view of this problem to the chain outlined above, much work needs to be done in specifying the underlying causal elements. The first risk factor described as “lack of parenting and family management skills” fails to fully specify a direction for intervention. Wolfe (1987), for example, suggests the following specification. For “lack of parenting and family management skills”, an allied risk factor may include poor coping skills for stressful life events. Protective factors may include a supportive spouse, economic stability, and success at work or school. Additional specification of the risk and protective factors provides greater promise for effective interventions.
Extensions and Refinements of the Risk and Protective Factor Model
As noted above, the public health approach embodied in the risk and protective factor model is an attempt to formulate a causal model of a problem behavior. As such, it raises a number of important questions surrounding the possibilities and limits of such models. At the outset, it should be clear that social policies and programs are most effective when their missions and interventions are based upon defensible causal models. Policies and programs founded upon an incomplete model of the social, political, or psychological environment may compel allocation of scarce resources to non-productive efforts. It is essential that policy decisions proceed from at least a provisional foundation of established research findings supporting the causal model they employ. One of the primary contributions of psychology to public health is to assess the adequacy of this model. Recent work on causal models of psychopathology and behavior disorders suggests limitations to the risk and protective factor approach.
It is important to consider that a risk and protective factor model (Catalano and Hawkins, 1996) is a limited model. It is limited in one sense that the risk factors are comprised of only two classes of causal variables. From the standpoint of a comprehensive behavioral assessment model, however, Haynes (1992) describes six such classes. This author describes risk factors as belonging to the classes of original and triggering causes. Haynes describes original causes as the initial elements in a sequence of causes leading to a behavior. Triggering causes are similar to antecedents in operant conditioning in that they occur immediately before the behavior. Other classes of causal variables described by Haynes include maintaining, magnitude (or duration)-affecting, topographic, and relapse-affecting variables. Maintaining causal variables are those that account for the current manifestations of the problem behavior. “Triggering” causes exist within this variable class. As such, they may be distinct from the original or “first cause” variables. Magnitude (or duration)-affecting variables influence the severity or longevity of a problem behavior. Topographic variables describe the qualitative aspects of a problem behavior. A taxonomy describing these variables may include motoric, cognitive, affective, or physiological aspects of an individual’s response. Finally, relapse-affecting variables influence the likelihood of recurrence of the presenting behavior. It is especially notable to consider how the risk and protective factor model may lead to a neglect of causal variables affecting relapse episodes. Over reliance upon this model may lead one to assume that if risk factors are eliminated and protective factors maximized (“accentuate the positive, eliminate the negative”), then the problem behaviors will vanish. This assumption is clearly questionable and begs the question of how one should manage the problem behaviors that will almost certainly recur given inevitable errors in any causal model of behavior. While it is likely that a certain degree of overlap exists among variables affecting post-treatment relapse and those affecting pre-treatment problem behaviors, it cannot be assumed that the same causal model will characterize these two conditions. Treatment strategies following relapse may differ substantially from those appropriate for initial intervention. The discernment and inclusion of causal variables affecting relapse is a necessary addition to the risk and protective factor model.
A similar differentiation of causal variables is made by Lyons and Kisiel (1996) who list three causal classes pertinent to domestic violence: predisposing, precipitating, and maintaining. They describe predisposing causes as characteristics of individuals or their environment that may exist prior to any violent act. Precipitating causes are characterized as occurring at the initiation of violence, while maintaining factors are responsible for continuation of the violent cycle. Risk factors, therefore, offer only a partial representation of the complex environment of causal variables surrounding problem behaviors. A more complete description of that environment would include other classes of causal variables.
The risk and protective factor model encounters additional limitations. Several of these proceed from consideration of general limitations on causal inferences summarized by Haynes and O’Brien (2000). The model may require the assumption of invariance of cause across individuals with similar presenting problems. As such, it may fail to adequately consider individual differences in the classes of original or maintaining causal variables. For example, individuals may present a risk of child abuse or neglect but the model may require that causes affecting that risk are invariant across the sample of affected individuals. Consequently, the same intervention may be employed for clients apparently presenting with the same disorder, but with possibly different causal variables. Individuals may differ in their exhibition of topographic variables—some may present with cognitive accompaniments to violence or abuse while others may present primarily with physiological arousal. Haynes and O’Brien also contend that causal relationships may operate within a limited domain. This refers to the personal or environmental conditions under which the relationship between a causal variable and the associated behavior is evident. Additionally, they stress the notion that causal relationships are almost inevitably dynamic, that they can change over time. At the very least, this would demand that assumptions regarding risk or protective factors be periodically reviewed and updated. Finally, Haynes and O’Brien include the caveats that causal relationships may be subjectively determined and should be viewed as hypotheses in continual need of refinement and revision.
Other Perspectives on Causal Models of Behavior Disorders
To this point, this review of refinements to the risk and protective factor model has focused on perspectives from causal modeling and behavioral assessment. Although these areas of study have important implications for characterizing behavior disorders, other disciplines have pursued similar efforts. A minimal model is one proceeding from the field of social epidemiology (Coreil, Bryant, and Henderson, 2001). These authors propose a “causality continuum” comprising at least three categories of risk factors: distal, intermediate, and proximate. This continuum reflects the commonsense notion that some causes are more closely allied with a given behavior and some more peripherally allied. In their model, the proximate variables exert the former influence and distal variables exert the latter one. An important implication noted by these authors is the differential capability to influence causal factors along the continuum. As factors become more distal, efforts to alter them become less effective. They note that such a relationship has prompted more public health efforts to address individual behavior change and fewer efforts to alter more distal variables such as cultural, economic or political elements. As such, this is a minimalist causal model.
Intervention and prevention efforts can be differentiated in a fashion analogous to the causality continuum of Coreil et al. (2001). If a range of causal factors is implied in the etiology or maintenance of a problem behavior, then an analogous range of interventions should be available to address those factors. Durlak (1997) proposed such a model for the prevention of childhood problems. This author outlines five levels of intervention, ranging from one involving individuals (“person-centered”) to four levels of “environment-centered” interventions (school, family, peers, and community). These distinctions parallel the proximal-intermediate-distal continuum of causal factors outlined above.
A final perspective on the development of behavior disorders proceeds from a critique of primary care medicine by Felitti (2002). The Family Policy Council, which oversees the operations of the Networks, has recently focused on this developmentally-oriented model of public health issues. Felitti essentially advances the argument that several adult medical problems (e.g. chronic obstructive pulmonary disease, intravenous drug use) may have their onset in adverse early childhood experiences. In conjunction with researchers at the Centers for Disease Control and Prevention, he has conducted a series of studies linking events such as childhood physical or sexual abuse and family dysfunction to the emergence of organic disorders in adulthood. This approach seeks to identify elements of the distal category of risk factors of Coreil et al. (2001). As such, it offers another opportunity to expand the range of causal variables and emphasize primary, rather than secondary or tertiary, prevention efforts.
Outcome Assessment in the Network Plan
Assessment of the effectiveness of interventions is an important component of program evaluation attempted by the Network. Early in the history of the Network, an important distinction was made between “outputs” and “outcomes”. A historical tendency to view “outputs” or services delivered (e.g. allocation of money or personnel) as representing the effects of interventions in social programs was challenged by a methodology that focused instead on “outcomes” (Knapp, 1999). This latter category was described as short- or long-term data reflecting progress on diminishing risk factors, enhancing protective factors, or reducing problem behaviors. For example, a short-term outcome may be increased reports of family cohesion among affected families while an associated long-term outcome would be a decrease in the incidence of child abuse or neglect. An important result of this fundamental re-orientation from outputs to outcomes was the formulation of a “Statement of Work” for each Network, a document which attempted to explicitly link problems, interventions, and outcomes. A supplement to this document, known as the “Service Report,” summarized individual projects supported by the Network, along with their measurable results, frequency and methods of data collection, and outcome indicators. All of these categories were intended to describe the efforts of the Network in ameliorating one or more of the selected problem behaviors. In like fashion, the Network utilized “logic models” to conceptualize movement from intervention strategies to outcomes.
The Networks have also begun to implement further advances in outcome assessment (Family Policy Council, 2002). An important distinction has emerged between “direct” and “indirect” outcomes, one that may supplement the distinction between short- and long-term outcomes. Direct outcomes proceed from the intended, specific purpose of a program or activity. Education about peer resolution or mediation within a school system, for example, should produce measurable changes in the knowledge or skills of such practices in the target population. The concept of indirect outcomes, though, takes a much broader view of effectiveness engendered by a collective system of services. Parent education programs, for example, may accompany peer mediation efforts, potentially resulting in a synergistic effect upon problem behaviors such as youth violence. This linkage between direct and indirect outcomes represents an important effort to establish an interlocking system of interventions, one that mirrors the complex network of causal variables.
This effort towards outcome evaluation represents a significant advance in providing accountability for social interventions. However, the range of outcome variables assessed by the Network is limited. For example, it is often restricted to parameters such as decrements in the incidence of risk factors and problem behaviors or increments in the rates of certain protective factors. The example of substance abuse is illustrative. Although social programs target reduction or elimination of problem drinking, for example, many other outcome variables could be affected by program efforts. These include duration, magnitude, variability, time to relapse, and rate of change of drinking behavior. Many measurable parameters other than frequency of problem behaviors, therefore, may be influenced by the alteration of causal variables such as risk and protective factors. As more parameters become available for measurement, the program evaluation of Network efforts may more accurately reflect the intervention outcomes.
The concept of clinical significance promises additional clarity in outcome measurement (Jacobson and Revenstorf, 1988). This concept views a successful outcome as a client’s return to normative levels of functioning. Thus, a teenager suffering from problem drinking may not be considered clinically “changed” if only abstinence had been achieved. The absence of disease or dysfunction is not a sufficient outcome criterion; instead, the client must have moved into a “functional distribution” (Jacobson and Truax, 1991). Further, clinical change should include some reliability criterion such as the “reliable change index”. This approach to outcome assessment represents some of the best psychological theory and research now available. Unfortunately, the public and private agency “infrastructure” which might provide raw data for the analysis of functional and dysfunctional populations is not well-established. Much work needs to be done to establish data collection methods that allow use of these sophisticated outcome measures. Agency management and staff must be educated in the use of measurement models that specify change more precisely.
Public health is an essential component of social efforts to reduce maladaptive behaviors. Several branches of psychology make important contributions to these efforts. Several important principles emerge from the present case study analysis. First, the identification and conceptualization of social problems should be accorded greater scrutiny. The heterogeneous composition of the Network boards has been a necessary first step, though group composition is not a sufficient guarantor of validity in the problem selection process. Enhanced awareness of certain principles from social and organizational psychology, as well as the psychology of judgment and decision-making may further this effort. Second, research on the etiology of behavior disorders in community populations needs greater emphasis. For example, we cannot assume that models of etiology or intervention from urban populations can be applied without modification to rural populations. Third, the risk and protective factor model in public health, although a useful conceptual tool, should be applied with full acknowledgment of its limitations. Increasingly sophisticated causal models of psychopathology have emerged from the work of Haynes (1992) and Haynes and O’Brien (2000). These models should be accorded greater consideration in the conceptualization and implementation of interventions for public health problems. Finally, outcome measurement of public health interventions should be supplemented by inclusion of multiple parameters of problem behaviors. Public health program evaluation could benefit greatly from recent advances in statistical methodology that more clearly specify the clinical significance of interventions.
Banks, S., Salovey, P., Greener, S., Rothman, A., Moyer, A., Beauvais, J., & Epel, E. (1995).
The effects of message framing on mammography utilization. Health Psychology, 14,
Becker, D. M. (2001). Public health and religion. In N. Schneiderman, M. Speers, J. Silva,
H. Tomes, & J. Gentry (Eds.), Integrating Behavioral and Social Sciences with Public
Health (pp. 351-368). Washington, DC: American Psychological Association.
Bell, D.E., Raiffa, H., & Tversky, A. (Eds.). (1988). Decision making: Descriptive, normative,
and prescriptive interactions. New York: Cambridge University Press.
Berman, A. L. & Schwartz, R. (1990). Suicide rates among adolescent drug users. American
Journal of Diseases of Children, 144, 310-314.
Catalano, R. & Hawkins, J. (1996). The social development model: A theory of antisocial
behavior. In J. Hawkins (Ed.), Deliquency and crime: Current theories (pp. 149-197).
New York: Cambridge University Press.
Cavaliere, F. (1995, July). APA and CDC join forces to combat illness. APA Monitor,
Cooley-Quille, M. & Lorion, R. (1999). Exposure to violence: Preliminary evidence of psycho-
physiological links for urban youth. Journal of Community Psychology, 27, 367-376.
Coreil, J., Bryant, C., & Henderson, J. (2001). Social and Behavioral Foundations of Public
Health. Thousand Oaks, CA: Sage Publications.
Durlak, J. (1997). Successful Prevention Programs for Children and Adolescents. New York:
Family Policy Council. (2002). Washington State field guide: Reviewing Community Efforts
Compiled for Washington’s Community Public Health and Safety Networks. Family
Policy Council: Olympia, WA.
Felitti, V.J. (2002). The relationship between adverse childhood experiences and adult health:
Turning gold into lead. The Permanente Journal, 6, 44-47.
Haynes, S. N. (1992). Models of Causality in Psychopathology: Toward Dynamic, Synthetic, and
Nonlinear Models of Behavior Disorders. New York: Macmillan Publishing Company.
Haynes, S.N. & O’Brien, W. H. (2000). Principles and Practice of Behavioral Assessment.
New York: Kluwer Academic/Plenum Publishers.
Jacobson, N.S. & Revenstorf, D. (1988). Statistics for assessing the clinical significance of
psychotherapy techniques: Issues, problems, and new developments. Behavioral
Assessment, 10, 133-145.
Jacobson, N. S. & Truax, P. (1991). Clinical significance: A statistical approach to defining
meaningful change in psychotherapy research. Journal of Consulting and Clinical
Psychology, 59, 12-19.
Janis, I. L. (1982). Groupthink: Psychological Studies of Policy Decisions and Fiascoes.
(2nd ed.). Boston: Houghton Mifflin.
Kahneman, D., Slovic, P., & Tversky, A. (Eds.). (1982). Judgment Under Uncertainty:
Heuristics and Biases. New York: Cambridge University Press.
Kinney, J., Haapala, S., and Happy, E. (2002). Family policy system in Washington—
Community Public Health and Safety Networks. Retrieved April 9, 2004 from
Kittitas County Community Public Health and Safety Network. (1997). Comprehensive Plan.
(Available from Kittitas County Community Network, P.O. Box 881, Ellensburg,
Knapp, K. (1999). Community Networks: Progress Toward Measuring Results. Olympia, WA:
Washington State Institute for Public Policy.
Kreitlow, B. (1988, December 27). Rural women less likely to seek social services help:
University of Iowa study. Cedar Rapids Gazette, p. C6.
Leviton, L. (1996). Integrating psychology and public health: Challenges and opportunities.
American Psychologist, 51, 42-51.
Lorion, R., Iscoe, I., DeLeon, P., & VandenBos, G. (1996). Psychology and Public Policy:
Balancing Public Service and Professional Need. Washington, DC: American
Lyons, J.S. & Kisiel, C. (1996). Domestic violence and social policy: Integrating theory and
practice in the prevention of spousal and child abuse. In C. Stout (Ed.). The Integration of
Psychological Principles in Policy Development (pp. 45-60). Westport, CT:
Piattelli-Palmarini, M. (1994). Inevitable illusions: How Mistakes of Reason Rule our Minds.
New York: John Wiley & Sons.
Sattler, J. (1998). Clinical and Forensic Interviewing of Children and Families: Guidelines for
the Mental Health, Education, Pediatric, and Child Maltreatment Fields. San Diego, CA:
Schneiderman, N., Speers, M., Silva, J., Tomes, H., & Gentry, J. (Eds.). (2001). Integrating
Behavioral and Social Sciences with Public Health. Washington, DC: American
Street, L., Niederehe, G., & Lebowitz, B. (2000). Toward greater public health relevance for
psychotherapeutic intervention research: An NIMH workshop report. Clinical
Psychology: Science and Practice, 7, 127-137.
Wolfe, D. A. (1987). Child Abuse: Implications for Child Development and Psychopathology.
Newbury Park, CA: Sage Publications.
Youth Violence Reduction Act (1994) Wash. Laws, HB 2319, Spec. Sess.