The Experience and Expression of Anger in Maritally Violent and Maritally Discordant-Nonviolent Men

Krista A. Barbour, University of Southern California, Christopher I. Eckhardt, Southern Methodist University, Gerald C. Davidson, University of Southern California, and Howard Kassinove, Hofstra University

We examined the experience and expression of anger in community samples of 31 maritally violent (MV), 23 maritally dissatisfied-nonviolent (DNV), and 34 maritally satisfied-nonviolent (SNV) men. Two methods were used to assess anger. First, participants completed the State-Trait Anger Expression Inventory (STAXI), which assesses respondents' level of anger experience (State Anger, Trait Anger) and usage of various anger expression styles (Anger In, Anger Out, and Anger Control). Second, participants verbally reported their emotions during a laboratory anger-induction paradigm (Articulated Thoughts in Simulated Situations; ATSS). Results indicated that, relative to DNV and SNV men, MV men scored significantly higher on the STAXI Trait Anger and Anger Out scales, and lower on STAXI Anger Control.



The Role of Offset Control in Anxious Responding: An Experimental Test Using Repeated Administrations of 20% Carbon Dioxide-Enriched Air

Michael J. Zvolensky, C.W. Lejuez, and Georg H. Eifert, West Virginia University

The present investigation attempted to clarify whether a lack of control affects self-reported anxiety and physiological reactivity during eight administrations of 20% carbon dioxide (CO2 )-enriched air. Thirty individuals who reported high levels of suffocation fear were randomly assigned to a condition that either permitted or did not permit control over the offset of CO2 gas inhalation. In contrast to participants with control, participants without control reported significantly more self-reported anxiety and intense panic experiences. Although 20% CO2-enriched air reliably evoked physiological arousal for both groups, no significant between-group differences for peripheral indices of somatic reactivity were observed. We discuss the implications of these findings for understanding how control over aversive environmental stimuli mediates anxious responding in panic disorder.



Evaluating the Function of Hair Pulling: A Preliminary Investigation

Raymond G. Miltenberger, Ethan S. Long, John T. Rapp, Vicki Lumley, and Amy J. Elliott, North Dakota State University

Although little is known about the operant function of hair pulling associated with trichotillomania, attempts to further our knowledge in this area have not employed experimental methods. In the present study we assessed the hair pulling behavior of two individuals using a functional analysis methodology. The results showed that hair pulling was most probable in the alone condition. Based on the functional analysis results and information from behavioral interviews, we hypothesized that hair pulling was maintained by self-stimulation in these two subjects. We discuss limitations of the current investigation and the need for future research.



A Comparison of Ecletic Treatment With Webster-Stratton's Parents and Children Series In a Children's Mental Health Center: A Randomized Controlled Trial

Ted K. Taylor, Oregon Research Institute, Fred Schmidt, Lakehead Regional Family Centre, Debra Pepler,York University, and Christine Hodgins, Lakehead Regional Family Centre


Few studies have evaluated the effectiveness of empirically supported therapy in applied settings, or typical service in such settings. In this study, parents seeking help at a children's mental health center for managing their 3- to 8-year-old children's behaviors were randomly assigned to one of three conditions: Webster-Stratton's Parents and Children Series (PACS) parenting groups (46 families), the eclectic approach to treatment typically offered at the center (46 families), or a wait-list control group (18 families). After 15 weeks, mothers in both treatments reported fewer child behavior problems than mother on the wait list. Mothers in the PACS program reported behavior problems and greater satisfaction with treatment than mothers in the electic treatment. These findings support the effectiveness of the PACS program, relative to typical service, for parents seeking help managing their children's behavior.



Cognitive Behavioral Interventions for Secually Abused Children Exhibiting PTSD Symptomatology

Stephanie P. Farrell and Anthony A. Hains, University of Wisconsin-Milwaukee, and W. Hobart Davies, Children's Hospital of Wisconsin

This investigation examined the effectiveness of a cognitive behavioral intervention with 4 sexually abused children exhibiting posttraumatic stress disorder (PTSD), using a multiple baseline design. Participants ranged in age from 8 to 10 years. Intervention effectiveness was measured with pre-, post-, and 3-month follow-up assessments of PTSD symptomatology and affective (depression and anxiety) measures. In addition, these instruments were administered during baseline and prior to each treatment session. The intervention focused on the training of relaxation skills, positive self-talk, and cognitive restructuring. All 4 participants reported decreases in their PTSD symptomatology. The 3 participants who had elevated levels of depression and anxiety during baseline showed decreases with treatment. The results of this study suggest that cognitive behavioral techniques can be an effective approach for working with this population.



Predicting Onset and Cessation of Bulimic Behaviors During Adolescence: A Longitudinal Grouping Analysis

Eric Stice and W. Stewart Agras, Stanford University

Because research has not examined the predictors of the onset and cessation of specific bulimic behaviors, the present study tested whether a series of putative risk factors predicted the onset and cessation of binge eating and compensatory behaviors over a 9-month period in a community sample of female adolescents (N = 218). Greater perceived social pressure to be thin, internalization of the thin ideal, body dissatisfaction, dieting, and negative affect predicted the onset of binge eating and compensatory behaviors. Further, lower thin-ideal internalization, body dissatisfaction, and dieting predicted the cessation of these behaviors. Multivariate analyses suggested that this set of risk factors predicted the onset of binge eating and compensatory behaviors, but showed weaker relations to the cessation of these behaviors. Results support the assertion that these variables constitute risk factors for the development of bulimic behaviors, but suggest that somewhat different processes may serve to maintain bulimic symptoms.



Prospective Relation of Dieting Behaviors to Weight change in a Community Sample of Adolescents

Eric Stice, Stanford University

Although treatment studies suggest that weight loss programs do not have lasting effects, little research has examined the relation of dieting to weight change in prospective community studies, and none have done so with adolescents. This study examined the effects of dieting on body mass over a 9-month period within a community sample of adolescents (N = 396). Continuous measures of dieting behaviors predicted a relative increase in body mass (for males and females), whereas self-labeled dieting predicted decreased body mass (for females). Divergent findings resulted because self-labeled dieting reflected more extreme dietary efforts. Analyses also suggested that dieting was quadratically related to body mass, wherein low to moderate dieting predicted weight gain and extreme dieting predicted weight loss. Results highlight the importance of measurement issues and suggest that dieting only leads to weight loss for those adolescents employing extreme dietary behaviors.



Concordance and Synchrony in Measures of Anxiety and Panic Reconsidered: A Hierarchical Model of Anxiety and Panic

Richard E. Zinbarg, University of Oregon

Based on the observation of low to moderate correlations among multiple measures of anxiety or panic, Lang (1968, 1978, 1979) and Rachman (1978a, 1978b, 1990) developed the three-systems model that rejects the notion of anxiety or panic as unitary latent constructs. This model is critically reviewed with an emphasis on discussing its implications for intervention, methodology, and theory. An alternative conceptualization, the hierarchical model of anxiety and panic, is offered in which anxiety and panic are seen as unitary latent constructs at a higher level of a hierarchical structure that coexist with several separable response systems at a lower level of the hierarchy. The hierarchical model retains many of the strengths of the three-systems model while overcoming many of the limitations of the three-systems model.



Hierarchical Views of Anxiety: What Do They Profit Us?

John D. Cone, United States International University

Zinbarg (1998) revisits the widely accepted triple-response mode conceptualization for assessing and treating anxiety and fear. the scientific validity of these constructs has been undermined considerably by empirical findings of Lang (1978) and others that they do not exist all of a piece, and are made up of distinguishable responses broadly categorizable as cognitive, motor, and physiological. Zinbarg argues that these constructs can still be useful if viewed as unitary latent traits at a higher level, coexisting with several distinguishable response systems at lower levels of a hierarchical model. Two major difficulties with his argument are discussed: (a) methodological and (b) conceptual. The first deals with the highly complex multicontent-multimethod-multibehavior analyses needed to validate triple response mode logic. The hierarchical model appears to require the addition of multisystem comparison, further complicating the analysis. Given that there are as yet no studies showing the separate and joint contributions of method and behavior to response system relationships, it is questionable whether even more difficult analysis will be undertaken. Conceptually, the hierarchical model looks like a return to abstract, highly inferential, medical model logic, which behavioral assessors and therapists have historically found distasteful. More parsimonious behavioral approaches are available for dealing with "anxiety" and "fear." Practitioners and researchers are encouraged to consider them before adopting a model that takes us even further from the actual behavior of clients.



Considerations of Methodological Complexity and Taste Versus Explanatory Power in the Evaluation of Competing Scientific Models: A Reply to Cone (1998)

Richard E. Zinbarg, University of Oregon

Cone (1998) charges that the hierarchical model (a) requires more complex methodology than the three-systems model, and (b) is a return to abstract, highly inferential, medical- model logic, which behavior therapists have historically found distasteful. The validity of both of cone's criticisms are found to be wanting: (a) the three-systems model requires the same multitrait-multisystem-multibehavior-multimethod analysis as does the hierarchical model; it is necessary to measure two or more responses in each system using two or more methods regardless of whether the systems are hypothesized to be fully independent (as in the three-systems model) or only partially independent (as in the hierarchical model); and (b) the hierarchical model rejects the two hallmark assumptions of the medical. It is also argued that even if Cone's criticisms are valid, considerations of explanatory power should take precedence over considerations of complexity and taste. The evidence rather clearly indicates that the three-systems model should be rejected in favor of models that include unitary anxiety and panic constructs.



Does Cognitive-Behavioral Therapy Cure Obsessive-Compulsive Disorder? A Meta-Analytic Evaluation of Clinical Significance

Jonathan S. Abramowitz, University of Memphis

Meta-analytic methods were employed to investigate the clinical significance of exposure therapy in the treatment of obsessive-compulsive disorder (OCD). Samples of patients treated in 16 outcome trials were compared with 9 normative samples to clarify how similar the symptoms of treated patients were to those without history of OCD. The Maudsley Obsessional-Compulsive Inventory (MOCI) was used as the basis for comparison. As expected, substantial improvements during the course of treatment were observed. Additionally, at posttest, the average patient was functioning at a level more similar to the general population than to individuals with untreated OCD. Follow-up scores indicated that clinically significant changes were stable for up to 5 months. Treated patients, however, did remain more symptomatic than members of the general population. The results are discussed in terms of the advantages of routinely including tests of clinical significance in outcome research.