Aggression and Violent Behavior 12 (2007) 280 – 299
Addressing “The criterion problem” in the assessment of aggressive behavior: Development of a new taxonomic system Dominic J. Parrott a,⁎, Peter R. Giancola b a b
Department of Psychology, Georgia State University, P.O. Box 5010 Atlanta, GA 30302-5010, United States Department of Psychology, University of Kentucky, 115 Kastle Hall, Lexington, KY 40506-0044, United States Received 16 September 2005; received in revised form 30 January 2006; accepted 11 August 2006 Available online 25 September 2006
Abstract Aggressive behavior is a highly complex construct that is very challenging to measure. While advancements in the assessment of aggression have been made, some fundamental problems persist. First, the operational definition of aggressive behavior and its various subtypes are frequently misinterpreted and lack sufficient conceptual clarity. Second, due to these definitional problems, assessment instruments frequently correspond to different conceptualizations of aggression. In the present review, we attempt to resolve these limitations by proposing a new taxonomic system of aggressive acts that (a) corresponds to a hybrid definition of aggressive behavior, and (b) increases conceptual clarity between subtypes of aggressive behavior. It is argued that this classification system will permit greater precision in the assessment of aggression and lead to the improvement of theories, diagnostic systems, and clinical interventions. © 2006 Elsevier Ltd. All rights reserved. Keywords: Aggressive behavior; Classification; Taxonomy; Assessment
Contents 1.
2.
Defining aggressive behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1. Determining a hybrid definition of aggressive behavior . . . . . . . . . . 1.1.1. Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1.2. Definitions of aggression . . . . . . . . . . . . . . . . . . . . . 1.1.3. Conceptual differences between aggression, anger, and hostility . 1.1.4. An operational definition of aggression. . . . . . . . . . . . . . 1.2. Defining subtypes of aggression . . . . . . . . . . . . . . . . . . . . . . 1.2.1. Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2.2. Limitations of current categorical efforts . . . . . . . . . . . . . 1.2.3. Theoretical justification for a new classification system . . . . . A new taxonomic system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1. Selection criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1.1. Adherence to an operational definition of aggression. . . . . . .
⁎ Corresponding author. Tel.: +1 404 651 1165; fax: +1 404 651 1391. E-mail address:
[email protected] (D.J. Parrott). 1359-1789/$ - see front matter © 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.avb.2006.08.002
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
282 282 282 282 283 283 283 283 284 284 285 286 286
D.J. Parrott, P.R. Giancola / Aggression and Violent Behavior 12 (2007) 280–299
2.1.2. Conceptual clarity/parsimony 2.1.3. Empirical support . . . . . . 2.2. Modes of expression . . . . . . . . . 2.2.1. Direct versus indirect . . . . 2.2.2. Active versus passive . . . . 2.3. Subtypes of aggression: active . . . . 2.3.1. Physical and verbal . . . . . 2.3.2. Postural . . . . . . . . . . . 2.3.3. Damage to property . . . . . 2.3.4. Theft . . . . . . . . . . . . . 2.4. Subtypes of aggression: passive. . . . 2.4.1. Physical and verbal . . . . . 2.4.2. Postural . . . . . . . . . . . 2.4.3. Damage to property . . . . . 2.4.4. Theft . . . . . . . . . . . . . 3. Assessment of aggression . . . . . . . . . . 3.1. Self-report inventories. . . . . . . . . 3.2. Observational techniques . . . . . . . 3.3. Clinical interviews . . . . . . . . . . 3.4. Laboratory paradigms . . . . . . . . . 3.4.1. Physical aggression . . . . . 3.4.2. Theft . . . . . . . . . . . . . 3.4.3. Verbal aggression . . . . . . 3.4.4. Summary. . . . . . . . . . . 4. Implications for diagnosis and treatment . . . 5. Summary . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
281
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . .
286 286 287 287 287 288 288 288 288 289 289 289 289 289 290 290 290 292 293 294 294 294 295 295 295 296 297
Aggressive behavior is a multifaceted construct that may be expressed behaviorally in a myriad of ways. Murder and verbally assaulting one's spouse are well-accepted forms of aggression. However, acts such as arson, “keying” a car, spreading viscous rumors, theft, or intentionally not waking up a roommate who has an examination in the morning may all be construed as aggressive. This reality creates a challenging task for the development of assessment instruments that aim to precisely measure this wide range of behaviors. As a result of these challenges, current research on aggressive behavior continues to suffer from limitations in measurement, also referred to as the criterion problem. We contend that the criterion problem exists on two primary levels. First, operational definitions of aggression and its various manifestations are frequently misinterpreted. For example, researchers have inconsistently differentiated between aggressive behavior and the related emotional and attitudinal/ cognitive constructs of anger and hostility, respectively. Indeed, it appears that these terms are oftentimes used interchangeably in the literature. Similarly, most investigators do not appropriately differentiate between the various routes by which aggression may be expressed (e.g., direct versus indirect). Thus, in order to address the current criterion problem, it is necessary for researchers to agree upon operational definitions of aggressive behavior and its various subtypes. Second, many instruments designed to assess aggressive acts do not adequately correspond to existing operational definitions. For example, a self-report measure of “aggression” may actually assess one's display of aggressive behavior and experience of anger. Similarly, an observational rating scale that purports to assess “verbal aggression” may actually measure theoretically distinct subtypes of verbal aggression (e.g., direct versus indirect). This lack of precision in measurement creates significant challenges for researchers attempting to identify risk factors that may differentially predict these subtypes of verbally aggressive behavior. Indeed, only via precise measurement can risk factors for direct verbal aggression (e.g., cursing at someone) be differentiated from determinants of indirect verbal aggression (e.g., spreading harmful rumors about someone). Consistent with this view, empirical evidence indicates that some risk factors differentially predict distinct manifestations of aggression. For example, recent research suggests that individuals who are more comfortable in
282
D.J. Parrott, P.R. Giancola / Aggression and Violent Behavior 12 (2007) 280–299
social situations tend to employ direct (i.e., face to face) aggressive tactics, whereas those who are more socially inhibited tend to use indirect aggressive strategies (Richardson & Green, 2003). A study by Kaukiainen et al. (1999) found that higher levels of social intelligence were positively associated with the expression of indirect, but not direct, aggression, whereas empathy was negatively correlated with the direct expression of physical and verbal aggression. Therefore, to the extent that investigators are not precise about their operational definition of aggression as well as their choice of pertinent assessment instruments, scientific understanding of risk factors for aggression and, ultimately, theory development will suffer. Given these arguments, the overarching aim of the present review is to facilitate the resolution of the criterion problem. Specifically, we will review the manner in which operational definitions of aggression are frequently misinterpreted as well as the lack of precision involved in the categorization of subtypes of aggression. Next, we will illustrate the disconnect between these operational definitions of aggressive behavior and instruments designed to assess it by reviewing limitations in the construct validity of these measures. While an exhaustive review of assessment instruments is beyond the scope of this article, we will attempt to review measures of aggression that most clearly highlight these limitations. After consideration of pertinent literature, we contend that the resolution of the criterion problem requires the endorsement of a hybrid operational definition of aggressive behavior and the creation of a new classification system. The proposed classification system represents a significant modification of a taxonomy originally outlined by Buss (1961). It first categorizes aggressive acts by their route of expression via the combination of two dichotomous variables: (a) direct–indirect, and (b) active–passive. Within each of the four possible routes of expression (e.g., active and direct), various aggressive subtypes are then outlined (e.g., physical, verbal, theft). Advantages of this system will be reviewed with a particular emphasis on its ability to facilitate the identification of risk factors for various subtypes of aggressive behavior and to spur new and more theoretically focused research. Suggestions will be offered regarding the use of current assessment instruments in order to more appropriately correspond with these definitions. Finally, implications for diagnosis and treatment will be discussed. 1. Defining aggressive behavior 1.1. Determining a hybrid definition of aggressive behavior 1.1.1. Overview There has been a great deal of debate in the literature on how to define aggressive behavior. As a result, many different definitions abound. In fact, Harre and Lamb (1983) reported that over 200 different definitions of aggression have been advanced. For example, Dollard, Doob, Miller, Mowrer, and Sears (1939) posited that aggression involves “an act whose goal-response is injury to an organism” (pg. 11), whereas Zillmann (1979) restricted aggressive acts to those that attempt to produce bodily injury to the victim. Indeed, the existence of so many different conceptualizations of aggression reflects the struggle by researchers to define this construct (Barratt & Slaughter, 1998; Coie & Dodge, 1998; Geen, 2001). The inability to form a consensus regarding an operational definition for aggressive behavior has led to a lack of precision in its measurement. That is, given the existence of so many working definitions, there is great variability between assessment instruments due to the fact that they are based upon different theoretical conceptualizations of aggression. Therefore, we contend that the resolution of the criterion problem must involve greater agreement in the literature regarding an operational definition of aggressive behavior. Of equal importance, researchers must refine and develop assessment techniques that are consistent with a consensus definition. 1.1.2. Definitions of aggression Historically, one of the more well-accepted definitions was put forth by Buss (1961), who defined aggression as “a response that delivers noxious stimuli to another organism” (pg. 1). Although Buss notes that the accidental delivery of harm should not be considered as aggressive, a literal interpretation of this definition includes instances in which an individual causes harm to another person without intending to do so (e.g., stepping on someone's foot by mistake). In addition, this definition encompasses acts in which an individual may welcome harm (e.g., sadomasochism, suicidal behavior) or in which the perpetrator may cause harm despite having prosocial intentions (e.g., a dentist causing pain in his patient to remove a tooth).
D.J. Parrott, P.R. Giancola / Aggression and Violent Behavior 12 (2007) 280–299
283
In response to these limitations, Baron (1977) and Baron and Richardson (1994) extended this definition by specifying the perpetrator's intentions and the victim's wish to avoid harm, stating that aggression is “any form of behavior directed toward the goal of harming or injuring another living being who is motivated to avoid such treatment” (Baron & Richardson, 1994; pg. 7). This definition has numerous strengths in that it includes many different forms of aggression while also excluding harmful acts that are not intended to be aggressive, harmful acts that the victim is not motivated to avoid, and prosocial acts that happen to cause harm to the victim. Similar definitions have also been advanced by other leading theorists (Anderson & Bushman, 2002; Berkowitz, 1993; Geen, 2001). 1.1.3. Conceptual differences between aggression, anger, and hostility Unfortunately, contemporary research on aggressive behavior has not fully adopted these recent definitions. This reality represents, in part, the first fundamental problem in the assessment of aggression. For example, the related theoretical constructs of anger and hostility continue to be used in lieu of aggression (Eckhardt, Norlander, & Deffenbacher, 2004; Spielberger, Jacobs, Russell, & Crane, 1983). Therefore, it is necessary to elucidate the conceptual distinctions between these constructs by reviewing their operational definitions. As noted above, aggression refers to a behavioral process that includes the goal of inflicting harm to another living being (i.e., not to an inanimate object) who is motivated to avoid the act. In contrast, anger is conceptualized as an emotional state that can vary in intensity, from mild annoyance to rage (Spielberger et al., 1983). Moreover, the experience of anger lacks a specific goal (Berkowitz, 1993) and is not necessary for aggression to occur. Unlike aggression or anger, hostility is an attitudinal/cognitive construct comprised of enduring cognitions that involve negative interpretations of the environment. As Buss (1961) writes, hostility “involves the interpretation and evaluation of stimuli, and the negative evaluations have no impact on others unless they are verbalized” (pg. 12). Indeed, once hostile attitudes are verbally or physically expressed, they may be more appropriately labeled as aggression. 1.1.4. An operational definition of aggression Despite the theoretical distinctions between aggressive behavior, anger, and hostility, the assessment of aggressive behavior continues to lack sufficient conceptual specificity. Therefore, we believe that assessment strategies would be significantly enhanced by adhering to the following theoretical guidelines reviewed thus far: 1) recognizing aggression as a behavioral process and, therefore, distinguishing it from related constructs such as anger or hostility; and 2) adopting a hybrid definition of aggression, consistent with that of contemporary theorists (e.g., Anderson & Bushman, 2002; Baron & Richardson, 1994; Berkowitz, 1993; Geen, 2001), that highlights the goal of inflicting harm to another living being who wishes to avoid such treatment. By reaching a consensus regarding this hybrid operational definition, the precision of research designs, assessment instruments, theoretical development, diagnosis, and clinical interventions will be significantly enhanced. 1.2. Defining subtypes of aggression 1.2.1. Overview A second issue pertinent to the definition of aggressive behavior involves its categorization into various subtypes. For over 40 years, theorists have proposed numerous subtypes of aggression in recognition of that the fact that there are various routes by which it may be expressed. These subtypes include, but are not limited to, direct versus indirect (Buss, 1961; Feshbach, 1969), physical versus verbal (Buss, 1961), active versus passive (Buss, 1961), rational versus manipulative (Bjorkqvist, Osterman, & Kaukiainen, 1992), proactive versus reactive (Dodge, 1991), antisocial versus prosocial (Sears, 1961), annoyance-motivated versus incentive-motivated (Zillmann, 1979), overt versus covert (Buss, 1995), targeted versus targetless (Buss, 1961), overt versus relational (Crick, 1996; Crick & Grotpeter, 1995), and relational versus social (Bjorkqvist, 2001; Underwood, Galen, and Paquette, 2001). In any discussion of aggression subtypes, it is worth noting that the literature has long differentiated between hostile and instrumental aggression (e.g., Bandura, 1973; Feshbach, 1964; Geen, 2001). This dichotomous conceptualization of aggressive behavior emerged, in part, from the observation that aggressive acts may differ in terms of their primary purpose, the presence or absence of anger, and the degree of planning involved. Hostile aggression (sometimes termed affective, impulsive, or retaliatory aggression) is an angry, unplanned act that intends to harm another person. In contrast, instrumental aggression is a premeditated, “cold blooded” act where the primary goal is to obtain some
284
D.J. Parrott, P.R. Giancola / Aggression and Violent Behavior 12 (2007) 280–299
nonaggressive incentive rather than to inflict harm. This view has been widely accepted by researchers over the past 40 years and facilitated the development of many important models of aggressive behavior. However, it has recently been argued that this distinction is no longer a useful descriptor of different types of human aggression (for a review, see Bushman & Anderson, 2001). In short, Bushman and Anderson (2001) posit that the division between hostile and instrumental aggressive behavior is confounded by the presence of both controlled and automatic information processing in many aggressive acts. That is, hostile aggression may involve some degree of planning and instrumental aggression may involve automatized, unscripted behaviors. Further, many aggressive acts are motivated by multiple goals and involve some degree of anger (i.e., hostile aggression) and planning (i.e., instrumental aggression). Notably, Barratt and Slaughter (1998) estimated that only 50% of aggressive acts could be categorized as either hostile or instrumental, leaving approximately 50% of acts to be categorized as “mixed”, “can't determine”, or “medically related”. This report underscores the lack of specificity provided by the hostile–instrumental distinction and supports the contention that this dichotomous conceptualization does not adequately distinguish between different forms of aggressive behavior. 1.2.2. Limitations of current categorical efforts Although the effort to categorize subtypes of aggression is of theoretical and practical value, the shear volume of proposed subtypes has become cumbersome (Underwood et al., 2001). A similar state of affairs exists in the classification of animal aggression, where numerous subtypes exist and refinement of these categories is needed (McEllistrem, 2004; Weinshenker & Siegel, 2002). Most importantly, many aggressive acts may be reasonably assigned to more than one category. For example, spreading rumors about a person to harm their social standing may be considered “relational aggression”, “social aggression”, or “indirect aggression” (Underwood, 2003; Underwood et al., 2001). As previously noted, many aggressive acts may also qualify as both “hostile” and “instrumental” (Bushman & Anderson, 2001). This limitation is at the foundation of the current criterion problem and directly relates to several other inadequacies in the current approach to classifying aggressive acts. First, because of the conceptual overlap that currently exists, the ability of investigators to determine whether risk factors differentially predict specific aggressive subtypes is compromised. By extension, the development of new theories as well as clinical interventions that target specific subtypes of aggressive behavior is also affected. Second, the construct validity of instruments designed to assess subtypes of aggression is sacrificed. Indeed, because many aggressive acts may fit into multiple categories, a measure of “social aggression” may include items that are also considered to reflect “indirect aggression” (Archer, 2001). Third, classification systems are advantageous in part because they facilitate communication between scientists. Due to the existing overabundance and conceptual overlap of aggressive subtypes, scientific communication is greatly impaired. Finally, without clear conceptual distinctions between subtypes of aggression, efforts to diagnose and treat violent individuals in clinical settings will continue to suffer as well. Indeed, existing diagnostic systems, such as the DSM-IV (American Psychiatric Association, 1994), possess strikingly little precision with regard to their conceptualization of aggression. For example, the only DSM-IV diagnosis devoted specifically to aggressive behavior is Intermittent Explosive Disorder, which is characterized by “discrete episodes of failure to resist aggressive impulses resulting in serious assaults or destruction of property” (American Psychiatric Association, 1994; pg. 609). In addition to blurring the distinction between aggressive and anger outbursts, this diagnosis lacks clearly defined subtypes of aggressive behavior. 1.2.3. Theoretical justification for a new classification system To address these limitations, two solutions are possible. Researchers could attempt to reduce conceptual overlap by continuing the challenging task of independently assigning every type of aggressive act to a discrete category. Assuming that this task is feasible, the result still would not resolve the limitations outlined above. In fact, an overwhelming number of aggression subtypes would emerge without any unifying classification system. Because the lack of conceptual clarity between different types of aggressive behavior would likely persist, theoretical and scale development would be stymied. Indeed, an unwieldy number of categories for aggressive acts would be of little value to researchers and would also exacerbate the existing problems in diagnosis and communication between researchers. These and other limitations of creating an overabundance of “personalized subtypes of aggression” (pg. 28) have been noted in the literature (Underwood, 2003). In contrast to the current approach, we contend that a more practical and effective solution involves the adoption of an inclusive, yet parsimonious, classification system that maintains conceptual clarity. The development of a new
D.J. Parrott, P.R. Giancola / Aggression and Violent Behavior 12 (2007) 280–299
285
taxonomic system is consistent with recent recommendations to discard the distinction between hostile and instrumental aggression (Bushman & Anderson, 2001). As noted by Bushman and Anderson (2001), such efforts are needed to facilitate the development of a “second generation paradigm of human aggression” (pg. 278). Other researchers share this contention (Ramirez & Andreu, 2003). That is, approaching the study of aggressive behavior through an empirically updated paradigm will stimulate more precise measurement and provide a foundation for the development of more advanced theories and treatment interventions. Indeed, to the extent that future research continues to focus upon the multitude of independently derived aggression subtypes, it is unlikely that these goals will be realized. The taxonomic system proposed herein addresses the limitations outlined above. First, a simplified system is more likely to eliminate the conceptual overlap between subtypes that presently exists. Indeed, the delineation of specific subtypes of aggression will help researchers determine different etiological pathways to each. Advancements in this area will also improve current theories, diagnostic strategies, and the development of different treatments designed to reduce or prevent these various types of aggression. Second, this approach will create a landscape in which scale development can be finely tuned to meet the needs of investigators who wish to assess clearly defined subtypes of aggressive behavior. Third, with broad consensus for a parsimonious classification system, researchers and clinicians alike will be able to communicate empirical, theoretical, and clinical progress more effectively. Finally, by adopting a single taxonomic system, separate diagnoses for maladaptive “anger episodes” (e.g., indiscriminately destroying property, yelling and screaming to nobody in particular) and discrete “aggressive episodes” may be envisioned. Indeed, given the distinctions between aggressive behavior and anger outlined above, an individual may presently receive the diagnosis of Intermittent Explosive Disorder for displaying outbursts of anger that result in non-harmful property destruction or for engaging in aggressive acts that cause harm to others. Therefore, the diagnosis of an “aggressive episode” would be characterized by any behavior motivated by the goal of harming another living being and could include the various subtypes proposed herein. While empirical and theoretical work has suggested the delineation of a class of “anger disorders” (Eckhardt & Deffenbacher, 1995), very little research has focused on the development of disorders of aggressive behavior. Indeed, conceptual refinement of the definition and subtypes of aggression will provide the foundation necessary to explore this possibility. 2. A new taxonomic system We contend that the classification system outlined below represents a step in helping to resolve the existing criterion problem. This simplified taxonomic system is based upon combinations of two dichotomous variables that reflect the various routes by which aggression is expressed: (a) direct versus indirect, and (b) active versus passive. Within each of the four possible routes of expression, various subtypes may be identified, including physical, verbal, postural, property damage, and theft (see Table 1). The theoretical distinctions of aggressive behavior inherent to this system are in accordance with the operational definition of aggression outlined above (e.g., Anderson & Bushman, 2002; Baron & Richardson, 1994; Berkowitz, 1993; Geen, 2001).
Table 1 Proposed taxonomic system of aggressive behaviors Direct expression
Indirect expression
Active expression
Subtypes Physical Verbal Postural Damage to property Theft
Subtypes Physical Verbal Postural Damage to property Theft
Passive expression
Subtypes Physical Verbal Damage to property Theft
Subtypes Physical Verbal Damage to property Theft
286
D.J. Parrott, P.R. Giancola / Aggression and Violent Behavior 12 (2007) 280–299
2.1. Selection criteria The proposed taxonomy was based on the following three criteria: Adherence to an operational definition of aggression, conceptual clarity/parsimony, and empirical support. These criteria were selected in order to facilitate reliable and valid assessment of observable aggressive acts. In contrast, criteria that were not consistent with this goal (e.g., criteria focusing on unconscious motives) were not utilized. 2.1.1. Adherence to an operational definition of aggression Given the operational definition outlined above (e.g., Anderson & Bushman, 2002; Baron & Richardson, 1994; Berkowitz, 1993; Geen, 2001), any classification system of aggression must consist entirely of behaviors that are intended to cause significant harm to a victim who wishes to avoid the act. Therefore, subtypes of aggression that include behaviors lacking the intent to cause harm were not considered. For example, behaviors categorized along the antisocial versus prosocial dimension (Sears, 1961) include prosocial acts that are motivated by the goal of helping, not harming, the target (e.g., a dentist inflicting pain in his patient by pulling a decayed tooth). Similarly, in order to adhere to this hybrid operational definition, a classification system must exclude harmful acts that are not avoided by the target (e.g., sadomasochism, suicidal behavior). 2.1.2. Conceptual clarity/parsimony A fundamental problem in the classification of aggressive behaviors is the lack of conceptual clarity and parsimony that presently exists. Therefore, any new classification system must make clear distinctions between subtypes while simultaneously including the majority of behaviors that may be labeled as aggressive. Below, we present several examples of current subtypes of aggression and discuss how they adapt to our system. Perhaps the most harmful, and widely researched, acts of aggression involve the infliction of bodily injury (i.e., physical aggression). Given the widespread attention to these types of behaviors, the distinction between acts of direct physical aggression (e.g., murder) and indirect physical aggression (e.g., hiring an assassin) has been noted in the literature (e.g., Bushman & Anderson, 2001) and fits within our taxonomy. Unfortunately, other conceptual divisions continue to lack sufficient clarity. For example, researchers have conceptualized property destruction as an act of indirect physical aggression (e.g., Buss, 1961; Geen, 2001) as well as a distinct subtype (e.g., Barratt, Stanford, Dowdy, Liebman, & Kent, 1999). To remedy this problem, a new taxonomic system must distinguish acts of property destruction that bring harm to the victim from those that cause actual bodily injury to the victim (i.e., acts of physical aggression). Further, direct and indirect expressions of property destruction may be considered aggressive only to the extent that the perpetrator intends to inflict harm on a living being. Another case in which subtypes lack sufficient conceptual clarity involves the distinctions between social, relational, and indirect aggression (Underwood, 2003; Underwood et al., 2001). Both social and relational aggression involve behaviors that attempt to harm the victim's social relationships, although the acts subsumed in these subtypes may be direct or indirect as well as verbal or non-verbal (Galen & Underwood, 1997). To address these problems, a taxonomy must clearly define direct versus indirect acts while simultaneously categorizing acts that are verbal or non-verbal (i.e., postural). In summary, the selection of aggression subtypes for a new classification system must be guided by clear conceptual boundaries that include a wide range of aggressive acts. As will be reviewed below, the proposed system appears to satisfy this criterion. 2.1.3. Empirical support The selection of aggression subtypes must be supported by empirical research. As previously noted, although the hostile–instrumental distinction has been useful in the past, pertinent literature indicates that it is no longer a useful descriptor of the various manifestations of aggression. As such, a distinction between these proposed types of aggressive behavior was not included in the proposed taxonomy. In contrast, research supports the inclusion of direct–indirect expressions of aggressive behavior. For example, studies indicate that different personal and situational risk factors are associated with the expression of direct and indirect aggression (Archer, 2004a; Kaukiainen et al., 1999; Richardson & Green, 2003; Walker, Richardson, & Green, 2000). Notably, the vast majority of investigations on human aggression have focused upon active, relative to passive, behaviors (see below for a detailed discussion of the active–passive distinction). Thus, studies investigating acts such
D.J. Parrott, P.R. Giancola / Aggression and Violent Behavior 12 (2007) 280–299
287
as fighting, cursing at someone, and negatively evaluating another person are all assessing active forms of aggressive behavior. On the other hand, relatively little research has been conducted on factors that relate to the expression of passive aggression. Nonetheless, acts of passive aggression are documented in the literature (e.g., Baron & Neuman, 1996), oftentimes as a precursor to active, physical aggression (Murphy & O'Leary, 1989). As such, we contend that despite a relative lack of empirical attention to passive relative to active aggression, the inclusion of this distinction is warranted in any taxonomic system. Indeed, including this dichotomy may spur new research into the understudied area of passive aggression. Empirical evidence also supports the selection of the specific subtypes of aggression proposed in our system. A multitude of studies have demonstrated the use of physical and verbal aggression (e.g., Archer, 2004b; Parrott & Zeichner, 2003; Schumacher & Leonard, 2005), and factor analytic studies support the delineation of these two subtypes (Buss & Perry, 1992). However, evidence also suggests that aggression may take forms that are non-verbal (Paquette & Underwood, 1999) or involve destruction (Fabiano et al., 2004; Rose, 2003) or theft of the victim's property (Ireland, 2004). Indeed, research on bullying indicates that aggressive acts may take the form of direct theft (Ireland, 2002) or damage to one's property (Glover, Gough, Johnson, & Cartwright, 2000). Some of these subtypes have also been established among young children. For example, confirmatory factor analysis of a longitudinal sample of 4–8 year old children indicated that indirect verbal aggression and direct physical aggression were separate constructs (Vaillancourt, Brendgen, Boivin, & Tremblay, 2003). In conclusion, research supports the contention that aggression may be expressed in different ways (e.g., direct versus indirect) and may take various forms (e.g., physical, postural, damage to property). As will be reviewed below, the development of our classification system was guided by these empirical findings. 2.2. Modes of expression 2.2.1. Direct versus indirect The fundamental distinction between direct and indirect aggression involves the route in which aggression is delivered, and, as a result, the victim's ability to identify the perpetrator. Direct means of aggression involve face-toface interactions in which the perpetrator is easily identifiable by the victim. In contrast, indirect aggression is delivered more circuitously, and the perpetrator is able to remain unidentified and thereby avoid accusation, direct confrontation, and/or counterattack from the target (Buss, 1961; Richardson & Green, 1997; 1999). Pertinent literature indicates that direct and indirect aggression are, in fact, separate constructs (Richardson & Green, 2003). Therefore, the direct– indirect distinction appears to be quite appropriate, given that it provides clearly defined conceptual boundaries for different expressions of aggressive acts and is supported by empirical research. It is worth noting that researchers have proposed further distinctions of indirect aggression, including social and relational aggression (Bjorkqvist, 2001; Crick & Grotpeter, 1995; Underwood, 2003; Underwood et al., 2001). These subtypes are characterized by the perpetrator's manipulation of social relationships to harm the victim. Social aggression, in particular, possesses the strength of including non-verbal acts (e.g., a glaring stare) that may cause harm to one's victim. However, as Archer (2001) writes, these subtypes “blur the direct–indirect distinction, since some [aggressive acts] may be used face-to-face” (pg. 268). Given these data, use of these subtypes to specify acts of indirect aggression appear unwarranted.1 2.2.2. Active versus passive The difference between active and passive forms of aggression lies in the extent to which the aggressor is engaged in behavior aimed at harming the victim. That is, active aggression involves the perpetrator's active engagement in harming the victim, whereas passive aggression is characterized by a lack of active responding that causes harm. According to Buss (1961), the majority of aggressive acts, whether direct or indirect, may be labeled as active. Examples of such acts include assault with a weapon (direct) or spreading harmful rumors (indirect). Acts of passive aggression are typically direct and detectable by the target. For example, an employer may intentionally not write a letter of recommendation for an employee, thereby resulting in the employee's incomplete application for a new job. Indirect forms of passive aggression, though rare, are possible and involve a complex chain of events. For example, a 1
The inclusion of non-verbal acts in the subtype of social aggression is noteworthy. We contend that these non-verbal displays (also termed “postural aggression”) that inflict harm to others deserve inclusion in a comprehensive taxonomic system. As such, these acts are reviewed below.
288
D.J. Parrott, P.R. Giancola / Aggression and Violent Behavior 12 (2007) 280–299
perpetrator may inflict harm on another person by refusing to move during a sit-in demonstration or to speak in someone's defense when the victim is being criticized. These methods of “inaction” do not directly cause harm to the victim. Rather, an aversive stimulus is delivered to the target through a series of events (e.g., a third party verbally abusing the target). It is worth noting that the perpetrator must intend to inflict harm upon the victim. Instances of forgetfulness that harm the victim (e.g., an employer forgetting that a letter had to be written) should not be considered aggressive. 2.3. Subtypes of aggression: active As reviewed thus far, the first level of our classification system identifies the route by which aggression may be expressed. Four possibilities exist, including acts that are active and direct, active and indirect, passive and direct, and passive and indirect. However, greater specificity is required to classify the multitude of aggressive acts that may adhere to a particular route of expression. As such, we propose that aggressive acts may be classified further as either physical, verbal, postural, damage to property, or theft. All subtypes may be expressed directly or indirectly. For clarity, these subtypes are first outlined with regard to active expressions of aggressive behavior. We then discuss their application to passive expressions of aggression. 2.3.1. Physical and verbal The vast majority of aggression research has involved the assessment of behaviors that are either physical or verbal in nature. This makes the distinction between these types of behaviors especially relevant. Active and direct expressions of aggression may be physical, in which the noxious stimuli delivered to the victim result in physical pain or injury (e.g., hitting, punching), or verbal, in which noxious stimuli are communicated orally and inflict psychological harm to the victim (e.g., criticism, cursing). Similarly, active and indirect expressions of aggression may also be physical (e.g., hiring an assassin) or verbal (e.g., gossiping, spreading rumors). 2.3.2. Postural Aggression may also be codified as postural, and the importance of this class of behaviors has been noted in the literature (Olweus, 1996). Postural aggression includes non-verbal acts that do not require physical contact with the victim. Acts may be direct (e.g. making threatening faces, rolling eyes, invading personal space during an interpersonal interaction) or indirect (e.g., non-verbal behaviors that result in intentionally excluding the victim from a group). The inclusion of postural aggression as a subtype in the proposed taxonomic system rectifies existing conceptual problems in the classification of aggressive behavior. As previously noted, the constructs of indirect, social, and relational aggression possess significant overlap because they include circuitous behaviors that may harm the social standing of the victim. However, proponents of the term “social aggression” note that only this subtype includes the use of non-verbal behaviors (Underwood et al., 2001; Underwood, 2003). Therefore, by including acts of direct and indirect postural aggression (i.e., non-verbal behaviors) in the proposed classification system and excluding terms such as social or relational aggression, conceptual clarity may be enhanced. 2.3.3. Damage to property A number of researchers have argued that acts such as slashing tires, arson, and other forms of property destruction constitute a form of aggression that is indirect and physical (e.g., Baron & Richardson, 1994; Buss, 1961). Without question, these acts are considered aggressive to the extent that they involve the delivery of harm to another individual. However, this type of classification does not account for a situation in which a perpetrator destroys a target's property in a face-to-face confrontation. For example, many conflicts between married couples involve one partner witnessing the destruction of his or her property by a spouse. Therefore, we contend that these acts are best classified as damage to property. In making this distinction, such acts may be expressed in direct or indirect ways. Specifically, a wife may tear the pages out of her husband's favorite book during an argument (i.e., active, direct property destruction), or an employee may secretly slash the car tires of his supervisor (i.e., active, indirect property destruction). Notably, instances of nonaggressive destruction may be misinterpreted as damage to property (Buss, 1961). For example, consider a person who, after being provoked during a phone conversation, hangs up and subsequently breaks the phone by throwing it against a wall. Although property destruction has occurred, the provocateur has no knowledge of the act. Moreover, the perpetrator is not reinforced by any harm brought to the provocateur. Therefore, this act of property destruction would not be considered aggressive.
D.J. Parrott, P.R. Giancola / Aggression and Violent Behavior 12 (2007) 280–299
289
2.3.4. Theft The final proposed subtype of aggression involves acts of theft. Such acts are only labeled aggressive to the extent that they serve the purpose of delivering harm to a target. While theft is frequently expressed indirectly (e.g., covertly stealing a person's car keys so that they can't drive to work), it may also be expressed in a direct manner. An example of an active and direct form of theft might involve a husband who, during a fight one morning with his wife, takes her car keys from the kitchen table and leaves home. In doing so, the husband's action is active and direct (i.e., his wife clearly knows that he took her keys) and was delivered with the intent to harm his wife (i.e., preventing her from driving to work). 2.4. Subtypes of aggression: passive The subtypes of aggression reviewed thus far may all be expressed actively as well as directly or indirectly. Moreover, research suggests that different risk factors may predict distinct behavioral expressions of these subtypes (e.g., Archer, 2004a; Walker et al., 2000). Acts of passive aggression, however, represent a unique class of behaviors. As noted above, while passive aggression may be expressed directly or indirectly, it is characterized by a lack of active responding. Therefore, it may be argued that it is not possible to subdivide “inaction” into physical, verbal, postural, damage to property, or theft subtypes. Indeed, any act consistent with these subtypes requires active engagement by the perpetrator. Further, with notable exceptions (Baron & Neuman, 1996), research has not consistently demonstrated that risk factors differentially predict various subtypes of passive aggressive behavior. Thus, in terms of creating a parsimonious classification system, there appears to be justification for not applying any subtypes to passive expressions of aggression. Despite these arguments, creating a typology of passive aggressive acts has several advantages. First, researchers will have a framework to determine whether different risk factors predict particular subtypes of passive aggression. Such efforts would facilitate theoretical development. Second, by creating a working typology of passive aggressive acts, diagnostic criteria for relevant psychopathology (e.g., passive–aggressive personality disorder) may be improved. Progress in this area would clearly enhance the development of new treatments and prevention techniques for individuals who frequently use these tactics. Third, and perhaps most importantly, the application of our subtypes to passive expressions of aggression will spur more research in this area to determine if they are indeed valid. Thus, while it may be argued that the inclusion of subtypes that relate to “inaction” (i.e., passive aggression) is awkward, we believe that the inclusion of these subtypes will advance further our scientific understanding of aggressive behavior. Taken together, these advantages are consistent with any classification system's overarching goal of facilitating new and improved research. 2.4.1. Physical and verbal Physical and verbal types of passive aggression are characterized by the absence of a physical or verbal behavior that could have prevented harm to the victim. The following examples of each were outlined by Buss (1961). For example, by participating in a sit-in demonstration, a passive assailant can directly cause harm by physically prohibiting the victim from obtaining some goal. Similarly, by refusing to move during a demonstration, the aggressor may indirectly bring harm to the victim. An example of passive, direct, verbal aggression involves refusing to talk to another person (e.g., giving “the silent treatment”). Likewise, an example of passive, indirect, verbal aggression involves one's intentional failure to make verbal comments that would prevent harm to another person (e.g., not speaking in one's defense). 2.4.2. Postural By definition, acts of postural aggression involve an active form of non-verbal behavior. Such acts may be perpetrated directly to the victim (e.g., a glaring stare) or indirectly to others that result in the victim's exclusion from a group (e.g., glaring at a group member in disapproval of the victim's inclusion). Therefore, although the inclusion of this subtype among expressions of direct and indirect forms of active aggression is justified, postural aggression is not applicable to expressions of passive aggression. 2.4.3. Damage to property Passive expressions of property damage may be perpetrated alone (i.e., directly) or involve a third party (i.e., indirectly). For example, consider a wife who tells her husband to water her favorite plant while she is away on a
290
D.J. Parrott, P.R. Giancola / Aggression and Violent Behavior 12 (2007) 280–299
business trip. In order to harm his wife, the husband intentionally does not water the plant and, as a result, it dies. In this case, the husband's aggression was expressed directly (i.e., his wife will know that he did not water the plant) and passively (i.e., his aggression was characterized by a lack of behavior). Imagine a similar scenario in which this couple's son is also aware of his mother's request to water the plant. If the son is motivated to harm his mother, he may “turn a blind eye” to his father's failure to water the plant. The son's “inaction” is indirect (i.e., his mother cannot prove that he knew of his father's behavior) and has the goal of inflicting harm to his mother. 2.4.4. Theft Passive expressions of theft, whether direct or indirect, are similar to examples of damage to property. However, instead of one's intentional failure to prevent damage to the victim's property, the passive assailant fails to stop others from stealing the victim's property. For example, intentionally “overlooking” some form of theft that would harm a victim may be construed as a passive form of theft. The failure to prevent theft would be considered direct if the victim could easily detect this failure by the passive assailant (e.g., a husband's failure to stop his friend from taking his wife's car keys in full view of his wife) or indirect if the victim would have significant difficulty identifying the passive assailant (e.g., a store clerk “looking the other way” while an accomplice shoplifts merchandise). 3. Assessment of aggression The second fundamental problem in the assessment of aggression involves the use of assessment instruments that do not adequately correspond to a hybrid definition of aggression or conceptually distinct subtypes. That is, many assessment tools purport to measure aggressive behavior but may actually be sensitive to emotional (e.g., anger) or attitudinal constructs (e.g., hostility). Further, among the various measures that do not blur this conceptual distinction, a lack of correspondence exists between item content and the aggression subtype purportedly being assessed. This problem is illustrated below by reviewing a select group of assessment strategies, including self-report inventories, observational measures, clinical interviews, and laboratory-based paradigms. This list is not exhaustive. Rather, it is meant to highlight these key limitations in the assessment of aggression. 3.1. Self-report inventories An instrument frequently employed to assess aggressive behavior is the Buss–Durkee Hostility Inventory (BDHI; Buss & Durkee, 1957). The BDHI is a 75-item true–false scale comprised of eight subscales: Assault, Indirect Hostility, Verbal Hostility, Irritability, Negativism, Resentment, Suspicion, and Guilt. Because many items do not assess aggressive acts, use of the BDHI total score is inappropriate for the assessment of aggressive behavior. As such, many researchers elect to use the Assault scale as an index of aggression. However, several limitations to this approach have been cited in the literature (Bendig, 1961; Biaggio, 1980; Bushman, Cooper, & Lemke, 1991). Most notably, research indicates that the BDHI is best conceptualized as two, highly related factors (e.g., expressive versus attitudinal, overt versus covert). Therefore, the Assault subscale does not appear to “stand on its own” as an index of aggressive behavior. Moreover, upon closer review of item content, some items do not specify the route in which aggression is expressed (e.g., “Once in a while I cannot control my urge to harm others” or “Whoever insults me or my family is just asking for a fight”). Finally, the Indirect and Verbal subscales possess some items that may be construed as aggressive (e.g., “I sometimes spread gossip about people I don't like” or “When people yell at me, I yell back”). However, these subscales possess far more items that do not directly assess aggressive behavior or specify the route by which aggression is expressed. The Verbal Aggressiveness Scale (VAS; Infante & Wigley, 1986) is a 20-item Likert-type scale that assesses the respondent's tendency to use verbally aggressive strategies when trying to influence others. Though not explicitly stated, the majority of items assess instances of active, direct, verbal aggression. Unfortunately, some items do not conform well to this category. For example, item #13 states: “I like poking fun at people who do things which are very stupid in order to stimulate their intelligence.” However, it is unclear whether “poking fun” is conducted directly (i.e., to the target's face) or indirectly (i.e., to peers of the target). Item #10 states: “When people criticize my shortcomings, I take it in good humor and do not try to get back at them.” A response indicating that this statement is almost never true (i.e., a respondent would almost always try to get back at the person) is coded as verbally aggressive. However, it remains unclear which tactic (e.g., physical, verbal, damage to property) the respondent would use to aggress toward
D.J. Parrott, P.R. Giancola / Aggression and Violent Behavior 12 (2007) 280–299
291
the target. Moreover, the route by which aggression would be expressed (i.e., direct–indirect, active–passive) is not specified. The Aggression Inventory (AI; Gladue, 1991) is a 28-item Likert-type scale that includes subscales for Physical and Verbal Aggression. However, a lack of clarity exists in these subscales. In the Physical Aggression subscale, several items assess acts that could be interpreted to be verbal as well as direct or indirect acts. For example, item #11 (“When another person is mean to me, I get even with him/her”) and item #22 (“I think it is OK to make trouble for an annoying person”) do not clearly indicate the route in which aggression would be delivered (i.e., direct–indirect, active–passive) or the tactic that would be used (e.g., physical, verbal, theft, etc.). Similarly, the Verbal Aggression subscale contains items that may reflect expressions of anger or assertiveness. For example, item #12 (“When a person is unfair to me, I get angry and protest”) does not specifically reflect the intent to harm the target and, most likely, assesses the general expression of anger. Similarly, item #20 (“When a person tries to ‘cut ahead’ of me in a line, I firmly tell him/her not to do so”) clearly does not reflect an intent to harm the target. Rather, it appears to measure one's level of assertiveness rather than direct verbal aggression. The Aggressive Acts Questionnaire (AAQ; Barratt et al., 1999) is a 22-item Likert-type scale that assesses qualities of the respondent's four most aggressive acts during the past 6 months. A major limitation of this scale is that aggressive acts are defined to the respondent as “hitting or verbally insulting another person or breaking objects because you are angry or frustrated.” Therefore, while active acts of direct physical and verbal aggression are included in this definition, respondents may also rate acts as aggressive despite the fact they were not perpetrated with the intent to harm another individual (e.g., breaking objects). Such acts were labeled by Buss (1961) as nonaggressive destruction. In addition, the primary function of this instrument is to differentiate between impulsive and premeditated aggression. As previously noted, this theoretical distinction possesses several limitations, including the fact that an aggressive act may be motivated by several goals in addition to harming the individual (Bushman & Anderson, 2001). The Past Feelings and Acts of Violence scale (PFAV; Plutchik & van Pragg, 1990) is a 12-item Likert-type scale designed to assess the respondent's past history of aggression. While five of the twelve items assess active, direct physical aggression (e.g., “Have you ever used a weapon to harm anyone?”), the remaining items clearly do not assess aggression in the broad sense or any proposed subtype of aggressive behavior. Rather, items assess one's tendency to become angry (e.g., “Do you find that you get angry very easily?”) or their access to weapons (e.g., “Do you keep weapons in your home that you know how to use?”). The Revised Conflict Tactics Scale (CTS-2; Straus, Hamby, Bony-McCoy, & Sugarman, 1996) is a 78-item, selfreport scale designed to assess strategies that respondents use to reduce conflict with their partners. The CTS-2 is composed of five subscales, including Physical Assault and Psychological Aggression. The items on these subscales adhere well to operational definitions of aggression in that they clearly reflect the goal of harming one's partner in some way. However, given the proposed taxonomic system, this scale is limited by the fact that it does not distinguish between the route of expression (e.g., direct–indirect) or between physical and verbal acts. For example, some items reflect direct physical aggression (e.g., “Have you choked your partner?”) while others reflect potentially indirect aggressive tactics (e.g., “Have you done something to spite your partner?”). In addition, the Physical Assault subscale contains the item, “Have you insulted or sworn at your partner?” that clearly assesses active, direct, verbal aggression. Likewise, the Psychological Aggression subscale contains items that clearly assess physical acts such as “Have you thrown something at your partner that could hurt?” or “Have you destroyed something that belonged to your partner?” Indeed, this latter item adheres well to the proposed damage to property subtype. In summary, these six measures possess significant limitations with regard to their adherence to a hybrid definition of aggressive behavior or its subtypes. Specifically, three of the six instruments (AI, AAQ, PFAV) include items that assess acts that do not intentionally cause harm to another individual. Two measures (AI, PFAV) possess items that assess one's experience of anger rather than their display of aggression. Moreover, five of the scales (BDHI, AI, AAQ, PFAV, CTS-2) do not appropriately distinguish between previously proposed subtypes of aggression and, by extension, do not adhere to the proposed taxonomic system. While the VAS purportedly assesses verbal aggression, it is not clear that certain items actually correspond with this subtype and, if so, whether these items reflect direct or indirect acts. In sum, these limitations result in the grouping together of aggressive acts that may otherwise be theoretically distinguished from one another. In an effort to update the BDHI and make it more compatible with modern definitions of aggression, Buss and Perry (1992) developed the Buss–Perry Aggression Questionnaire (BAQ). Indeed, the BAQ adheres particularly well to our conceptualization of aggression and its subtypes. This 29-item Likert-type scale assesses an individual's disposition
292
D.J. Parrott, P.R. Giancola / Aggression and Violent Behavior 12 (2007) 280–299
toward physical aggression, verbal aggression, anger, and hostility. The BAQ clearly makes the theoretical distinction between aggressive behavior (either physical or verbal) and the related constructs of anger and hostility. However, while many researchers appropriately employ these subscales, others report the total score of the questionnaire as an index of “aggression” (e.g., Adams, Wright, & Lohr, 1996; Dougherty, Bjork, Marsh, & Moeller, 1999; Tremblay & Belchevski, 2004). When used in this way, researchers are not adhering a hybrid operational definition of aggressive behavior and thereby reduce the conceptual clarity of the instrument. Notably, BAQ items that comprise the physical and verbal aggression subscales reflect active, direct aggressive behavior. As such, it will be important for future studies to acknowledge this fact when interpreting results. In doing so, researchers may elucidate those risk factors that predict active, direct aggression relative to other routes by which aggression may be expressed. A second measure that conforms well to the proposed taxonomic system is the Richardson Conflict Response Questionnaire (RCRQ; Richardson & Green, 2003). This 28-item Likert-type scale assesses the respondent's past history (e.g., over the past 6 months) of direct and indirect forms of physical and verbal aggression. Notably, all items represent active forms of aggression. Thus, this scale permits researchers to determine whether particular risk factors differentially influence various manifestations of active aggression. The BAQ and RCRQ possess two primary strengths. First, the item content of these measures conforms well to a hybrid definition of aggressive behavior (e.g., Anderson & Bushman, 2002; Baron & Richardson, 1994; Berkowitz, 1993; Geen, 2001). Second, each measure possesses appropriate correspondence between item content and the subtype of aggression purportedly being assessed. One minor limitation of these measures is that, although they assess active forms of aggression, they do not explicitly state this fact. This may lead to a lack of conceptual clarity for researchers attempting to investigate passive subtypes of aggression. It is recommended that future studies on scale development build upon these strengths while simultaneously minimizing the primary limitations of existing measures. 3.2. Observational techniques The Modified Overt Aggression Scale (MOAS; Kay, Wolkenfield, & Murrill, 1988) is a 20-item observational rating scale designed to assess aggression in psychiatric populations. Various subtypes of aggression are identified, including physical aggression, verbal aggression, aggression against property, and autoaggression. While the authors do not explicitly state whether aggressive acts are direct/indirect or active/passive, closer inspection of item content reveals that the physical aggression subscale reflects active, direct acts of physical aggression. However, significant limitations exist in the remaining subscales. First, in the verbal aggression subscale, it is unclear whether acts are direct or indirect, or whether they are aggressive at all. For example, raters are asked to indicate whether the patient “shouts angrily, curses mildly, or makes personal insults.” Indeed, endorsement of this item may reflect an instance in which the patient curses at another patient (i.e., active, direct, verbal aggression) or in which a patient is simply shouting angrily at nobody in particular (i.e., no intention to harm another person). Second, in the aggression against property subscale, acts may reflect active, indirect, damage to property (e.g., surreptitiously ripping someone's favorite shirt to ‘get back at them’). Alternatively, such destruction of property may not be motivated to harm another individual at all. The Social Dysfunction and Aggression Scale (SDAS; Wistedt et al., 1990) is an 11-item observational rating scale that assesses outward and inward aggression. Though not explicitly stated, all items included in the outward aggression subscale involve active forms of aggressive behavior. However, several significant limitations exist. First, some items clearly do not conform to the hybrid definition of aggression. Specifically, there is no indication to suggest that the perpetrator intended to harm a target (e.g., “general shouting, screaming, swearing,” “lack of feeling for situations,” or “uncooperative”). Second, for certain items, it remains unclear whether the observed physical behavior was intended to directly or indirectly harm the target (e.g., “kicking furniture, destroying things”). Third, some behaviors that may be better categorized as direct physical aggression, direct verbal aggression, or direct/indirect damage to property are combined within the outward aggression subscale. This compromises the conceptual clarity of the instrument. In sum, while this scale purports to assess aggressive behavior, it appears that other emotional and nonaggressive behavioral constructs are also being measured. The Staff Observation Aggression Scale (SAOS; Palmstierna & Wistedt, 1987) was designed to assess the degree and frequency of aggressive acts among psychiatric inpatients. Most of the possible response options reflect active forms of direct physical or verbal aggression. Pertinently, the rater is able to discriminate between acts directed toward harming others, acts with no particular human target, and acts directed toward objects. However, all of these acts are considered to be aggressive, and the scale does not indicate whether behaviors directed toward objects represent nonaggressive destruction or are motivated by the intent to harm another person (i.e., damage to property).
D.J. Parrott, P.R. Giancola / Aggression and Violent Behavior 12 (2007) 280–299
293
In summary, these measures appear to assess aggressive behavior as well as anger and nonaggressive destruction. This broad conceptualization of aggressive behavior likely has great utility in residential or inpatient settings. However, in research settings, a precise operational definition of aggression and its subtypes is necessary in order to test existing theories, develop new theoretical models, and enhance clinical interventions. As such, use of these observational measures in research settings poses significant challenges. To remedy these limitations, revisions of these or similar scales could differentiate between “anger outbursts” that have no direct target and “aggressive outbursts” that are intended to harm another individual. Indeed, delineation of these different types of “outbursts” lies at the heart of the development of diagnostic criteria for distinct anger and aggressive disorders (Eckhardt & Deffenbacher, 1995). In addition, noting the route in which aggression is delivered (i.e., direct–indirect) and the type of aggressive tactic that is employed (e.g., physical, verbal, damage to property) would greatly increase researchers' ability to refine existing theories of aggressive behavior. 3.3. Clinical interviews The Direct and Indirect Aggression Scales (DIAS; Bjorkqvist, Lagerspetz, & Osterman, 1992) constitute a structured interview designed to assess aggression in children. Several subscales have been identified through factor analyses, including Physical aggression (e.g., hits, kicks, shoves), Verbal Aggression (e.g., yells, insults, calls the other names), and Indirect Aggression (e.g., gossips, ignores, says bad things being the back). While the physical and verbal aggression subscales seem to adequately represent active forms of direct physical and verbal aggression, the indirect aggression subscale lacks sufficient homogeneity. Specifically, acts such as “ignoring” reflect a passive act of aggression, whereas acts such as gossiping or saying negative things behind one's back are quite active. In addition, some acts included in this subscale may be better conceptualized as direct (e.g., criticizes the other one's hair or clothing) or indirect verbal aggression (e.g., gossips, tells the other one's secret to a third person). Therefore, while this subscale makes great strides to measure indirect forms of aggressive behavior, a lack of conceptual clarity remains due to the grouping together of active versus passive and direct versus indirect acts. The Life History of Aggression assessment (LHA; Coccaro, Berman, & Kavoussi, 1997) is a semi-structured interview consisting of three subscales: Aggression, Consequences/Antisocial Behavior, and Self-Directed Aggression. The Aggression subscale includes the following five categories within which behaviors may be assigned: a) verbal aggression, b) indirect aggression (aggression directed toward inanimate objects), c) non-specific fighting, d) physical assault against people (with evidence of intent to harm), and e) temper tantrums. While the non-specific fighting and physical assault against people items clearly assess active and direct expressions of physical aggression, the remaining items are less well defined. For example, it is unclear whether the verbal aggression item reflects direct or indirect behavior. Further, the indirect aggression item may reflect active, indirect damage to property (e.g., slashing tires) or property destruction without the intent to harm the owner. Finally, the temper tantrum item appears to reflect one's disposition to experience anger and implies that the subject is not directing his/her anger towards any particular target. The remaining subscales also pose threats to the conceptual clarity of the scale as a measure of aggression, in that the majority of their items do not directly reflect displays of aggression. For example, items from the Consequences/ Antisocial Behavior subscale include “antisocial behavior not resulting from police involvement” (e.g., selling drugs, prostitution) and “problems with supervisors” (e.g., demotions, firings, warnings). Likewise, the Self-Directed Aggression subscale includes self-injurious and suicidal behavior. Given that the individual is not attempting to avoid these behaviors, they do not constitute an aggressive act (Baron & Richardson, 1994). Taken together, it appears that use of the LHA total score to index trait aggressivity is misleading, as many of the items do not assess actual aggressive behaviors. Moreover, with regard to the Aggression subscale, it is difficult to determine which subtypes of aggressive behavior are being assessed. In summary, the limitations of these clinical interviews again highlight a fundamental problem in the assessment of aggression. Specifically, the behaviors that are assessed either do not correspond to a precise definition of aggressive behavior or are not placed within an appropriate categorical framework that differentiates meaningful subtypes of aggressive behavior. As a result, these and similar interview techniques blur the theoretical distinction between aggression and related constructs. Use of such instruments also creates challenges for those investigators aiming to determine different etiological pathways to various subtypes of aggression. To address these limitations, future research is needed in order to develop interview strategies that more precisely assess acts that cause harm to others and identify the route in which these acts are expressed.
294
D.J. Parrott, P.R. Giancola / Aggression and Violent Behavior 12 (2007) 280–299
3.4. Laboratory paradigms 3.4.1. Physical aggression At present, there exist two well-known laboratory measures of physical aggression. The first, termed the Buss Aggression Machine or the Teacher–Learner Aggression Paradigm, was developed by Buss (1961). Participants are seated in a room and are told that they will be working on a mathematical task with a supposed “partner” in an adjacent room. During each trial of the Teacher–Learner task, the “partner” must answer a particular arithmetic question. Participants are told that their job is to help their partner solve some arithmetic problems by administering electric shocks to this person each time s/he provides an incorrect answer. If the response to the question is correct, the participant reinforces the partner by illuminating a light. However, if the response is incorrect, the participant administers a shock to the partner. With regard to our taxonomic system, this task measures direct and active expressions of physical aggression. However, as will be seen below, there is some debate about whether aggression on this task is confounded by other constructs. The second task is the Taylor Aggression Paradigm (TAP; Taylor, 1967), which was developed in response to the criticisms of the Buss Teacher–Learner task (see below). Participants are told that they will compete against an opponent (seated in an adjacent room) on a reaction-time task. Depending on whether participants win or lose a reaction-time trial, they will either administer or receive an electric shock to/from their opponent. Of course, no opponent actually exists and the sequence of wins and loses as well as the intensity of the shocks set by the “opponent” is predetermined by the experimenter. The intensity of shocks given to participants increases across trials in order to determine the effects of increasing provocation on aggressive responding. The bogus competition merely exists to deceive individuals into believing that they are engaged in a competitive and adversarial relationship with another person. The TAP operationalizes aggression as the average shock intensity selected over trials and, as such, can be said to measure direct and active expressions of physical aggression within our taxonomic system. More recent modifications of the TAP have added the time duration of shock delivered as another dependent measure of aggression (Giancola & Zeichner, 1994; Zeichner & Pihl, 1979). Other shock-aggression tasks have also used the frequency of shocks administered as a dependent variable (Berkowitz & LePage, 1967; Zeichner, Frey, Parrott, & Butryn, 1999). The TAP was constructed in response to criticisms directed at the Buss Teacher–Learner task. The main difference between the two paradigms is that the TAP creates an adversarial relationship between the participant and the opponent by instituting a competitive task that results in the participant being provoked by the opponent and then being allowed to retaliate. On the other hand, the Buss task requires participants to take the role of a teacher so that they can “help” their “partner” learn a particular task. The main criticism marshaled against the Buss Teacher–Learner task is that responses may be influenced by altruistic intentions inasmuch as the goal of the paradigm is to help the partner learn a task rather than engage him/her in an adversarial competition (Baron & Eggleston, 1972; Taylor, Gammon, & Capasso, 1976). Finally, the fact that participants do not receive shocks during the Teacher–Learner task, and are therefore not provoked to behave in an aggressive manner, may also serve to mitigate the desired response (i.e., aggression). 3.4.2. Theft A third commonly used laboratory measure of aggression is the Point Subtraction Aggression Paradigm (PSAP). The PSAP was developed by Cherek (1981) and has its origins in experimental behavioral pharmacology research. The PSAP requires participants to sit in front of a response panel typically containing two buttons. Pressing the first button 100 times earns the individual a point which is worth a specified amount of money and pressing the second button 10 times subtracts one point from the participant's fictitious opponent. Points are also deducted from participants by their “opponent” on a predetermined basis. The dependent measure of aggression is the number of times the point subtraction button is pressed. The main difference between the TAP and the PSAP is the way in which aggression is operationalized. Whereas the TAP operationalizes aggression as a physical retaliation to provocation in the form of electric shock, the PSAP operationalizes aggression as the removal of points (money) from a fictitious opponent. One can therefore argue that the two tasks measure different types of aggression: physical versus theft. Loosely stated, one can also argue that the Taylor task employs a more “harsh” expression of aggression than the Point Subtraction task. Given this, it is important to note that the threshold for a physical attack is probably higher than that for a monetary attack. A recent study presents some data and reviews evidence to support this contention (Gussler-Burkhardt & Giancola, 2005). Within our taxonomic system, the PSAP measures active and indirect theft. We have classified the PSAP as an indirect measure of aggression inasmuch as participants are told that their opponents are unknown to them and located
D.J. Parrott, P.R. Giancola / Aggression and Violent Behavior 12 (2007) 280–299
295
in a nearby building. If the opponent could be viewed as in the case of the TAP, the task would then be a measure of direct aggression. Nevertheless, as noted in the description of our taxonomic system, the PSAP appears to measure theft that is motivated by the intent to harm another person. An analogous case of “real world” aggression would involve an individual stealing another's car keys from their office so that they could not drive home. 3.4.3. Verbal aggression A number of experimental paradigms have been developed to assess direct and indirect expressions of verbal aggression. Participants are typically led to believe that they are engaged in the experiment with another person (i.e., a confederate). Alternatively, some studies involve the use of research assistants, or even the experimenter, as a provocateur to whom the participant may aggress. By implementing a cover story for the study, the experimenter is able to control the route by which the participant aggresses. For example, following some form of provocation, the participant may make comments directly to the confederate (direct) or rate the provocateur on a self-report scale (indirect). All such paradigms measure active expressions of aggressive behavior. A more detailed examination of the methodology employed by these paradigms illustrates the precision in which subtypes of aggressive behavior may be assessed. For example, to assess direct verbal aggression, a confederate may verbally provoke the participant or make socially undesirable comments. Next, the experimenter records the participant's vocal responses to the confederate. These responses are then coded for aggressive content. Because the participant believes that the confederate can hear these comments through an intercom system, any display of verbal aggression may be considered direct. To assess indirect verbal aggression, the participant may be instructed to rate the provocateur (e.g., the experimenter) by completing a self-report form. The participant is led to believe that the experimenter will not see the ratings and that the director of the laboratory will use them for making future hiring decisions. As a result, the participant may inflict harm on the provocateur while simultaneously avoiding direct detection. 3.4.4. Summary As with some self-report measures, laboratory aggression paradigms possess several strengths. First, they assess behaviors that are directed toward the goal of harming another individual. With respect to the TAP and its modified versions, shocks are administered in order to harm one's opponent (although not seriously). Moreover, given the competitive nature of the task, it is clear that the recipient of the shocks is motivated to avoid their noxious and aggravating effects. Regarding the PSAP, whereas the subtraction of points from an opponent may not be physical in nature, the event is nonetheless threatening and harmful to the individual's ego and certainly harmful in a financial sense. Moreover, based on observations of individuals competing on this paradigm, it is clear that they are highly motivated to avoid losing points/money (Cherek, personal communication, 1995). Given these descriptions, we would argue that both of these tasks conform to our definition of aggression. Second, with the exception of the Teacher– Learner task, each paradigm appears to possess a sufficient degree of conceptual clarity. For example, it is readily apparent that the TAP is measuring active, direct physical aggression and not some other type of behavior (e.g., damage to property). Similarly, laboratory paradigms can effectively differentiate the various routes that aggression is expressed (e.g., direct–indirect) and the type of aggressive tactic that is employed (e.g., verbal aggression). Given the lack of conceptual clarity in many other measures of aggressive behavior, this strength is especially noteworthy. On balance, critics have argued that laboratory aggression tasks/paradigms have a number of limitations. These criticisms have been debated by scientists for a number of years (Anderson & Bushman, 1997; Anderson, Lindsay, & Bushman, 1999; Berkowitz, 1989; Berkowitz & Donnerstein, 1982; Giancola and Chermack, 1998; Gottfredson & Hirschi, 1993; Kane, Joseph, & Tedeschi, 1976; Orne, 1962; Tedeschi & Quigley, 1996, 2000). However, the most often cited criticism of laboratory paradigms of aggression is that they do not measure what is known as “real world” aggression. We will not debate this issue here inasmuch as both sides of the argument are well represented in the citations above. However, we will say that the measurement of aggression must reflect the multifaceted nature of the construct. We, therefore, advocate for the use of multiple measures of aggression with the strict stipulation that the measures be supported by empirical data demonstrating their validity. 4. Implications for diagnosis and treatment At present, the only DSM-IV (American Psychiatric Association, 1994) diagnosis characterized primarily by aggressive behavior is Intermittent Explosive Disorder (IED). Unfortunately, compared to anxiety and mood disorders,
296
D.J. Parrott, P.R. Giancola / Aggression and Violent Behavior 12 (2007) 280–299
relatively little research has focused on the diagnosis and treatment of IED. Perhaps due to this fact, the DSM-IV currently includes 11 Axis I disorders for anxiety-related problems and nine Axis I disorders for mood disorders. Indeed, Coccaro, Kavoussi, Berman, and Lish (1998) note that “despite the common occurrence of inappropriate, impulsive–aggressive behavior in our society, the [DSM-IV] continues to be deficient in identifying individuals with problematic behaviors of this type” (pg. 368). It may be argued that the relative paucity of research in this area corresponds to the DSM-IV claim that IED is “apparently rare” (pg. 611). However, epidemiological data suggest otherwise. Research indicates that the lifetime prevalence of IED ranges from 3.3% (or 9.3 million Americans) to 9.3% (or 26 million Americans) (Coccaro, Schmidt, Samuels, & Nestadt, 2004). Indeed, these data indicate that IED is by no means a “rare” psychological disorder. In addition, when coupled with epidemiological reports that speak to the pervasiveness of violence in the general population (U.S. Department of Justice, 2002), these data underscore the urgent need to develop a more comprehensive classification system for aggressive behaviors. In response to this gap in the current diagnostic literature, it is necessary to initiate a systematic research program to determine whether the creation of distinct “aggressive disorders” is warranted. In support of this view, researchers have recently called for the DSM to include of a class of disorders characterized primarily by aggressive behavior (see review, by Widiger & Sankis, 2000). Such proposals echo the movement to adopt diagnostic criteria for “anger disorders” that has been underway for several years (Eckhardt & Deffenbacher, 1995). The proposed classification system is the necessary first step toward achieving this aim. Without a theoretically sound taxonomy of aggressive acts, a diagnostic class of aggressive disorders would be built upon a weak foundation and ultimately be of little clinical value. Indeed, the current lack of conceptual clarity in the aggression literature does not provide an adequate framework for the development of theoretically grounded diagnostic criteria. In contrast, the proposed classification system possesses clear theoretical distinctions between subtypes of aggressive behavior. As discussed earlier, IED is limited by the fact that it blurs the conceptual distinction between aggressive behavior and the expression of anger. By adopting the hybrid definition of aggressive behavior reviewed above (Anderson & Bushman, 2002; Baron & Richardson, 1994; Berkowitz, 1993; Geen, 2001), the proposed taxonomic system permits a precise class of disorders based on aggression to be distinguished from a class of disorders based upon anger (Eckhardt & Deffenbacher, 1995). Although much research needs to be done to determine the exact diagnostic criteria that would constitute disorders of aggression, the proposed taxonomic system puts forth several possibilities. For example, diagnosis of an “aggressive episode” would be characterized by any behavior motivated by the goal of harming another living being and could include the various subtypes proposed herein. These criteria are currently absent in the diagnosis of IED. In addition, diagnoses may be distinguished by the route in which the perpetrator tends to express his or her aggression. This possibility is especially likely given the fact that pertinent literature has identified several individual difference variables that differentially predict direct and indirect aggression (Archer, 2004a; Kaukiainen et al., 1999; Richardson & Green, 2003; Walker et al., 2000). Similarly, to the extent that the proposed system stimulates new research on determinants of passive aggression, distinct diagnostic criteria for “passive–aggressive” disorders may be envisioned. Notably, the DSM-IV includes Passive–aggressive Personality Disorder as a diagnostic entity requiring further study. This disorder is characterized by a negativistic personality style that can include overt acts of aggression. Although recent research supports these criteria (Berman, Fallon, & Coccaro, 1998), this disorder is not characterized by one's lack of behavior that subsequently causes harm. Thus, to the extent that research continues to support this disorder, the proposed taxonomic system suggests that it may be more appropriately termed “Indirect–Aggressive Personality Disorder”. This change in terminology highlights that fact that aggression may be expressed overtly (i.e., actively) yet in a circuitous, more indirect manner. 5. Summary Research on aggressive behavior currently suffers from problems in measurement. We contend that two fundamental issues underlie these limitations. First, operational definitions of aggression and its various subtypes lack sufficient conceptual clarity. For example, constructs such as anger or hostility are oftentimes included in the conceptualization of aggression. In addition, subtypes of aggression frequently include acts that may be assigned to more than one category. Second, many assessment instruments used in the study of aggression do not appropriately
D.J. Parrott, P.R. Giancola / Aggression and Violent Behavior 12 (2007) 280–299
297
correspond to precise definitions of aggression or its subtypes. Collectively, these definitional problems pose significant challenges for the identification of risk factors that differentially predict aggressive subtypes. The development of theories, diagnostic systems, and clinical interventions is also comprised. As a first step toward resolving these limitations, a new taxonomic system is proposed. The development of this classification system was based upon a hybrid definition of aggressive behavior, clear distinctions between subtypes of aggressive behavior, and pertinent empirical findings. Despite adherence to these criteria, it is clearly necessary for future research to validate further the proposed taxonomy. For example, factor analytic techniques could be utilized to corroborate existing literature suggesting that the various dimensions and subtypes of aggressive acts discussed herein are, in fact, distinct. Correlational techniques may also be employed to examine theoretically derived relations between dispositional and/or situational factors and various subtypes of aggressive acts. Indeed, while studies employing these methods have empirically differentiated between direct and indirect aggression (e.g., Richardson & Green, 2003) and physical and verbal aggression (e.g., Vaillancourt, Brendgen, Boivin, & Tremblay et al., 2003), additional work directed toward passive expressions of aggression as well as postural, damage to property, and theft subtypes is needed. Despite the need for additional validation, this proposed approach to the categorization of aggressive acts will bring greater conceptual clarity to the assessment of aggression and advance empirical, theoretical, diagnostic, and clinical work in this highly complex area. References Adams, H. E., Wright, L. W., & Lohr, B. A. (1996). Is homophobia associated with homosexual arousal? Journal of Abnormal Psychology, 105, 440−445. American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders, (4th ed.). Washington DC: Author. Anderson, C., & Bushman, B. (2002). Human aggression. Annual Review of Psychology, 53, 27−51. Anderson, C. A., & Bushman, B. J. (1997). External validity of “trivial” experiments: The case of laboratory aggression. Review of General Psychology, 1, 19−41. Anderson, C. A., Lindsay, J., & Bushman, B. J. (1999). Research in the psychological laboratory: Truth or triviality? Current Directions in Psychological Science, 8, 3−9. Archer, J. (2001). A strategic approach to aggression. Social Development, 10, 267−271. Archer, J. (2004a). Sex differences in aggression in real-world settings: A meta-analytic review. Review of General Psychology, 8, 291−322. Archer, J. (2004b). Which attitudinal measures predict trait aggression? Personality and Individual Differences, 36, 47−60. Bandura, A. (1973). Aggression: A social learning theory analysis. Englewood Cliffs, NJ: Prentice Hall. Baron, R. A. (1977). Human aggression. New York, NY: Plenum Press. Baron, R., & Eggleston, R. (1972). Performance on the “aggression machine”: Motivation to help or harm? Psychonomic Science, 26, 321−322. Baron, R. A., & Neuman, J. H. (1996). Workplace violence and workplace aggression: Evidence of their relative frequency and potential causes. Aggressive Behavior, 22, 161−173. Baron, R. A., & Richardson, D. R. (1994). Human aggression, (2nd Edition). New York, NY: Plenum Press. Barratt, E. S., & Slaughter, L. (1998). Defining, measuring, and predicting impulsive aggression: A heuristic model. Behavioral Sciences and the Law, 16, 285−302. Barratt, E. S., Stanford, M. S., Dowdy, L., Liebman, M. J., & Kent, T. A. (1999). Impulsive and premeditated aggression: A factor analysis of selfreported acts. Psychiatry Research, 86, 163−173. Bendig, A. Q. (1961). A factor analysis of scales of emotionality and hostility. Journal of Clinical Psychology, 22, 189−192. Berkowitz, L. (1989). Laboratory experiments in the study of aggression. In J. Archer & K. Browne (Eds.), Human aggression: Naturalistic approaches (pp. 42−61). New York: Routledge. Berkowitz, L. (1993). Aggression: Its causes, consequences, and control. New York: McGraw Hill. Berkowitz, L., & Donnerstein, E. (1982). External validity is more than skin deep. American Psychologist, 37, 245−257. Berkowitz, L., & LePage, A. (1967). Weapons as aggression-eliciting stimuli. Journal of Personality and Social Psychology, 7, 202−207. Berman, M. E., Fallon, A. E., & Coccaro, E. F. (1998). The relationship between personality psychopathology and aggressive behavior in research volunteers. Journal of Abnormal Psychology, 107, 651−658. Biaggio, M. K. (1980). Assessment of anger arousal. Journal of Personality Assessment, 44, 289−298. Bjorkqvist, K. (2001). Different names, same issue. Social Development, 10, 272−275. Bjorkqvist, K., Lagerspetz, K. J., & Osterman, K. (1992). The Direct and Indirect Aggression Scales. Vasa, Finland: Abo Akademi University, Department of Social Sciences. Bjorkqvist, K., Osterman, K., & Kaukiainen, A. (1992). The development of direct and indirect aggressive strategies in males and females. In P. Niemela & K. Bjorkqvist (Eds.), Of mice and women: Aspects of female aggression (pp. 51−64). San Diego, CA: Academic Press. Bushman, B. J., & Anderson, C. A. (2001). Is it time to pull the plug on the hostile versus instrumental aggression dichotomy? Psychological Review, 108, 273−279. Bushman, B. J., Cooper, H. M., & Lemke, K. M. (1991). Meta-analysis of factor analyses: An illustration using the Buss–Durkee Hostility Inventory. Personality and Social Psychology Bulletin, 17, 344−349.
298
D.J. Parrott, P.R. Giancola / Aggression and Violent Behavior 12 (2007) 280–299
Buss, A. (1961). The psychology of aggression. New York: John Wiley. Buss, A. H. (1995). Personality: Temperament, social behavior, and the self. Boston, MA: Allen and Bacon. Buss, A. H., & Durkee, A. (1957). An inventory for assessing different kinds of hostility. Journal of Consulting Psychology, 21, 343−349. Buss, A. H., & Perry, M. (1992). The aggression questionnaire. Journal of Personality and Social Psychology, 63, 452−459. Cherek, D. (1981). Effects of smoking different doses of nicotine on human aggressive behavior. Psychopharmacology, 75, 339−345. Coccaro, E. F., Berman, M. W., & Kavoussi, R. J. (1997). Psychiatry Research, 73, 147−157. Coccaro, E. F., Kavoussi, R. J., Berman, M. E., & Lish, J. D. (1998). Intermittent explosive disorder-revised: development, reliability, and validity of research criteria. Comprehensive Psychiatry, 39, 368−376. Coccaro, E. F., Schmidt, C. A., Samuels, J. F., & Nestadt, G. (2004). Lifetime and 1-month prevalence rates intermittent explosive disorder in a community sample. Journal of Clinical Psychiatry, 65, 820−824. Coie, J. D., & Dodge, K. A. (1998). Aggression and antisocial behavior. In N. Eisenberg (Ed.), Handbook of Child Psychology (pp. 779−862). New York: Wiley. Crick, N. R. (1996). The role of overt aggression, relational aggression, and prosocial behavior in the prediction of children's future social adjustment. Child Development, 67, 2317−2327. Crick, N. R., & Grotpeter, J. K. (1995). Relational aggression, gender, and social–psychological adjustment. Child Development, 66, 710−722. Dodge, K. A. (1991). The structure and function of reactive and proactive aggression. In D. Pepler & K. Rubin (Eds.), The Development and Treatment of Childhood Aggression (pp. 201−218). Hillsdale, NJ: Erlbaum. Dollard, J., Doob, L. W., Miller, N. E., Mowrer, O. H., & Sears, R. R. (1939). Frustration and aggression. New Haven, CT: Yale University Press. Dougherty, D. M., Bjork, J. M., Marsh, D. M., & Moeller, F. G. (1999). Influence of trait hostility on tryptophan depletion-induced laboratory aggression. Psychiatry Research, 88, 227−232. Eckhardt, C., Norlander, B., & Deffenbacher, J. (2004). The assessment of anger and hostility: A critical review. Aggression and Violent Behavior, 9, 17−43. Eckhardt, C. I., & Deffenbacher, J. L. (1995). Diagnosis of anger disorders. In H. Kassinove (Ed.), Anger disorders: Definition, diagnosis, and treatment (pp. 27−48). Washington, DC: Taylor and Francis. Fabiano, G. A., Pelham, W. E., Manos, M. J., Gnagy, E. M., Chronis, A. M., Onyango, A. N., et al. (2004). An evaluation of three time-out procedures for children with attention deficit/hyperactivity disorder. Behavior Therapy, 35, 449−469. Feshbach, N. (1964). The function of aggression and the regulation of aggressive drive. Psychological Review, 71, 257−272. Feshbach, N. (1969). Gender differences in children's modes of aggressive responses toward outsiders. Merrill–Palmer Quarterly, 15, 249−258. Galen, B. R., & Underwood, M. K. (1997). A developmental investigation of social aggression among children. Developmental Psychology, 33, 589−600. Geen, R. (2001). Human aggression, (2nd ed.). Buckingham: Open University Press. Giancola, P. R., & Zeichner, A. (1994). Neuropsychological performance on tests of frontal-lobe functioning and aggressive behavior in men. Journal of Abnormal Psychology, 103, 832−835. Giancola, P. R., & Chermack, S. T. (1998). Construct validity of laboratory aggression paradigms: A response to Tedeschi and Quigley (1996). Aggression and Violent Behavior, 3, 237−253. Gladue, B. A. (1991). Aggressive behavioral characteristics, hormones, and sexual orientation in men and women. Aggressive Behavior, 17, 313−326. Glover, D., Gough, G., Johnson, M., & Cartwright, N. (2000). Bullying in 25 secondary schools: Incidence, impact, and intervention. Educational Research, 42, 141−156. Gottfredson, M., & Hirschi, T. (1993). A control theory interpretation of psychological research on aggression. In R. Felson & J. Tedeschi (Eds.), Aggression and violence: Social interactionist perspectives (pp. 47−68). Washington DC: APA. Gussler-Burkhardt, N. L., & Giancola, P. R. (2005). A further investigation of gender differences in alcohol-related aggression. Journal of Studies on Alcohol, 66, 413−422. Harre, R., & Lamb, R. (1983). The encyclopaedic dictionary of psychology. Great Britain: Basil Blackwell Publisher Limited. Infante, D. A., & Wigley, C. J. (1986). Verbal aggressiveness: An interpersonal model and measure. Communication Monographs, 53, 61−69. Ireland, J. L. (2002). Official records of bullying incidents among young offenders: What can they tell us and how useful are they? Journal of Adolescence, 25, 669−679. Ireland, J. L. (2004). Nature, extent, and causes of bullying among personality-disordered patients in a high-secure hospital. Aggressive Behavior, 30, 229−242. Kane, T., Joseph, J., & Tedeschi, J. (1976). Person perception and the Berkowitz paradigm for the study of aggression. Journal of Personality and Social Psychology, 33, 663−673. Kaukiainen, A., Bjorkqvist, K., Lagerspetz, K., Osterman, K., Salmivalli, C., Rothberg, S., et al. (1999). The relationships between social intelligence, empathy, and three types of aggression. Aggressive Behavior, 25, 81−89. Kay, S. R., Wolkenfield, F., & Murrill, L. M. (1988). Profiles of aggression among psychiatric patients I. Nature and prevalence. Journal of Nervous and Mental Diseases, 176, 539−546. McEllistrem, J. E. (2004). Affective and predatory violence: A bimodal classification system of human aggression and violence. Aggression and Violent Behavior, 10, 1−30. Murphy, C. M., & O'Leary, K. D. (1989). Psychological aggression predicts physical aggression in early marriage. Journal of Consulting and Clinical Psychology, 57, 579−582. Olweus, D. (1996). Bully/victim problems at school: Facts and effective intervention. Journal of Emotional and Behavioral Problems, 5, 15−22. Orne, M. (1962). On the social psychology of the psychological experiment: With particular reference to demand characteristics and their implications. American Psychologist, 17, 776−783.
D.J. Parrott, P.R. Giancola / Aggression and Violent Behavior 12 (2007) 280–299
299
Palmstierna, T., & Wistedt, B. (1987). Staff observation scale, SOAS: Presentation and evaluation. Acta Psychiatrica Scandinavica, 76, 657−663. Paqutte, J. A., & Underwood, M. K. (1999). Young adolescents' experiences of peer victimization: Gender differences in accounts of social and physical aggression. Merrill–Palmer Quarterly, 45, 233−258. Parrott, D. J., & Zeichner, A. (2003). Effects of hypermasculinity on physical aggression against women. Psychology of Men and Masculinity, 4, 70−78. Plutchik, R., & van Pragg, H. M. (1990). A self-report measure of violence risk, II. Comprehensive Psychiatry, 31, 450−456. Ramirez, J. M., & Andreu, J. M. (2003). Aggression's typologies. International Review of Social Psychology, 16, 145−159. Richardson, D. R., & Green, L. R. (1997). Circuitous harm: Determinants and consequences of nondirect aggression. In R. M. Kowalski (Ed.), Aversive interpersonal behaviors (pp. 171−189). New York, NY: Plenum Press. Richardson, D. R., & Green, L. R. (1999). Social sanction and threat explanations of gender effects on direct and indirect aggression. Aggressive Behavior, 25, 425−434. Richardson, D. R., & Green, L. R. (2003). Defining direct and indirect aggression: The Richardson Conflict Response questionnaire. International Review of Social Psychology, 16, 11−30. Rose, S. M. (2003). Community interventions concerning homophobic violence and partner violence against lesbians. Journal of Lesbian Studies, 7, 125−139. Schumacher, J. A., & Leonard, K. E. (2005). Husbands' and wives' marital adjustment, verbal aggression, and physical aggression as longitudinal predictors of physical aggression in early marriage. Journal of Consulting and Clinical Psychology, 73, 28−37. Sears, R. R. (1961). Relation of early socialization experiences to aggression in middle childhood. Journal of Abnormal Psychology, 63, 466−492. Spielberger, C. D., Jacobs, G., Russell, S., & Crane, R. (1983). Assessment of anger: The State-Trait Anger Scale. In J. N. Butcher & C.D. Spielberger (Eds.), Advances in personality assessment, Vol. 2 (pp. 159−187). Hillsdale, NJ: Erlbaum. Straus, M. A., Hamby, S. L., Bony-McCoy, S., & Sugarman, D. B. (1996). The revised Conflict Tactics Scales (CTS2): Development and preliminary psychometric data. Journal of Family Issues, 17, 283−316. Taylor, S. (1967). Aggressive behavior and physiological arousal as a function of provocation and the tendency to inhibit aggression. Journal of Personality, 35, 297−310. Taylor, S., Gammon, C., & Capasso, D. (1976). Aggression as a function of the interaction of alcohol and threat. Journal of Personality and Social Psychology, 34, 938−941. Tedeschi, J., & Quigley, B. (1996). Limitations of laboratory paradigms for studying aggression. Aggression and Violent Behavior: A Review Journal, 1, 163−177. Tedeschi, J., & Quigley, B. (2000). A further comment on the construct validity of laboratory paradigms: A response to Giancola and Chermack. Aggression and Violent Behavior, 5, 127−136. Tremblay, P. F., & Belchevski, M. (2004). Did the instigator intend to provoke? A key moderator in the relation between trait aggression and aggressive behavior. Aggressive Behavior, 30, 409−424. Underwood, M. K. (2003). Social aggression among girls. New York, NY: Guilford. Underwood, M. K., Galen, B. R., & Paquette, J. A. (2001). Top ten challenges for understanding gender and aggression in children: Why can't we all just get along? Social Development, 10, 248−266. United States Department of Justice (2002). Criminal victimization in the United States, 2002. Retrieved on May 11, 2005 from http://www.ojp.usdoj. gov/bjs/abstract/cvusst.htm Vaillancourt, T., Brendgen, M., Boivin, M., & Tremblay, R. E. (2003). A longitudinal confirmatory factor analysis of indirect and physical aggression: Evidence of two factors over time? Child Development, 74, 1628−1638. Walker, S., Richardson, D. S., & Green, L. R. (2000). Aggression among older adults: The relationship of interaction networks and gender role to direct and indirect responses. Aggressive Behavior, 26, 145−154. Weinshenker, N. J., & Siegel, A. (2002). Bimodal classification of aggression: Affective defense and predatory attack. Aggression and Violent Behavior, 7, 237−250. Widiger, T. A., & Sankis, L. J. (2000). Adult psychopathology: Issues and controversies. Annual Review of Psychology, 51, 377−404. Wistedt, B., Rasmussen, A., Pedersen, L., Malm, U., Traskman-Bendz, L., Wakelin, J., et al. (1990). The development of an observer-scale for measuring social dysfunction and aggression. Pharmacopsychiatry, 23, 249−252. Zeichner, A., & Pihl, R. (1979). Effects of alcohol and behavior contingencies on human aggression. Journal of Abnormal Psychology, 88, 153−160. Zeichner, A., Frey, F. C., Parrott, D. J., & Butryn, M. (1999). Measurement of laboratory aggression: A new response choice paradigm. Psychological Reports, 82, 1229−1237. Zillmann, D. (1979). Hostility and aggression. Hillsdale, NJ: Erlbaum.