Analog assessment of frustration tolerance: Association with self-reported child abuse risk and physiological reactivity

Child Abuse & Neglect 46 (2015) 121–131

Contents lists available at ScienceDirect

Child Abuse & Neglect

Research article

Analog assessment of frustration tolerance: Association with self-reported child abuse risk and physiological reactivity夽 Christina M. Rodriguez a,∗ , Mary Bower Russa b , John C. Kircher c a b c

University of Alabama at Birmingham, USA Grand Valley State University, USA University of Utah, USA

a r t i c l e

i n f o

Article history: Received 13 October 2014 Received in revised form 18 February 2015 Accepted 26 February 2015 Available online 19 March 2015 Keywords: Child maltreatment Physical abuse Child abuse potential Frustration intolerance Analog tasks Assessment

a b s t r a c t Although frustration has long been implicated in promoting aggression, the potential for poor frustration tolerance to function as a risk factor for physical child abuse risk has received minimal attention. Instead, much of the extant literature has examined the role of anger in physical abuse risk, relying on self-reports of the experience or expression of anger, despite the fact that this methodology is often acknowledged as vulnerable to bias. Therefore, the present investigation examined whether a more implicit, analog assessment of frustration tolerance specifically relevant to parenting would reveal an association with various markers of elevated physical child abuse risk in a series of samples that varied with regard to age, parenting status, and abuse risk. An analog task was designed to evoke parenting-relevant frustration: the task involved completing an unsolvable task while listening to a crying baby or a toddler’s temper tantrum; time scores were generated to gauge participants’ persistence in the task when encountering such frustration. Across these studies, low frustration tolerance was associated with increased physical child abuse potential, greater use of parent–child aggression in discipline encounters, dysfunctional disciplinary style, support for physical discipline use and physical discipline escalation, and increased heart rate. Future research directions that could better inform intervention and prevention programs are discussed, including working to clarify the processes underlying frustration intolerance and potential interactive influences that may exacerbate physical child abuse. © 2015 Elsevier Ltd. All rights reserved.

Classic psychological theory has long underscored the role of frustration in promoting aggression (Dollard, Doob, Miller, Mowrer, & Sears, 1939). Prominent theories have emphasized that frustration can arise from blocked end goal attainment, with frustrating situations that lead to negative affect having a particularly pronounced impact on the tendency for aggression (Berkowitz, 1989, 2012). In fact, goal blockages that are perceived as intentional and illegitimate (versus unintentional and justified) are among the most likely to lead to aggressive responding (Anderson & Huesmann, 2003; Berkowitz, 2012). Although frustration is often treated interchangeably with anger in the broader literature (e.g., Wranik & Scherer, 2010), within contemporary models of aggression (e.g., Berkowitz, 2012), frustration and other forms of negative affect (e.g., sadness, depression, irritability) are viewed as common precursors to anger and aggression. Anger and aggression are thus viewed as possible, but not inevitable, outcomes of experiencing frustration (Berkowitz, 2012). Although the triggers for these two

夽 This study was supported in part by a University of Utah Faculty Research and Creative Activities Grant and a University of North Carolina at Greensboro Faculty Research Grant. ∗ Corresponding author at: University of Alabama at Birmingham, Department of Psychology, 1720 2nd Avenue South, Birmingham, AL 35294, USA. http://dx.doi.org/10.1016/j.chiabu.2015.02.017 0145-2134/© 2015 Elsevier Ltd. All rights reserved.

122

C.M. Rodriguez et al. / Child Abuse & Neglect 46 (2015) 121–131

emotional reactions often appear to coincide in time, they are theoretically distinct: frustration ensues when an individual is thwarted in achieving a goal (Anderson & Bushman, 2002); anger appears to be the consequence of attributions of blame for an event (Power & Dalgleish, 2008). Thus, frustration due to circumstances not perceived as aversive or intentional may not precipitate anger (Anderson & Huesmann, 2003; Berkowitz, 2012). When parenting contexts are considered, child misbehavior (which is often judged as intentional and aversive) seems particularly likely to evoke reactions of frustration and/or anger in parents. Indeed, when faced with what is often viewed as willful noncompliance within a coercive parent–child exchange (e.g., Patterson, 1982), parents may shift from initial frustration at not being able to successfully affect the child’s behavior (i.e., blocked goal attainment) to anger. The parental disciplinary response may shift, in tandem, from instrumental aggression (i.e., aggression intended to achieve an objective such as changing the child’s behavior) to hostile, emotional aggression (i.e., aggression intended to harm; see Berkowitz, 1989 and Feshbach, 1964 for more on these two classic types of aggression). Consistent with this view, physical child abuse often ensues when parents escalate the severity of an initially more benign physical discipline approach (Herrenkohl, Herrenkohl, & Egolf, 1983; Whipple & Richey, 1997), implying that putatively instrumental aggression may become hostile aggression when a parent is faced with perceived child noncompliance they view as intentional. For example, in interviews with abusive parents, 91% of parent’s first response to the child was nonabusive, but abuse arose when the parents escalated to punitive physical discipline and abuse (Kadushin & Martin, 1981). As a result, many conceptualize parent–child aggression (PCA) as occurring along a continuum (Graziano, 1994; Greenwald, Bank, Reid, & Knutson, 1997; Rodriguez, 2010; Straus, 2001a, 2001b; Whipple & Richey, 1997), wherein milder forms of physical discipline on one end can progress to physical abuse further along the continuum. Child abuse potential (Milner, 1994) predicts the likelihood that an individual will move along this continuum from milder forms of PCA toward physical abuse. Considerable attention has been devoted to the role of anger in the risk for PCA, with relatively little consideration of its precursor, frustration. Anger is presumed to intensify general tendencies for aggression (Baumeister & Bushman, 2007) and is often implicated in risk for parents’ physical abuse perpetration (Black, Heyman, & Smith Slep, 2001; Stith et al., 2009). The self-reported tendency to overtly express anger is associated with greater parental child abuse potential (Robyn & Fremouw, 1996; Rodriguez & Green, 1997; Rodriguez & Richardson, 2007). Rarely, however, is low frustration tolerance considered specifically as a risk factor for child abuse. Although frustration has been demonstrated to influence the types of cognitive processing that could lead to physical abuse perpetration (e.g., Russa, Rodriguez, & Silvia, 2014), research has yet to focus explicitly on frustration as a risk factor for child abuse. Ultimately, increased understanding of the association between frustration and PCA risk could serve to inform future intervention programming to reduce abuse risk. Only a limited number of child abuse intervention programs include anger management skills (e.g., Acton & During, 1992; Donohue, Miller, Van Hasselt, & Hersen, 1998; Sanders et al., 2004), whereas many other intervention programs emphasize relationship building, stress management, social support enhancement, and/or child behavior management (e.g., Fantuzzo, Stevenson, Kabir, & Perry, 2007; Herschell & McNeil, 2007; Swenson, Schaeffer, Henggler, & Faldowski, 2010). If low frustration tolerance is associated with risk for PCA, targeted interventions focused on parental decision-making and behavioral control at this earlier level of emotional arousal (i.e., frustration), prior to escalation to anger and physical abuse, might be useful. Methodological issues in studies of PCA risk, however, have hampered research. A few laboratory paradigms have employed behavioral simulations of aggressive behavior as a proxy for child abuse (e.g., Crouch et al., 2012; Passman & Mulhern, 1977; Vasta & Copitch, 1981), but the field continues to rely heavily on the assessment of concepts like anger via self-report. Self-report is highly susceptible to response bias and distortion, and it may result in intentional, or even unconscious, misrepresentations (Fazio & Olson, 2003). For example, respondents may be inclined to embellish their abilities to control their anger or tolerate frustration, motivated to present themselves in a socially desirable manner, and/or motivated by positive self-perception biases. Because of the shortcomings of self-report assessments of frustration, and the desire for experimental designs, analog approaches are sometimes favored in aggression research. Many analog tasks are designed to reduce response biases by assessing the target concept through implicit means. Because the respondent is not fully aware of the task’s intent or how it will be evaluated, it is less likely that responses will be consciously or unconsciously manipulated (Fazio & Olson, 2003). Analog approaches can complement and enrich the information available from self-report strategies (Nosek, 2007). In one early study, a small sample of undergraduates was placed in a frustrating teaching condition, and participants’ subsequent “aggressive” responses to a fictional child were assumed to arise from increased frustration due to aversive child behavior (Vasta & Copitch, 1981). Analog tasks like the Mirror Tracing Persistence Task (Strong et al., 2003) and the Paced Auditory Serial Addition Task (Lejuez, Kahler, & Brown, 2003) can provide an alternative to a self-report assessment of frustration, but neither of these analog tasks pertain to parenting nor have any measures of frustration tolerance been utilized to predict PCA risk. Given the desire to explore the association between frustration tolerance and PCA while recognizing the methodological limitations of self-reports, the current investigation evaluated whether an analog assessment of frustration tolerance (Frustration Intolerance Task, FIT; McElroy & Rodriguez, 2008) designed to specifically elicit parenting-related frustration would be associated with child abuse potential. Consistent with the literature that frustration arises when an individual is thwarted in achieving a goal, the FIT requires participants seek a store exit (simulated by working through an insoluble computerized maze attempting to find the exit) while listening to a crying child. Thus, in addition to arousing frustration by

C.M. Rodriguez et al. / Child Abuse & Neglect 46 (2015) 121–131

123

presenting participants with an insoluble “parenting related” task, the FIT was designed based on early research indicating that infant cries are a frustrating, aversive stimulus for abuse-risk mothers (Frodi & Lamb, 1980). Recent research confirms that abuse-risk parents report more hostile feelings and consider infant crying as more aversive relative to comparison parents (Crouch, Skowronski, Milner, & Harris, 2008). The FIT has been successfully used in past research to explore the impact of frustration in parenting situations (Russa et al., 2014). This investigation utilized a series of studies to systematically explore and extend findings regarding FIT frustration tolerance and PCA risk in six different samples. We considered a multifaceted range of risk factors reflective of the PCA continuum. Parents who evidence high child abuse potential are more likely to rely on physical tactics, including abusive tactics, in disciplinary situations (Rodriguez, 2010), and to demonstrate coercive, dysfunctional parenting styles (Haskett, Scott, & Fann, 1995; Margolin, Gordis, Medina, & Oliver, 2003). We also considered a range of participants, including samples of undergraduate students (pre-parents who could become parents at any time), a community sample of mothers, and groups of at-risk mothers (mental health sample). By using samples of varying risk and a variety of indicators of PCA risk, we could investigate whether an association between PCA risk and low frustration is evident independent of the nature of the sample or how PCA risk is measured. This investigation will be considered in two parts: (1) in the first set of studies (Part 1), the FIT with an infant crying stimulus was utilized. Respondents with lower scores (i.e., poorer frustration tolerance) were expected to evidence elevated self-reported child abuse potential, more use of physical aggression, greater dysfunctional parenting practices, and attitudes favoring physical discipline use; (2) in a second set of studies (Part 2), a potentially enhanced version of the FIT was used to replicate and extend the findings from Part 1 in a set of new samples. Part 1: Methods Measures Across Part 1 Studies The Frustration Intolerance Task (FIT; McElroy & Rodriguez, 2008) is an analog approach designed to assess a respondent’s capacity to tolerate frustration in a parent–child situation. A computer simulation presents a two-dimensional visual maze of grocery store aisles. After a practice trial of how to navigate the maze with arrow keys, the FIT instructions ask the respondents to imagine they are in a grocery store and must find their way out of the store because their child has begun crying. Although movement with the arrow keys implies progress through the store/maze, no solution is actually possible. On the screen is a large “Stop” button; participants are asked to continue searching for an exit to the store/maze until they succeed or until they wish to stop. The task terminates in five minutes if the participant does not quit. Throughout their search for an exit, they hear an auditory overlay of a crying baby. Across all studies, FIT scores measured the time, in seconds, each participant took to quit the task. Poorer frustration tolerance was operationalized as shorter duration on the FIT. The Child Abuse Potential Inventory (CAPI; Milner, 1986) is a well-known instrument designed to screen for physical child abuse risk, evaluating rigidity and interpersonal and intrapersonal qualities identified in substantiated abuse perpetrators. This instrument contains few explicit questions regarding parenting or discipline beliefs. Respondents indicate whether they agree with 160 statements, although only 77 of these statements are variably weighted to contribute to an Abuse Scale total score. The remaining items in the instrument involve experimental scales and distortion indices. Higher scores on the CAPI Abuse Scale suggest greater child abuse potential. Psychometric evidence for the CAPI confirms high internal consistency for the Abuse Scale (Milner, 1986), with split-half reliability ranging from .96 (for control groups) to .98 (for abuse samples), and Kuder–Richardson reliability coefficients ranging from .92 (for control samples) to .95 (for abuse groups). CAPI scores also demonstrate predictive validity, with a correct classification rate of 89.2% of confirmed child abusers and 99% of controls (Milner, 1994). The Adult–Adolescent Parenting Inventory-2 (AAPI-2; Bavolek & Keene, 2001) is a self-report measure of parenting and child-rearing attitudes characteristic of abusive and neglectful parenting. This measure has been conceptualized as a measure of beliefs associated with child abuse potential (Conners, Whiteside-Mansell, Deere, Ledet, & Edwards, 2006), providing an adjunct to the CAPI self-report of abuse risk that more explicitly addresses parenting. The AAPI-2 contains 40 items wherein participants rate their level of agreement using a five-point Likert scale. Scoring on the AAPI-2 was oriented such that high AAPI-2 Total scores suggest greater abuse risk (to be consistent with the CAPI direction). With regard to content validity, AAPI-2 Total scores are correlated with the Parenting Discipline Methods Interview (Baydar, Reid, & Webster-Stratton, 2003) which assesses caregiver disciplinary responses to child behaviors. A psychometric evaluation of the AAPI-2 (Conners et al., 2006) reported adequate internal consistency for the AAPI-2 Total score (˛ = .85). The Parent–Child Conflict Tactics Scale (CTS-PC; Straus, Hamby, Finkelhor, Moore, & Runyan, 1998) presents 22 behaviorally specific items in which a parent estimates the frequency with which they have implemented the behavior during parent–child conflicts in the past year, ranging from never to more than 20 times. Thirteen items directly address varying degrees of physical aggression toward children, and these items comprise the Physical Assault subscale (these include minor assault/corporal punishment, severe assault/physical maltreatment, and very severe assault/severe physical maltreatment). Moderate internal consistency is reported for the Physical Assault Scale (˛ = .55) which likely reflects the broad range of behavior assessed on the scale (from spanking to assault with a deadly weapon; Straus et al., 1998). The Parenting Scale (Arnold, O’Leary, Wolff, & Acker, 1993) is designed to identify parents’ dysfunctional disciplinary style. Using a 7-point scale, parents indicate their responses to 30 typical parent–child conflict situations, with contrasting parent reactions at each endpoint of the scale. Based on the original factor analysis (Arnold et al., 1993), overall dysfunctional

124

C.M. Rodriguez et al. / Child Abuse & Neglect 46 (2015) 121–131

disciplinary style incorporates three disciplinary response styles: Overreactivity (10 items representing a harsh, angry discipline style), Laxness (reflecting a permissive approach to parenting), and Verbosity (in which parents rely on verbal persuasion even when ineffective). The Parenting Scale yields a Total score that represents overall dysfunctional disciplinary style based on the average across items, with high scores indicative of more frequent use of dysfunctional strategies. Internal consistency reported for the Total score is moderately high at .84 (Arnold et al., 1993). Over a two-week period, test–retest reliability was relatively high for the Total scores, at .84, and scores were significantly related to clinical observations of parent–child situations (Arnold et al., 1993). The Attitudes Toward Spanking (ATS; Holden, 2001) is a 10-item measure of self-reported attitudes toward physical discipline using a 7-point Likert scale from Strongly Disagree to Strongly Agree. Participants are asked to indicate the extent to which they agree with the use of spanking in various settings. High internal reliability (ranging from .89 to .91) and test–retest reliability over three weeks (.76) have been reported (Holden, 2001). High scores on the ATS are positively correlated with parent weekly reports of spanking (r = 73) and use of physical discipline (Ateah & Durrant, 2005; Holden, 2001). Study 1: Participants and Procedures A pilot study of two analog tasks, including the FIT, was conducted with 53 undergraduate students enrolled for course credit in undergraduate education courses. This sample included 27 males and 26 females with a mean age of 22.55 years (SD = 4.41). This sample was predominantly White (83%), single (92.5%), and childless (92.5%). This study aimed to explore whether the FIT showed utility in a pre-parent population prior to administering to a parent sample. Upon arrival to the lab, participants provided informed consent and were escorted to a private room to complete measures on a computer that displayed items individually. Measures extracted for this investigation consisted of time scores for the FIT (last in the protocol) and responses to self-report questionnaires, which included the AAPI-2 and ATS. The program automatically stored responses into a database identified only by a randomly assigned identification number. Thus, participants were assured of anonymity and encouraged to respond candidly. The university Institutional Review Board granted approval for this study. Study 2: Participants and Procedures A sample of 73 mothers was recruited for a parenting study of maternal caregivers raising children ages 5–12 with diagnosed behavior problems (i.e., Oppositional Defiant Disorder, Attention Deficit Hyperactivity Disorder, Conduct Disorder). This group of mothers was presumed to be at higher abuse risk for abuse perpetration due to the challenges in disciplining and parenting children of this age range with behavior problems. In contrast to our previous pre-parent sample, use of mothers with school-age children also insured that these mothers had a significant history and range of experiences parenting in contexts that might involve noncompliance and coercive parent–child exchanges of the type that can lead to disciplinary escalation and abuse. This sample and the findings reported below were previously published (see details in McElroy & Rodriguez, 2008) and are included here for the purposes of comparison with the other samples. On average, parents were 40.51 years of age (SD = 10.53), primarily Caucasian (82.2%), living with a partner (71.6%), with an estimated annual family income of $41,016 (SD = $33,058). Of the total sample, 43.8% reported vocational training beyond high school, 11.0% had a college degree, 5.5% had some graduate education, and the remaining 23.3% reported no education beyond high school. Ethical approval was obtained from the university Institutional Review Board. Participants were recruited from community mental health agencies and a school district by distributing flyers directly to eligible participants. In an in-home session, parents entered their responses on a laptop computer with no identifying information, permitting responses to be saved anonymously. For this study, the protocol included the FIT (last in the protocol) as well as the CAPI, CTS-PC, and Parenting Scale. Families received $20 for participating in this study. Study 3: Participants and Procedures A community sample of 70 mothers, with children ages 7–12, who completed the FIT were extracted from a larger study about physical discipline focused on parenting elementary-age children. While this was a sample of experienced parents (versus pre-parents), we considered this sample to be at lower risk for abuse perpetration than the previous sample, where the children have diagnosed behavioral problems. Participants were recruited from flyers distributed at local after-school programs and churches. Mothers were on average 36.38 years (SD = 5.98). The sample was primarily Caucasian (89.6%), living with a partner (89.9%), with a median estimated annual family income of $60,000. The majority of the sample had some vocational training beyond high school (34.8%) or a college degree (39.1%). Sessions were conducted in the family home where mothers completed the protocol by entering responses anonymously onto a laptop computer. Instruments of interest for this study included the FIT (last in the protocol) and the CAPI, AAPI-2, Parenting Scale, and ATS questionnaires. Families received $20 for participating. This study was approved by the university Institutional Review Board.

C.M. Rodriguez et al. / Child Abuse & Neglect 46 (2015) 121–131

125

Table 1 FIT means, standard deviations, and correlations across studies. FIT M (SD), sec Study 1 (n = 53 undergrads) 5-min cry AAPI-2 Attitudes toward Spanking Study 2a (n = 73 at-risk mothers) 5-min cry CAPI Abuse Scale CTS-PC Physical Assault Parenting Scale Total Study 3 (n = 70 mothers) 5-min cry CAPI Abuse Scale AAPI-2 Parenting Scale Total Attitudes toward Spanking Study 4a (n = 112 undergrads) 10-min cry AAPI-2 APT Physical Total APT Escalation Total PANAS-X Frustration Study 4b (n = 45) 10-min cry Final 30-s Heart Rate Final 30-s Skin Conductance Level Final 30-s Skin Conductance Number Study 5 (n = 105 undergrads) 10-min tantrum AAPI-2 Study 6 (n = 46 at-risk mothers) 10-min tantrum AAPI-2 Child Vignettes Punishment Child Vignettes Attribution

FIT(r)

193.26 (104.87) −.30* −.27* 222.50 (92.07) −.48*** −.31* −.38*** 218.91 (90.54) −.31** −.25* −.30** −.20 428.94 (190.22) −.27** .13 −.21* −.27* 488.51 (156.47) −.36* −.15 −.12 317.41 (173.34) −.25** 224.02 (209.61) −.50*** −.29* −.33*

Note: FIT = Frustration Intolerance Task; CAPI = Child Abuse Potential Inventory; CTS-PC = Parent–Child Conflict Tactics Scale; AAPI-2 = Adult–Adolescent Parenting Inventory-2; APT = Analog Parenting Task. a From McElroy and Rodriguez (2008). * p ≤ .05. ** p ≤ .01. *** p ≤ .001.

Part 1: Results Statistical Analyses Across the samples in Studies 1–3, demographic comparisons were first conducted to assess potential gender differences using t-tests and to evaluate potential socioeconomic associations using correlational analyses. Subsequent analyses then considered whether analog task scores evidenced an association with the indicators of PCA risk across samples. Demographic Considerations Gender differences could only be considered in Study 1 (Studies 2 and 3 consisted of only mothers). In Study 1, males’ FIT scores suggest they appear less tolerant (M = 154.37, SD = 99.97) than females (M = 233.65, SD = 95.67), t(51) = 2.95, p ≤ .01. In contrast, in Study 1, comparable mean scores were attained for males and females for the self-reported abuse risk attitudes on the AAPI-2 and discipline attitudes on the ATS. With regard to respondent age, which could be evaluated in the parent samples that demonstrate appreciable variability in age (Studies 2 and 3), younger parents were somewhat less tolerant on the FIT (r = −.23, p = .05) in the community sample of parents in Study 3 but no significant association was observed in the risk sample of Study 2 (r = .06, p > .05). Younger parents also attained higher CAPI Abuse Scale scores (r = −.31, p ≤ .01) and Parenting Scale scores (r = −.30, p ≤ .01) in the at-risk sample of Study 2. Younger age was unrelated to CAPI scores or Parenting Scale scores in the community sample of mothers (p > .05). For these parent samples, annual family income was not associated with FIT scores in either Studies 2 or 3 (r = .02 and r = −.03, respectively). However, family income was significantly related to CAPI Abuse Scale scores in Study 2 (r = −.34, p ≤ .05), marginally related to CAPI Abuse Scale scores in Study 3 (r = −.19, p = .05), and significantly related to Parenting Scale Total scores (r = −29, p ≤ .01) in Study 3. Correlational Analyses As seen in Table 1, the pattern of associations across Studies 1–3 indicates that FIT scores were significantly associated with elevated child abuse potential (CAPI, AAPI-2), more frequent use of physical discipline tactics (CTS-PC Physical Assault Total), greater use of dysfunctional parenting style practices (Parenting Scale Total), and to some extent, greater endorsement

126

C.M. Rodriguez et al. / Child Abuse & Neglect 46 (2015) 121–131

of physical punishment (ATS). In the at-risk sample of Study 2, controlling for age and income did not substantively alter the magnitude of the obtained associations with FIT scores reported in Table 1. The undergraduate pre-parents in Study 1, who are younger than the parent samples, also appear significantly less tolerant than the parents, terminating the maze more rapidly (even in comparison to the lower-risk community sample, t(68) = 2.35, p ≤ .05200). These three studies indicate that poor frustration tolerance on the FIT is consistently associated with a variety of indices of elevated PCA risk across pre-parent, low-risk parent, and high-risk parent samples. Part 2: Introduction Part 2 was designed to replicate and extend findings from Part 1 using a modified FIT. Specifically, the maximum time limit for the FIT was extended from five to ten minutes, thereby reducing ceiling effects for some of the participants in Part 1 (i.e., respondents reaching the time limit without quitting, limiting variability in FIT scores). An increased variability in FIT scores would improve the ability to statistically assess associations between the FIT and PCA risk. Additionally, in order to extend Part 1 findings, the association between FIT scores and physiological arousal was examined. Frustration has been associated with high physiological arousal (e.g., Lobbestael, Artnz, & Wiers, 2008), which in turn can be experienced as aversive (Anderson & Huesmann, 2003). Although the literature is sometimes equivocal (e.g., McCanne & Milner, 1991), high arousal in terms of cardiac reactivity is generally associated with aggressive behavior presumably because aversive cues evoke frustration that can then prompt aggressive responding (Patrick & Verona, 2007). With regard to physical child abuse risk, increased psychophysiological reactivity has been implicated as a risk factor (Black et al., 2001; Milner & Dopke, 1997), with one meta-analysis combining anger and hyper-reactivity indicating a large effect size for physical abuse (Stith et al., 2009). Increased arousal among abuse-risk mothers is apparent for both child-related and non-child related stressors (Casanova, Domanic, McCanne, & Milner, 1992). Early work has indicated that abusive mothers are also more physiologically responsive to infant cries, which they consider aversive (Frodi & Lamb, 1980). Finally, although infant crying may be experienced as aversive and potentially frustrating, respondents might attribute crying or infant distress to benign intent (e.g., hunger or physical discomfort). Given that the attribution of blame in combination with frustration theoretically provokes anger (Kalat & Shiota, 2007; Power & Dalgleish, 2008), a frustration stimulus that increases perception of blame would be expected to lead to increased physical abuse risk. Therefore, in another FIT modification (Studies 5 and 6), the infant cry in the FIT was replaced with a toddler temper tantrum wherein the toddler refuses to accept a time-out, representing a more clearly oppositional child which could elicit negative attributions. In summary, in Part 2 of this investigation, a 10-minute FIT was utilized in which lower frustration tolerance was expected to relate to elevated self-reported abuse risk and increased willingness to use PCA as well as heightened physiological reactivity. The final two samples utilized the 10-minute temper tantrum adaptation in which the stimulus more clearly evokes blame to confirm that FIT scores remain related to self-reported child abuse risk markers. Part 2: Methods Measures In addition to the FIT and the AAPI-2 (described previously), the Analog Parenting Task (APT; Russa & Rodriguez, 2010; Zaidi, Knutson, & Mehm, 1989) was used in Study 4. The APT contains 26 images of child behavior (e.g., dangerous, destructive, rule violation, control), requesting the participants’ initial discipline reaction to the behavior as well as their response should the child persist in the behavior. Responses on this instrument yield scores for their selection of physical punishment (APT Physical Total) and escalation of discipline to more extreme physical discipline responding in dealing with continued child noncompliance (APT Escalation Total). Internal consistency estimates for these two scales are strong (APT Physical, ˛ = .91–.93; APT Escalation, ˛ = .84; Russa & Rodriguez, 2010). With respect to content validity, APT scores modestly correlate with the CAPI and AAPI-2 (Russa & Rodriguez, 2010). In addition, the Positive and Negative Affect Schedule-Expanded Form (PANAS-X; Watson & Clark, 1994) was administered in Study 4. The PANAS-X is a self-assessment of emotional states in which participants are asked to report on positive or negative affect. In Study 4, participants reported on their positive and negative affect immediately after the FIT on a scale of 1 (very slightly or not at all) to 5 (extremely), with a particular focus in this investigation on their response to their reported experience of frustration specifically post-FIT. The Beck Depression Inventory (Beck, Steer, & Brown, 1996), a frequently used self-report inventory of depression, was administered in Study 6 and was included in the analyses as an independent predictor of child abuse risk, given prior research implicating depression as a physical child abuse risk factor (Black et al., 2001). The BDI present 21 symptoms in which the participant indicates which of four statements best describes them, with higher scores indicative of greater depressed mood. Strong internal consistency and validity have been previously reported for the BDI (Beck et al., 1996). Finally, in Study 6, parents were also asked to respond to Child Vignettes (CV; Plotkin, 1983), which assess parental attributions and expected punishment for child behavior following eighteen short vignettes. Parents are asked to imagine the child in the vignette is their own and rate the perceived intentionality and the degree of punishment they would employ following each situation. Attributions and punishment expectations are rated on 9 point scales, ranging from “My child did not mean to annoy me at all” to “The only reason my child did this was to annoy me” and “I would not punish my child

C.M. Rodriguez et al. / Child Abuse & Neglect 46 (2015) 121–131

127

at all” to “I would punish my child a great deal,” such that higher scores indicate greater negative attributions and greater intention to punish. Abusive parents perceive greater child intentionality and are willing to implement more punishment than comparison parents (Haskett, Scott, Willoughby, Ahern, & Nears, 2006). The CV subscales also demonstrate good internal consistency (˛ = .83) for both scores (Haskett et al., 2006). Study 4: Participants and Procedures This study involved a sample of 112 pre-parent undergraduates participating for course credit at two universities with a mean age of 21.30 years (SD = 6.06), with 83.7% female. This sample was predominantly White (81.7%), single (84.3%), and childless (89.4%). Participants at the first university site were administered the 10-min crying baby FIT, immediately followed by the PANAS-X, as well as the AAPI-2 and APT, all anonymously completed by computer. At the second site (n = 45), the protocol was identical and conducted entirely while the participant was monitored physiologically. The Institutional Review Boards of the two universities granted approval for the study. A multi-channel BioPac MP physiological data recorder (BIOPAC Systems Inc., Goleta, CA) was used to track participants’ skin conductance and heart rate continuously at 60 Hz. Skin conductance was recorded from adhesive, pre-gelled disposable Ag–AgCl electrodes attached to two fingers on the non-preferred hand. Vasomotor activity was recorded from a photoplethysmograph attached with a Velcro strap to the tip of another finger on the same hand. The computer derived heart rate from the intervals between successive systolic points of the analog vasomotor signal. To minimize artifacts in the recordings, the heel of the hand rested on the end of the arm of a chair or table, with the fingers dangling over the edge, and the sensors on the fingers did not come in contact with any other object. In addition, participants were cautioned to avoid movement. Minor artifacts were edited from the recordings prior to data analysis. Reactivity for skin conductance was measured in terms of mean skin conductance level and the number of skin conductance responses that exceed 0.02 ␮S conductance. Reactivity for heart rate was measured as the change in mean and maximum heart rate expressed as a proportion of mean baseline. Prior to beginning physiological recording, the participant was led to a sink to wash their hands with soap and warm water. They were then led to a private room, and the sensors were described and attached to fingers on their non-preferred hand. Headphones were provided and the participant was left alone in the room during the administration of the 10-min crying baby FIT. Physiological data were recorded continuously during the task, and the final 30-sec of data prior to their termination of the FIT maze were analyzed. Study 5: Participants and Procedures This study provided a preliminary examination of the 10-min temper tantrum version of the FIT. A total of 105 psychology, pre-parent, undergraduates participated for course credit in a larger study piloting a number of analog tasks for childless students ages 18–20. The sample was 75.2% female and predominantly Caucasian (70.5%). All responses were individually and anonymously completed on a computer in a private room in the lab. Participants provided responses to the 10-minute tantrum FIT (toward the end of the protocol) and the AAPI-2. The university Institutional Review Board approved the study. Study 6: Participants and Procedures This final sample includes 46 mothers (age M = 37.07, SD = 5.80 years) from an ongoing study of at-risk parenting of children ages 7–12. Mothers were recruited from community mental health agencies wherein either the mother (27.9%), the child (46.5%), or both (25.6%) were receiving mental health services (children diagnosed with disruptive behavior disorders, comparable to those in Study 2). Like Study 2, this group was considered to be at risk given the greater challenges of parenting under those conditions. This diverse sample included 51.2% who identified as African-American (7.3% Other, 41.5% Caucasian), with 10.3% of the total sample also identifying as Hispanic/Latina. The majority of this low-income sample were single parents (69%) with a reported average annual family income of $13,000 or less. The majority of the sample indicated they had at least some college or vocational training. Similar to Studies 2 and 3, mothers participate in an in-home session by anonymously completing measures administered on a laptop computer, in which responses to the FIT (last in the protocol), BDI, AAPI-2, and Child Vignettes were extracted for this investigation. Mothers received $25 for participating in this study. The study was approved by the university Institutional Review Board. Part 2: Results Statistical Analyses Across samples in Studies 4–6, demographic comparisons were first performed in order to assess potential gender differences using t-tests and potential socioeconomic associations using correlational analyses. Analyses then considered whether FIT analog task scores were associated with the markers of PCA risk across samples, with additional attention to controlling for personal mental health issues among mothers in Study 6.

128

C.M. Rodriguez et al. / Child Abuse & Neglect 46 (2015) 121–131

Demographic Considerations With regard to gender, which can be considered in Studies 4 and 5, males attained comparable scores on the FIT (p > .05) but significantly lower scores on the AAPI-2 in Study 4, t(111) = 2.66, p ≤ .01 and marginally lower scores on the AAPI-2 in Study 5, t(103) = 1.82, p = .07. No gender differences were observed in Study 4 on the APT (p > .05). With regard to age, which could be considered in Study 6, younger parents expected to punish children more on the CV-Punishment scale (r = −.31, p ≤ .05) and received higher AAPI-2 scores (r = -.31, p = .06), but age was unrelated to FIT scores. Correlational Findings As seen in Table 1, lower FIT scores on this longer FIT version continue to be associated with greater self-reported abuse risk on the AAPI-2 across studies, replicating the findings from Part 1. This pattern was particularly apparent in the at-risk sample in Study 6. Furthermore, pre-parents with lower FIT scores showed an expected pattern of greater self-reported frustration immediately following the FIT, and these lower FIT scores were significantly associated with an increased propensity to escalate to physical discipline (APT Escalation) when faced with child misbehavior in Study 4. FIT scores were significantly correlated with increased heart rate, but not skin conductance, in the physiological portion of Study 4. At-risk mothers in Study 6 who had lower FIT scores were also significantly more likely to attribute negative intent to children and intended to punish perceived child misbehavior, as reported in the vignettes. Using Study 6, multiple regression analyses were conducted to predict AAPI-2 scores of child abuse risk, controlling first for demographics (age, income, race, education level) and then controlling for maternal depressive BDI symptoms, to determine if FIT scores uniquely predicted AAPI-2 scores beyond these covariates. Results indicate that the overall model was significant, F(6, 40) = 7.61, p ≤ .001, R2 = .53. Specifically, all demographics and BDI scores explained 49% of the variance in AAPI-2 scores, with FIT scores accounting for an additional 5.3% of the variance in AAPI-2 scores, t = 1.96, p ≤ .05. It is notable that the pre-parent undergraduates receiving the temper tantrum adaptation in Study 5 were significantly less tolerant than their pre-parent peers in Study 4, where the stimulus was a crying baby, t(104) = 6.59, p ≤ .001. When parent and pre-parent reactions were compared, mothers in the at-risk sample in Study 6 averaged even lower FIT scores than pre-parent undergraduates in Study 5, t(104) = 4.98, p < .001. These findings suggest a potential impact of parenting status (pre-parent versus parent) on FIT (frustration tolerance) scores. Additionally, the pattern suggests that respondents’ attributions of blame in the cry versus temper tantrum version of the FIT led to lower levels of frustration tolerance, a finding not inconsistent with research suggesting that frustrating episodes in which blame can be attributed may be more likely to lead to anger and aggressive responding. In summary, Part 2 replicates key findings from Part 1, and extends these findings to demonstrate expected patterns of association between FIT scores and disciplinary choices (including escalation), physiological responding, patterns of attributions, and punishment intentions. Concluding Discussion Prior theory has proposed that frustration perceived as aversive and intentional serves as a precursor for anger, which may lead to aggression, particularly when conditions of blame can be attributed in the situation (Power & Dalgleish, 2008). The present investigation explored whether frustration tolerance was associated with various markers of risk for PCA. In order to avoid the limitations of a self-report design, frustration tolerance was assessed using an analog measure (Frustration Intolerance Task, FIT) specifically designed to elicit parenting-related frustration. Findings across six studies utilizing a range of indices of risk for physical abuse and samples of varying risk levels (pre-parents, community parents, at-risk parents) yielded consistent patterns whereby low frustration tolerance was associated with markers of risk for physical abuse. Specifically, FIT scores were associated with: increased self-reported frustration, increased child abuse potential (as assessed by questionnaires), greater use of parent–child aggression in discipline encounters, dysfunctional disciplinary style, support for physical discipline use and disciplinary escalation, increased heart rate responding, more negative child attributions, and increased punishment intentions. Although more research is needed, these data are consistent with the notion that low frustration tolerance may lead to increased risk for PCA. Although frustration has long been a focus of the aggression literature (Berkowitz, 1989; Dollard et al., 1939), exploration of the potential role of frustration and frustration tolerance in PCA has been nearly absent. These findings extend previous findings that self-reported propensity for anger (e.g., Robyn & Fremouw, 1996; Rodriguez & Green, 1997; Rodriguez & Richardson, 2007), a potential outcome of frustration, is associated with increased risk for child abuse. Unfortunately, much of the anger research has relied on individuals to accurately convey the extent of their own anger expression and control. For example, although anger and hyper-reactivity have been identified as demonstrating a strong effect on child abuse risk in previous reviews (e.g., Stith et al., 2009), much of the referenced literature is based on self-reported anger. The current findings underscore the potential value of using analog methods to explore the role of frustration or anger in risk for PCA. The present study explored the link between frustration and PCA risk using an implicit frustration task for which participant scores were unlikely to reflect impression management, social desirability, positive self-perception, or other biases that may influence self-report results. The FIT was designed to elicit frustration via inclusion of the elements identified in research literature to lead to frustration. The insoluble maze placed participants in a task in which goal attainment (finding the exit) was blocked, and simultaneously, participants were exposed to parenting-relevant auditory stimuli (crying baby or

C.M. Rodriguez et al. / Child Abuse & Neglect 46 (2015) 121–131

129

tantruming toddler) noted in the literature to be cues for frustration. As expected, FIT frustration tolerance was associated with self-reported frustration immediately following the task (Study 4a). In addition to its association with a wide range of PCA risk markers across samples of varying risk levels and parenting backgrounds, frustration tolerance as measured by the FIT contributed unique variance in explaining both child abuse potential and dysfunctional parenting style in at risk mothers (Study 2, McElroy & Rodriguez, 2008). The replication of this pattern in a second sample of at-risk women, even while controlling for critical socio-demographic and mental health risk factors, further increased confidence in these findings. Taken together, these data lend support to the premise that frustration tolerance could be an important area for further inquiry in understanding PCA. This study leaves several lingering questions. In the present study, data for some samples were collected in the home, while data for other samples were collected in the laboratory. The effect size for the association of frustration tolerance with the index of risk for abuse (AAPI-2) in the community sample of mothers (administered at home) was comparable to that of the pre-parent undergraduates (in the lab). However, we are unable to gauge directly whether scores are affected by setting administration, a potential area for future research consideration. The logistics of data collection in the home also precluded the collection of physiological data for parents; further work considering physiological correlates of frustration tolerance in community and at-risk mothers would be particularly intriguing. And while our physiological findings indicated expected associations between heart rate and frustration tolerance, we found no effect for skin conductance. These data are consistent with an emphasis on cardiac reactivity with aggression (Patrick & Verona, 2007), and prior significant associations between overreactive discipline and heart rate reactivity but no association between overreactive discipline and electrodermal responding (Lorber & O’Leary, 2005). Future physiological investigations would be helpful in further verifying the replicability and possible source of this pattern. Evaluation of how frustration tolerance operates in more socioeconomically and racially/diverse parent samples is warranted, particularly utilizing samples of varied levels of PCA risk. For instance, note that the effect sizes between frustration tolerance scores and PCA risk were largest for the samples of mothers with greater risk because they are dealing with mental health issues (e.g., for the AAPI-2, moderate effect size in magnitude, comparable for both Studies 2 and 6 despite the potential demographic differences between these two samples). In contrast, the lower risk pre-parent undergraduate students and community sample of parents demonstrated mild effect sizes for this association. If this trend is consistent, explorations of the association between frustration tolerance and markers of PCA in samples substantiated for physical abuse may yield stronger effects. Continued replication of other moderate effect sizes in other at-risk samples may also be informative for targeted intervention efforts. Age and gender may also influence these patterns in ways that we had limited ability to explore in the present study. With our focus on community and high risk mothers, in light of expected gender differences in aggression (Patrick & Verona, 2007), research investigating fathers’ frustration tolerance is sorely needed. Additionally, in our findings, maternal age was associated with many PCA markers but was not associated with frustration tolerance scores in the two at-risk samples. While the source of these differences is unclear, given that age was only mildly related to frustration tolerance scores in the community sample of mothers, it is possible that these rather modest effects of age were simply over-ridden by other, more prominent high risk features in the higher risk samples. Interestingly, the younger, pre-parent samples appeared to be significantly less tolerant on the FIT than the parent samples, but it is hard to know whether this difference is attributable to the age of the pre-parents, or a lack of parenting experience. In addition, parents were raising elementary age children in all parent samples, and it would be intriguing to investigate whether the obtained results are similar with parents of children of toddlers/preschoolers, closer in age to the child represented in the FIT auditory stimuli. Although the FIT task was purposely designed to be directly relevant to parenting, consideration of the extent to which scores reflect more non-specific levels of frustration tolerance could be informative (e.g., self-report measures are scarce, but the Mirror Tracing Persistence Task or the Paced Auditory Serial Addition Task may be analog options to which one could compare). The association between frustration tolerance as measured by the FIT and other related qualities such as task persistence or compliance also remains unclear; it would be fruitful to further explore the likely complex interactive processes that lead respondents to tolerate or persist both on this task and when faced with aversive child behavior. Additionally, given the more extensive consideration of anger in the extant literature on parent–child aggression, a consideration of the association between FIT scores and parents’ experience and/or expression of anger may also be worthwhile. Because the potential markers for abuse risk in this investigation were largely based on self-report, a particularly intriguing direction for research would be to examine the association of FIT frustration tolerance scores with analogs of PCA or observations of parenting behavior, although the latter would be susceptible to reactivity. Continued efforts to develop behavioral simulations and analog tasks along these lines would enhance our confidence in findings with regard to prediction of abuse risk, a field replete with doubts arising from concerns over response bias. Finally, given the apparent amplification in responding in some samples as a consequence of the change in auditory stimuli (cry to tantrum), closer examination of how frustration interacts with perceived intentionality and other cognitive appraisals could uncover critical interactive influences. For example, the potential for blame to interact with frustration to lead to anger and ultimately exacerbate abuse risk would be consistent with theory (Power & Dalgleish, 2008). Although some abuse intervention programs noted earlier include anger management (Donohue et al., 1998; Sanders et al., 2004), many others do not directly target improving parents’ frustration tolerance. A stronger focus on frustration tolerance could be valuable to intervention and prevention programs alike to avert anger. Strategies to enhance desensitization to aversive stimuli, relaxation training, refocusing or distraction in the face of frustration, and attribution and cognitive

130

C.M. Rodriguez et al. / Child Abuse & Neglect 46 (2015) 121–131

retraining to mitigate the experience of frustration that can occur in parent–child interactions, may be useful components for programming. Continued identification of dispositional qualities such as frustration tolerance may facilitate a more comprehensive approach to intervention given the complexity of factors that likely ultimately converge to evoke parent–child aggression.

References Acton, R. G., & During, S. M. (1992). Preliminary results of aggression management training for aggressive parents. Journal of Interpersonal Violence, 7, 410–417. Anderson, C. A., & Bushman, B. J. (2002). Human aggression. Annual Review of Psychology, 53, 27–51. Anderson, C. A., & Huesmann, L. R. (2003). Human aggression: A social-cognitive view. In M. A. Hogg, & J. Cooper (Eds.), The Sage handbook of social psychology (pp. 296–323). Thousand Oaks: Sage. Arnold, D. S., O’Leary, S. G., Wolff, L. S., & Acker, M. M. (1993). The Parenting Scale: A measure of dysfunctional parenting in discipline situations. Psychological Assessment, 5, 137–144. Ateah, C. A., & Durrant, J. E. (2005). Maternal use of physical punishment in response to child misbehavior: Implications for child abuse prevention. Child Abuse & Neglect, 29, 169–185. Bavolek, S. J., & Keene, R. G. (2001). Adult–Adolescent Parenting Inventory (AAPI-2): Administration and development handbook. Park City, UT: Family Development Resources, Inc. Baumeister, R. F., & Bushman, B. J. (2007). Angry emotions and aggressive behaviors. In G. Steffgen, & M. Gollwitzer (Eds.), Emotions and aggressive behavior (pp. 61–75). Ashland, OH: Hogrefe & Huber. Baydar, N., Reid, M. J., & Webster-Stratton, C. (2003). The role of mental health factors and program engagement in the effectiveness of a preventive parenting program for Head Start mothers. Child Development, 74, 1433–1453. Beck, A. T., Steer, R. A., & Brown, G. K. (1996). Beck Depression Inventory – Second edition manual. San Antonio: The Psychological Corporation. Berkowitz, L. (1989). Frustration–aggression hypothesis: Examination and reformulation. Psychological Bulletin, 106, 59–73. Berkowitz, L. (2012). A cognitive-neoassociation theory of aggression. In P. A. M. Van Lange, A. W. Kruglanski, & E. T. Higgins (Eds.), Handbook of theories of social psychology (pp. 99–117). Thousand Oaks: Sage. Black, D. A., Heyman, R. E., & Smith Slep, A. M. (2001). Risk factors for child physical abuse. Aggression and Violent Behavior, 6, 121–188. Casanova, G. M., Domanic, J., McCanne, T. R., & Milner, J. S. (1992). Physiological responses to non-child-related stressors in mothers at risk for child abuse. Child Abuse & Neglect, 16, 31–44. Conners, N. A., Whiteside-Mansell, L., Deere, D., Ledet, T., & Edwards, M. C. (2006). Measuring the potential for maltreatment: The reliability and validity of the Adult Adolescent Parenting Inventory-2. Child Abuse & Neglect, 30, 39–53. Crouch, J. L., Irwin, L. M., Wells, B. M., Shelton, C. R., Skowronski, J. J., & Milner, J. S. (2012). The Word Game: An innovative strategy for assessing implicit processes in parents at risk for child physical abuse. Child Abuse & Neglect, 36, 498–509. Crouch, J. L., Skowronski, J. J., Milner, J. S., & Harris, B. (2008). Parental responses to infant crying: The influence of child physical abuse risk and hostile priming. Child Abuse & Neglect, 32, 702–710. Dollard, J., Doob, L., Miller, N., Mowrer, O., & Sears, R. (1939). Frustration and aggression. New Haven, CT: Yale University Press. Donohue, B., Miller, E. R., Van Hasselt, V. B., & Hersen, M. (1998). An ecobehavioral approach to child maltreatment. In V. B. Van Hasselt, & M. Hersen (Eds.), Handbook of psychological treatment protocols for children and adolescents (pp. 279–356). Mahway, NJ: Lawrence Erlbaum Associates. Fantuzzo, J., Stevenson, H., Kabir, S. A., & Perry, M. A. (2007). An investigation of a community-based intervention for socially isolated parents with a history of child maltreatment. Journal of Family Violence, 22, 81–89. Fazio, R. H., & Olson, M. A. (2003). Implicit measures in social cognition research: Their meaning and use. Annual Review of Psychology, 54, 297–327. Feshbach, S. (1964). The function of aggression and the regulation of aggressive drive. Psychological Review, 71, 257–272. Frodi, A. M., & Lamb, M. E. (1980). Child abusers’ responses to infant smiles and cries. Child Development, 51, 238–241. Graziano, A. M. (1994). Why we should study subabusive violence against children. Journal of Interpersonal Violence, 9, 412–419. Greenwald, R. L., Bank, L., Reid, J. B., & Knutson, J. F. (1997). A discipline-mediated model of excessively punitive parenting. Aggressive Behavior, 23, 259–280. Haskett, M. E., Scott, S. S., & Fann, K. D. (1995). Child Abuse Potential Inventory and parenting behavior: Relationships with high-risk correlates. Child Abuse & Neglect, 19, 1483–1495. Haskett, M. E., Scott, S. S., Willoughby, M., Ahern, L., & Nears, K. (2006). The Parent Opinion Questionnaire and child vignettes for use with abusive parents: Assessment of psychometric properties. Journal of Family Violence, 21, 137–151. Herrenkohl, R. C., Herrenkohl, E. C., & Egolf, B. P. (1983). Circumstances surrounding the occurrence of child maltreatment. Journal of Consulting and Clinical Psychology, 51, 424–431. Herschell, A. D., & McNeil, C. B. (2007). Parent–child interaction therapy with physically abusive families. In J. M. Briesmeister, & C. E. Schaefer (Eds.), Handbook of parent training: Helping parents prevent and solve problem behaviors (pp. 234–267). Hoboken, NJ: Wiley & Sons. Holden, G. W. (2001). Attitude toward spanking (ATS). In J. B. Touliatos, R. Perlmutter, & G. W. Holden (Eds.), Handbook of family measurement techniques (Vol. 2) Abstract (p. 209). Thousand Oaks, CA: Sage. Kadushin, A., & Martin, J. A. (1981). Interview study of abuse–event interaction. In A. Kadushin (Ed.), Child abuse: An interactional event (pp. 141–224). New York, NY: Columbia University Press. Kalat, J. W., & Shiota, M. N. (2007). Emotion. Belmont, CA: Wadsworth. Lejuez, C. W., Kahler, C. W., & Brown, R. A. (2003). A modified computer version of the Paced Auditory Serial Addition Task (PASAT) as a laboratory-based stressor. The Behavior Therapist, 26, 290–293. Lobbestael, J., Artnz, A., & Wiers, R. W. (2008). How to push someone’s buttons. A comparison of four anger-induction methods. Cognition and Emotion, 22, 353–373. Lorber, M. F., & O’Leary, S. G. (2005). Mediated paths to overreactive discipline: Mothers’ experienced emotion, appraisals, and physiological responses. Journal of Consulting and Clinical Psychology, 73, 972–981. Margolin, G., Gordis, E. B., Medina, A. M., & Oliver, P. H. (2003). The co-occurrence of husband-to-wife aggression, family-of-origin aggression, and child abuse potential in a community sample. Journal of Interpersonal Violence, 18, 413–440. McCanne, T. R., & Milner, J. S. (1991). Physiological reactivity of physically abusive and at-risk subjects to child-related stimuli. In J. S. Milner (Ed.), Neuropsychology of aggression (pp. 147–166). Boston: Kluwer Academic. McElroy, E. M., & Rodriguez, C. M. (2008). Mothers of children with externalizing behavior problems: Cognitive risk factors for abuse potential and discipline style and practices. Child Abuse & Neglect, 32, 774–784. Milner, J. S. (1986). The Child Abuse Potential Inventory: Manual (2nd ed.). Webster, NC: Psyctec. Milner, J. S. (1994). Assessing physical child abuse risk: The Child Abuse Potential Inventory. Clinical Psychology Review, 14, 547–583. Milner, J. S., & Dopke, C. (1997). Child physical abuse: Review of offender characteristics. In D. A. Wolfe, R. J. McMahon, & R. D. Peters (Eds.), Child abuse: New directions in prevention and treatment across the lifespan (pp. 27–54). Thousand Oaks, CA: Sage. Nosek, B. A. (2007). Understanding the individual implicitly and explicitly. International Journal of Psychology, 42, 184–188. Passman, R. H., & Mulhern, R. K. (1977). Maternal punitiveness as affected by situational stress: An experimental analogue of child abuse. Journal of Abnormal Psychology, 86, 565–569.

C.M. Rodriguez et al. / Child Abuse & Neglect 46 (2015) 121–131

131

Patrick, C. J., & Verona, E. (2007). The psychophysiology of aggression: Autonomic, electrocortical and neuro-imaging findings. In D. J. Flannery, A. T. Vazsonyi, & I. D. Waldman (Eds.), The Cambridge handbook of violent behavior and aggression (pp. 111–150). New York: Cambridge University Press. Patterson, G. R. (1982). Coercive family process. Eugene, OR: Castalia. Plotkin, R. (1983). Cognitive mediation in disciplinary actions among mothers who have abused or neglected their children: Dispositional and environmental factors Unpublished doctoral dissertation. University of Rochester. Power, M., & Dalgleish, T. (2008). Cognition and emotion: From order to disorder. New York: Psychology Press. Robyn, S., & Fremouw, W. J. (1996). Cognitive and affective styles of parents who physically abuse their children. American Journal of Forensic Psychology, 14, 63–79. Rodriguez, C. M. (2010). Parent–child aggression: Association with child abuse potential and parenting styles. Violence and Victims, 25, 728–741. Rodriguez, C. M., & Green, A. J. (1997). Parenting stress and anger expression as predictors of child abuse potential. Child Abuse & Neglect, 21, 367–377. Rodriguez, C. M., & Richardson, M. J. (2007). Stress and anger as contextual factors and pre-existing cognitive schemas: Predicting parental child maltreatment risk. Child Maltreatment, 12, 325–337. Russa, M. B., & Rodriguez, C. M. (2010). Physical discipline, escalation and child abuse potential: Psychometric evidence for the Analog Parenting Task. Aggressive Behavior, 36, 251–260. Russa, M. B., Rodriguez, C. M., & Silvia, P. J. (2014). Frustration influences impact of history and disciplinary attitudes on physical discipline decision making. Aggressive Behavior, 40, 1–11. Sanders, M. R., Pidgeon, A. M., Gravestock, F., Connors, M. D., Brown, S., & Young, R. W. (2004). Does parental attributional retraining and anger management enhance the effects of the Triple-P Positive Parenting Program with parents at risk of child maltreatment? Behavior Therapy, 35, 513–535. Stith, S. M., Liu, T., Davies, L. C., Boykin, E. L., Alder, M. C., Harris, J. M., Som, A., McPherson, M., & Dees, J. (2009). Risk factors in child maltreatment: A meta-analytic review of the literature. Aggression and Violent Behavior, 14, 13–29. Straus, M. A. (2001a). Beating the devil out of them: Corporal punishment in American families and its effects on children. New Brunswick, NJ: Transaction. Straus, M. A. (2001b). New evidence for the benefits of never spanking. Society, 38, 52–60. Straus, M. A., Hamby, S. L., Finkelhor, D., Moore, D. W., & Runyan, D. (1998). Identification of child maltreatment with the Parent–Child Conflict Tactics Scales: Development and psychometric data for a national sample of American parents. Child Abuse & Neglect, 22, 249–270. Strong, D. R., Lejuez, C. W., Daughters, S., Marinello, M., Kahler, C. W., & Brown, R. A. (2003). The computerized mirror tracing task, Version 1. Unpublished manual. Swenson, C. C., Schaeffer, C. M., Henggler, S. W., & Faldowski, R. (2010). Multisystemic therapy for child abuse and neglect: A randomized effectiveness trial. Journal of Family Psychology, 24, 497–507. Vasta, R., & Copitch, P. (1981). Simulating conditions of child abuse in the laboratory. Child Development, 52, 164–170. Watson, D., & Clark, L. A. (1994). The PANAS-X: The manual for the Positive and Negative Affect Schedule-Expanded Form. Ames, IA: Department of Psychology, University of Iowa. Whipple, E. E., & Richey, C. A. (1997). Crossing the line from physical discipline to child abuse: How much is too much? Child Abuse & Neglect, 5, 431–444. Wranik, T., & Scherer, K. R. (2010). Why do I get angry? A componential appraisal approach. In M. Potegal, G. Stemmler, & C. Spielberger (Eds.), International handbook of anger: Constituent and concomitant biological, psychological, and social processes (pp. 243–266). New York: Springer. Zaidi, L. Y., Knutson, J. F., & Mehm, J. G. (1989). Transgenerational patterns of abusive parenting: Analog and clinical tests. Aggressive Behavior, 15, 137–152.