Intimate partner violence (IPV) is a problem in society, but even with the abundance of research conducted in the field to better understand the behaviors and consequences involved in these types of relationships, inconsistency is a reoccurring issue. According to Hamby (2014), these issues will remain unresolved without systematic and technological advancements in the way researchers conceptualize and evaluate IPV. This paper aimed to evaluate additional studies to see whether Hamby’s critique could be expanded to studies other than those she mentioned in her review. Conclusively, and in agreement with Hamby, it appears the biggest issue involves misconceptualizing IPV through the use of the Conflict Tactics Scale (CTS), a scale that perhaps is applied much too broadly. Indeed, without more thorough, generalizable studies and an accurate definition of what exact violent behaviors are unique to IPV and not other crime-ridden, violent epidemics, this field will remain stagnant. The additional studies under review provide support of this.
Intimate Partner Violence: Strengths and Weaknesses of Past Studies, Further Critiqued
Intimate partner violence (IPV) is defined by the presence of physical and mental abuse between partners in an intimate relationship; these relationships are prevalent in today’s society and often yield negative mental effects. According to a wide range of studies evaluated in Filson, Ulloa, Runfola, and Hokoda’s study (2010), the consequences of IPV are much more serious in women, and sexual violence is a large-scale issue. Scientific field studies that focus on IPV have made efforts to describe these types of relationships and the factors that contribute to the behaviors involved, to further develop an understanding of intimate partner relational dynamics, and in turn, move toward preventative measures for these violent behaviors.
Perhaps due to the ethical restrictions, in order to examine these types of relationships, most of the research in this field consists of correlational data. Of the five studies evaluated in this review, all of them are correlational studies in which data was typically collected through self-report measurements. According to Hamby (2014), self-report measures are habitual to this field and little has changed in the way IPV has been observed (p. 150). Additionally, studies have typically been done on homogenous, convenience samples. A majority of the participants in each study were Caucasian, ranging from 45.7-100% of the participants in each study; four of the studies’ participants were recruited from subgroups. Only two of the five studies reviewed in this paper consisted of college students, and most of the five studies included participant volunteers. A majority was comprised of either male or female participants, and only two of the studies included both. In addition, sample sizes were relatively small to medium, ranging from 68-451 among the five studies.
Though there is a plethora of studies that examine IPV relationships, it appears that data findings are relatively mixed and most research focuses on a variety of different aspects and their link to IPV. In other words, overall, study designs are similar in nature, but the variables and hypotheses evaluated are more varied, which perhaps explains for the inconsistencies in the field. The biggest finding up for debate seems to be the idea that IPV is bidirectional, that is, there is gender parity in the perpetration and victimization of violent behaviors in these types of relationships. This inconsistency may be due, in part, to the innumerable ways in which researchers have chosen to examine and operationalize IPV and conduct their study.
Hamby recognizes these inconsistencies, but aims to explain them through her critique of prior literature, evaluating anywhere from the measurements used to the systematic processes and biases in getting published. Her findings focus mainly on the systemic and technological “drags” in research that may account for the overstatement in gender parity. This particular review attempts to expand on her analysis by focusing, in more detail, on the individual studies themselves, which Hamby neglects to do. Review of the five additional studies, when compared with Hamby’s, reveal that perhaps it is a matter of technological “drags” that are keeping the field stagnant, but also a sheer miss by most researchers to evaluate the variables under study in more extensive detail—and most limiting, researchers may be misconceptualizing IPV all together. A more in depth and meticulously studied understanding of the variables under examination would better explain the complexity of IPV and the inconsistencies found within research. In turn, future research might meet the innovative expectations of Hamby, improve generalizability, which is otherwise lacking, and result in better preventive efforts that would best fit the needs of those involved, and eventually reduce rates of IPV.
Conceptual Issues
According to Hamby, a majority of studies fall victim to misconceptualization due to “hyperspecialization.” This is the idea that behaviors of IPV, most specifically noted by Hamby, violent behaviors, can be isolated and studied (Hamby, 2014, p. 151). Her biggest critique of this is that most researchers will use outdated scales, like the Conflict Tactics Scale (CTS), that have not been advanced or undergone maintenance since they first surfaced. As a result, researchers may be measuring similar behaviors like bullying, community violence, or family violence that score similarly to IPV behaviors on self-report scales. Hamby considers a wide range of violent acts in the form of sexual assault, aggravated and simple assault, robbery, homicide, and intimate partner homicide in evidence against gender symmetric patterns of IPV. She goes on to conclude that individual rates of these violent acts are most perpetrated by men and that these findings should not be ignored when assessing patterns of IPV across gender. She suggests rather than using default measures, a progressive approach would be to look at gender patterns across different forms of violence to further evaluate a more practical relationship of gender disparity in IPV and thus, this information should be incorporated for future works in the conceptualization of IPV (p. 152). As we see in the five additional articles, deficient conceptualization may be an overarching problem across literature.
The overall questions and theories that drive the studies that examine behaviors of IPV are grounded in similarities, though specific theories are not present in more than one study examined in this review. In other words, all five of the studies under examination analyzed different aspects and theories behind IPV, which in turn, reveals this field to be expansive in nature. There appears to be consensus that IPV includes both violent physical and mental abuse, resulting in injurious outcomes for those involved. Most of the research questions evaluated in the five additional studies pertain to the contributing factors and possible predictors of IPV behaviors to attempt to better explain the direction and patterns of these relationships across gender.
To begin, Edwards, Mattingly, Dixon, and Banyard (2014) introduced the social disorganization theory to explain IPV in their study. The social disorganization framework is meant to provide an explanation of the role of economic and neighborhood compositional factors on crime (p. 199). Though this framework has only recently been applied to the study of IPV, Edwards et al. attempt to link impoverished communities and higher rates of IPV and bystander intervention of IPV with perceived collective efficacy. In this way, Edwards et al. expand on the social disorganization framework. IPV was conceptualized based on the Conflict Tactics Scale Revised (CTS2) victimization and perpetration items while the Bystander Behavior Scale (BBS) conceptualized bystander intervention. Here, Hamby would attest to conceptualization issues in the use of the CTS.
Derefinko, DeWall, Metze, Walsh, and Lynam’s focus (2011) was more on the individual, rather than community factors. They introduce the General Theory of Crime, rather briefly. This theory proposes that low self-control is the most important factor in understanding and predicting criminality (p. 224). Derefinko et al. goes on to explain the link of self-control and aggression toward intimate partners, but questions how this association changes in different types of aggressive behaviors. Derefinko et al. aimed to conceptualize self-control through impulsivity by using a standardized self-report measure, called the UPPS Impulsive Behavior Scale (UPPS), that evaluates Positive and Negative Urgency, lack of Premeditation, lack of Perseverance, and Sensation Seeking. Researchers also used the Crime and Analogous Behavior Scale (CAB) to conceptualize both general violent behavior (GVB) and intimate partner violent behavior (IPVB), as well as the NEO PI-R, which evaluated personality dimensions. In addition, prior research had found that severe violent batterers were likely to show hypoarousal heart rates. In this case, autonomic arousal was conceptualized by blink magnitude, which was measured by an electromyogram (EMG). Though Derefinko et al. do not use the typical CTS self-report, it is fair to question whether they studied the population who experience IPV or the population of impulsive persons.
Filson, Ulloa, Runfola, and Hokoda (2010) examine powerlessness and its effects on depression in those who have been victimized by their partners in IPV relationships. Their research question stems from prior literature that found links between violence and power, and power and depression. Their definition of violence in this study focused on more aggressive violent acts and sexual abuse among romantic partners, similar to Hamby’s focus, as well as Derefinko’s study (Derefinko et al., 2011, p. 401). Filson introduces the “theory of gender and power” that states men hold more power than women in society, and this lack of equality is preserved, in part, through violence. Also introduced by these researchers is Emerson’s “social exchange theory” that evaluates power as “the extent to which one’s resistance can be potentially overcome by someone else.” As recognized in prior literature, Filson et al. introduce the theory that violence in IPV may cause a sense of powerlessness, and in turn, result in depression. Therefore, Filson et al. choose these variables to study and used a modified version of the CTS to conceptualize victimization, Sexual Relationship Power Scale (SRPS) to conceptualize power, and the Center for Epidemiological Study Depression Scale (CES-D) to conceptualize depression (p. 406). While Hamby would, again, argue against the conceptualization of IPV based on the CTS, it should also be of concern that perhaps Filson et al. are focused too much on the link between power and depression, rather than IPV and depression or IPV and power.
The most complex study of the five can be awarded to Lohman, Neppl, Senia, and Schofield’s (2013). Lohman et al. chose to evaluate individual and family risk factors, backed up by prior research that emphasizes the possibility of these variables as predictors rather than rooted in theory. Additionally, past literature recognizes the possibility of familial influences like early exposure of IPV and psychologically abusive parenting on the presence of IPV in adulthood. Thus, stemming from the intergenerational transmission theory, Lohman et al. evaluate psychological abuse, family stress, and the stress between parents as predictors of IPV, Each of these factors was conceptualized by a variety of self-report and observation methods. Additionally, these researchers evaluated substance use problems, sexual activity, antisocial behaviors, low self-esteem, association with deviant peers, negative emotionality, and the gender of the participants to predict IPV behaviors in adulthood.
While Edwards, Derefinko, and Filson attempt to evaluate predictors of IPV, Crane, Hawes, Mandel, and Easton (2014) attempt to evaluate the actual relationship patterns of IPV over time. Similar to Lohman et al.’s study, but without specific predictor components, Crane et al. focus on male against female partner violence (MFPV) and female against male partner violence (FMPV). These relationships were conceptualized to include any sexually or romantically involved partner. IPV was distinguished by the presence of severe physical aggression. The CTS2 was used to conceptualize aggression and victimization. Their decision to analyze MFPV and FMPV originated from previous theories in the bidirectionality pattern of gender violence in IPV relationships.
This leads to the discussion of the biggest disagreement among all studies, which lies in the theory that both men and women contribute to IPV equally. According to Hamby, there is no way to test gender patterns with the outdated technologies still used today: the CTS is often used in studies that corroborate with the bidirectionality theory, and Hamby believes this outdated measure is the source to blame for the inflated results that are in support of gender parity. It is noticeable that a majority of the additional articles use the CTS to conceptualize victimization and/or perpetration of IPV. Therefore, any problems with the CTS would threaten the studies that have included this scale in their conceptualization of IPV, and their findings and the way they produce their study should be analyzed with caution.
In comparison with Hamby’s study, it appears that most of the additional studies also attempt to examine similar aspects of violent physical and sexual abuse behaviors present in IPV and attempt to conceptualize IPV behaviors in a similar way, mostly through self-report measures, and most specifically, through the use of the CTS. Only one study, Lohman et al.’s, seems to evaluate the violent psychological behaviors found in IPV relationships, but does not neglect the matter of physical abuse. The five studies appear to base their conceptualization on different theoretical frameworks, and expectedly, examine different variables in their study in comparison to one another. Their use of the CTS in evaluating a multitude of different predictors proves difficult. The use of the CTS in such broad contextual situations, without efforts to manipulate items geared toward the specifics of a new study, may explain the reason for inconsistencies in research—inconsistencies most specifically revolving around the inconclusive gender patterns. However, all five studies are unique from one another and provide evidence of the complexity and broad knowledge present in current literature on IPV—but not without concern for faulty conceptualizations that may explain our limited understanding.
Methodological Issues
Research Design
Of all five additional studies, correlational designs were used. This appears to be one of the biggest similarities with the research Hamby used in her review. Therefore, it is apparent that a correlational study is the most common way in which researchers and those interested in this field gain information on IPV relationships. Data was mostly documented through the use of self-report measures. Additionally, the correlational data has appropriately expanded current understanding of what behaviors are involved in IPV and has incorporated a variety of possible predictors for these relationships. Although similar in their initial design, the researchers under review take on a variety of different approaches in their attempts to better explain and predict IPV, through the use of different predictors and participant recruitments. Furthermore, they accurately report their findings without suggesting causality. This is a key strength in the use of non-experimental research. However, the inability to demonstrate causality could also be considered a weakness of all the studies mentioned. As these studies are inquired further, obvious strengths and weaknesses grow more apparent, but all contribute to the cumulative knowledge of IPV.
Hypotheses.
Researchers evaluated different hypotheses in accordance with the variables they chose to review. A variety of predictions were analyzed among the five additional studies and all the researchers attempted to bridge some sort of gap in prior literature.
Edwards et al. recognized that most studies have been conducted in urban communities; they expand on earlier findings by questioning the factors of poverty and collective efficacy as predictors of perpetration, victimization, and bystander intervention in rural communities. Edwards et al. hypothesized that different community factors like poverty and collective efficacy would influence IPV rates. That is, they predicted that higher levels of poverty would be associated with higher rates of victimization and perpetration, and bystander effect would be positively correlated with perceived collective efficacy.
Derefinko et al. also evaluated possible predictors of IPV victimization and perpetration, but through individual traits as evaluated by the UPPS-P and autonomic response predictors. They hypothesized that impulsive personality traits and higher levels of autonomic response would predict IPV victimization and perpetration.
Crane et al. focused on the patterns of IPV over time. They predicted that both MFPV and FMPV patterns would present similar patterns and any changes in MFPV would show similar changes in FMPV.
Lohman et al. also focused on gender patterns but also individual and family predictors. They hypothesized that adolescent exposure to IPV would be positively correlated with the presence of partner psychological violence in adulthood, in addition to specific individual traits, exposure to parent-child psychological violence, and family stress. They also predicted they would see similar patterns of IPV throughout time.
Crane et al. and Lohman et al.’s findings would likely best be used against Hamby’s suggestion that gender symmetry is a fluke in research. Though it is important to note that both recognize that the severity of these behaviors may differ between men and women. Filson et al. predicted that power would be a mediator of IPV and depression. Evidently, all five articles have little in common in regards to the variables under study, especially in regards to Hamby’s article where only a few articles directly evaluate the patterns of IPV across gender. However, all the researchers’ hypotheses accurately stem from their conceptualization, theories, and prior findings, and the largest concern is whether researchers like Derefinko et al. and Filson et al. are too focused on other behaviors (impulsivity and power) that strays them from making conclusions for the actual IPV population. The only hypothesis that appears to be complex is Lohman et al.’s, whose hypotheses fail to meet the Occam’s razor criteria of parsimony and coherence (Weston, 2019). This is because their study evaluates a multitude of different individual factors and five different hypotheses that make their study highly complex in comparison to the other four.
Methods
Participants. The biggest problem in participant selection across all five additional articles was their sample participants were recruited through nonprobability sampling and all were volunteers. According to Meltzoff and Cooper (2018, p. 68), participants that are selected through volunteerism threaten the generalizability of the study. Their voluntary contributions mean that for whatever reason, these self-selected participants were aware of the study over others who may have been a part of the target population. This also means that these participants may have similar characteristics, and similar access to computers, flyers, or articles, and may be limited to a specific area of town. This would make the researchers’ study sample more homogenous in nature, something that randomized sampling would have resolved (Meltzoff & Cooper, 2018, p. 66).
Another issue made apparent in Crane et al.’s study is the source from which researchers recruited their volunteers. In this particular study, it is crucial to be critical of the males recruited who were referred by the court. This means that all those situations where an official record of IPV was not filed were left out of the study, in turn, limiting the study’s generalizability. However, as generalizability is the most evident limitation across all the studies, these limitations are recognized by the researchers, which should not go unnoticed as the studies’ strength. Participants across all five studies were heterosexual and limited diversely, providing further evidence that generalizability is severely limited in this field. Data is very minimal in regards to nonnuclear relationships. Without data on samples that are representative of the target population, findings cannot be conclusive and generalized to that target population. However, it can be argued that other studies may provide similar findings and have a different sample population that may cover the gaps in another study that had a different range of diversity. This means it might be beneficial to examine IPV studies as a whole, especially when generalizability in each individual study is lacking.
An additional weakness identified in the foundation article is that most studies collect data from college students. As Hamby notes, there is actually a large gap in current knowledge regarding violence in men and women in the college population (Hamby, 2014, p. 154). It could also be argued that the college population, by nature, experiences lower levels of violence and aggression. Despite her claims, it seems that college students are the most frequently sampled, perhaps for their convenience. Consider Filson et al.’s study, where 327 college students were recruited for their study and the 131 male college students used in Derefinko et al.’s study. The most apparent weaknesses of these studies are that no researchers considered other factors that may have influenced their results. For example, living on or off campus may make a difference in the victimization and perpetration of IPV. Perhaps those living off campus are more likely to experience higher levels of IPV because they are not living under a protected college community. Therefore, findings from these studies are questionable in terms of external validity.
Additionally, college students were likely to have been recruited in their classes, meaning participant characteristics may be more similar because of their interests in similar classes or their programs instituted by the college. Other students may have been selected for their preexisting IPV behaviors, like those selected in Derefinko et al.’s study. If researchers are selecting participants who will likely provide support for their hypothesis, internal and external validity is further threatened.
Moreover, both Filson et al. and Derefinko et al.’s studies only evaluated males or females, which further limits the generalizability of their findings and does not account for gender differences. Hamby would argue against studies that are limited to college students, especially in neglecting gender differences. She would also debate that without a foundation and broader view of the prevalence of IPV at this age, studies would not be able to generalize to the actual target population of all those that experience IPV. Hamby attempts to fill in these gaps by providing information on youth, college students, and adults in violence, but does not provide statistics of those specific to IPV. She also states that gender differences may be less apparent in younger people (Hamby, 2014, p. 153), which might explain why bidirectionality is so apparent in literature today, since most studies examine younger age groups, especially college students. Furthermore, most of the additional studies evaluated lower-risk individuals of IPV as well, which further brings to question whether their findings are actually significant when applied to the target population.
Measures. As noted prior, Hamby recognizes that scales like the CTS, a self-report method typically used to measure IPV may be inconclusive. She spends most of her report critiquing current literature findings. Though she only focuses on the CTS scale, she argues that it is reliable, but not valid. She contends that the CTS will not discern aggressive or violent behaviors that are due to sources other than IPV (Hamby, 2014, p. 151). This is problematic because researchers that use the CTS report their findings on IPV when really, they might not have been studying a population of people who experience IPV to begin with. Hamby states the use of scales like the CTS that have been around for 30-40 some years have contributed to a flawed operationalization of IPV (p. 154). She goes on to evaluate the CTS2 in her own study, which manipulated items on the scale in terms of partner-specific and nonspecific language (Hamby, 2014, p. 155). Reformatted items included “My partner pushed me” for partner-specific statements, and “Someone pushed me” for nonspecific statements.
Findings showed men were more likely to report IPV with partner-specific language while women were more likely to report IPV on nonspecific language items. Hamby concludes that researchers should not be so quick to conclude gender patterns based on measurements, like the CTS, that have been around for decades, when there is a lack of understanding of such methods and the way which people respond to questions on scales that are meant to assess violence. She advises that technological advancements must be made to focus in on these types of interactions that may be wrongfully identifying patterns of gender parity. Without further evidence and evaluation of current measurements of violence, her findings should not be considered as truth without critique. Hamby fails to consider other scales that measure violence like the CAB used in Derefinko et al.’s study. However, perhaps because the scale is less commonly used, it missed Hamby’s radar whose main focus was to critique more cliché measurements.
Despite the debate on the validity of the CTS, Edwards et al., Crane et al., and Filson et al. use this measurement, and the other two studies use a variation of other self-report measures. Seeing that the CTS is used in such a broad range of studies that evaluated different variables, it does raise the question of whether it is truly valid without any kind of innovative changes in the scale over the last few decades or changes geared specifically toward the studies’ participants. Without being geared to the study’s purpose and participants, it seems far-fetched to apply it in any possible IPV situation. Though none of the additional articles add to the field according to Hamby’s standards, what was revolutionary about Lohman et al.’s study was their use of their own self-report measurements as well as the addition of an observer rating.
In Lohman et al.’s study, they videotaped interactions between the participants’ parents, parents and participants, and the participant and his/her romantic partner and an observer coded for verbal attacks to come up with a final score. There are obvious limitations to the use of a camera to measure behaviors: the camera itself may influence the way participants interacted with one another because of simple awareness that they are being recorded. Additionally, each conversation was given limited time but the time was not equated across each recording period, which might influence the results as well. It brings to question whether participants would grow more comfortable and more likely to behave naturally as time passes over shorter recorded periods. However, these researchers expand on simple self-report measures. Recording interactions in the hopes to observe some aspect of IPV is novel to the field, though ethically limited. Therefore, Lohman et al. were able to evaluate psychologically violent behaviors associated with IPV through verbal attacks, but not severe physical violence. Accurately, Lohman et al. introduce a new way to ethically study aspects of IPV that contribute to a better understanding of the behaviors involved.
Additionally, Derefinko et al. came up with their own measure that evaluated autonomic arousal. They called this the startle response activity in which participants were presented with three images of each category: pleasant, neutral, and unpleasant images. Images were randomized and shown for six seconds at a time. A loud noise was presented after the sixth image. An EMG calculated blink magnitude, which they used to measure participants’ reactivity. They failed to consider possible order effects in response reactions after the sixth image. It is quite possible that reactions might be different depending on whether a pleasant, neutral, or unpleasant stimulus was presented. Additionally, they are unable to distinguish the difference in response between pleasant and unpleasant stimuli. They both may present with similar physiological reactions, which makes it questionable whether tracking blink responses is the most effective way to analyze the reactions.
Though self-report measures may include a range of participant biases like social desirability bias that influence the study’s results (Weston, 2019), it is the surest way to examine intimate behavior like IPV in an ethical way. Due to its intimacy, researchers are limited to measurements that face restrictions themselves, which may explain why the technological advancements Hamby suggests are essential to grow in this field are difficult to actually initiate advancements in. This is not to say researchers are not attempting novel ways to evaluate their variables under study, as was shown in the additional studies.
However, it is important to be aware that some of the measures may not have been examining what researchers intended them to. Consider the use of the Bystander Behavior Scale (BBS) in Edwards et al.’s study. Items on this scale prompted respondents to reply with “yes” or “no” based on items like “Approach a friend if I thought they were in an abusive relationship and let them know that I’m here to help” or “If I heard a friend insulting their partner I would say something to them.” This scale does not allow researchers to predict the respondents’ behavior if the person in need was a stranger, which severely limits the scale and does not evaluate all aspects of bystander intervention.
It can be argued that the researchers could have done a better job in describing the validity of the measurements included in their study, to further strengthen their decision to use biased self-report measures. Most of the studies mention reliability, and slim to none mention validity. In most cases, if validity was mentioned, another study had validated the measure for them, which appears to be the easy way out.
Procedures. Though Hamby fails to delve into the procedural aspects in her literature review, it is not necessary for her to do so, considering that her study is a review of the statistical findings of past research. Surprisingly, the additional studies under review present rather vague descriptions of the procedures implemented, thus contributing to yet another commonly seen weakness in the field. Without a clear procedural approach, the ability to replicate a study is threatened, which further weakens the external validity of a study (Meltzoff & Cooper, p. 62).
Consider Lohman et al’s study, which was the most complex of all five additional studies. First, it is not verified whether the Iowa Youth and Families Project (IYFP) and Family Transitions Project (FTP) were the recruiters for the participants in the study, but rather it is implied. Second, it is not explained which interaction tasks were initiated first, since the order might influence participants’ behaviors in each videoed task. Additionally, researchers did not verify how long they recorded parent and adolescent engagements. Without this information, it seems even more difficult to replicate an already complex study. Also consider Filson et al.’s study, the most simple of all five additional studies. Procedure information was at minimum. This could be due to the simplicity of the study and variables under examination. The extent of their procedure was the issuance of waivers and confidentiality agreements and then the surveys to collect data. Nothing was mentioned of time and place the survey was initiated, or whether a proctor was available.
In Crane et al.’s study, participants underwent a twelve-week treatment, as mentioned in the procedure section. Since the study itself did not examine direct effects of this treatment, the treatment details were limited. Researchers eventually state the treatment was a Cognitive Behavioral Treatment (CBT), but even these details were not included until the discussion section, and there was no explanation of what the treatment may have entailed. This should be considered a huge issue in Crane et al.’s study because the treatment itself may have some sort of effect on the results, and would lead to questions of whether the same findings in this study would actually be reproduced without the implementation of a treatment. Thus, the treatment may be a confounding variable of Crane et al.’s study. Since all five studies were imprecise about their procedures, and Hamby neglects procedures all together, it is apparent that procedures are not emphasized as important in this field, which stands as a clear weakness of all additional studies and the foundation article, and in turn, might explain the limited generalizability of studies across the board in this field.
Analysis/Results
As expected, researchers use a variety of different correlational analyses to produce their results. The researchers of the five additional articles found overwhelming evidence and support for their hypotheses. These results should not go without critique since questions were raised of the validity of many of the measurements used to collect data, as well as researchers’ possible misconceptualization of IPV. However, most analyses include t-tests and correlational coefficients. Edwards et al. used Poisson distribution equations to examine both community and individual level scores simultaneously (p. 202). They found support that higher poverty levels predicted higher IPV victimization (B = .12, p<.001) and perpetration (B = .07, p<.001) levels. Though they found collective efficacy influence victimization and perpetration reports in men (B = -.37, p<0.001), but not women (B = .06, ns).
Derefinko et al. found support for the hypotheses, that is they found Lack of Premeditation to be correlated with GVB (r = .17, p<.05), and Urgency to be correlated with IPVB (r = .26, p<.01). The fact that they did not find IPVB to be correlated with Lack of Premeditation suggests a distinct quality of GVB and IPVB. Additionally, though they did not find a link between autonomic arousal and GVB as their hypothesis stated, they did find the link with IPVB (r = .16).
Crane et al. found a decrease of victimization and perpetration rates of both male and female partners from the first time point to the second. Severe FMPV was less common than minor acts at both times (t(66) = 3.17, p = .02) and (t(41) = 1.00, p = .32). Minor acts of FMPV decreased from the first to second assessment, whereas severe FMPV was unlikely at both assessments. Consider these findings in minor aggression t(41) = 2.55, p =.02, severe aggression, t(41) = 2.08, p = .04, psychological responding, t(41) = 2.87, p < .01, and active responding, t(37) = 3.88, p<.01. Consider t-test scores for MFPV psychological aggression, t(41) = 2.34, p = .02, and physical aggression over time, t(41) = 1.89, p = .07. Additionally males reported co-occurring activity, which lends support of both their hypotheses.
Lohman et al. found support for all of their hypotheses, except that IPV between parents would predict psychological violence in the target adolescent. Negative emotionality and sexual activity predicted higher levels of psychological violence during both emerging adulthood and adulthood. Additionally, academic difficulties and gender positively predicted psychological violence during emerging adulthood. Evaluations for substance use, antisocial behaviors, low self-esteem, and deviant peers were not statistically significant. Exposure to parental intimate partner violence did not predict intimate partner psychological violence at either time points. However, parent to child psychological violence did predict psychological violence in emerging adulthood and adulthood, as did exposure to family stress. Family stress seemed to predict psychological violence in adulthood, but not emerging adulthood. However, numbers were not provided for any of the above findings, which is a limitation of their study. Lohman et al. also found that there was consistency in the pattern of victimization (r = .34 for males, .47 for females) and perpetration (r = .17 for males, .53 for females) in emerging adulthood to adulthood. Victimization and perpetration were correlated from both time points (Emerging adulthood r = .78 for males, .70 for females and adulthood r = .68 for males, .69 for females).
Lastly, Filson et al. found support of both hypotheses: They found associations between the CTS and SRPS r(1, 323) = .404, p<.001 , in addition to an association between the SRPS and CES-D r(1, 322) = .275, p < .001. There was also a statistically significant relationship found between the CTS and CES-D r(1, 324) = .214, p < .001. When scores were regressed onto the CTS and SRPS, the significance decreased [r(2, 321) = .122, p = .037], which established power as a potential mediator of IPV victimization and depression.
A lot of the above studies set up the possibility for future research, but provide the bear minimum to their study, neglecting areas where they could have greatly expanded their findings in more thorough detail. Consider Crane et al.’s study where the ability to perhaps examine the direction of the relationship, as well as the differences in severity of the behaviors of IPV, presented itself, but no efforts were made to do so. If Hamby were to review these studies, she would argue they do little to add to systemic and technological advances. However, it can be argued that Lohman et al. introduce a novel element of observation, which suggests further developments in this area pertaining to improvements in data collection and a broadened range of observable behaviors to code.
As made apparent across all five studies, individual traits, family factors, and community influences may predict IPV behaviors. Additionally, the supplementary studies’ findings suggest similar patterns of IPV among males and females, both across time and within the relationship, which contradicts Hamby’s argument for gender disparity. This further proves that the additional studies have added little to expand on Hamby’s implications. Crane et al.’s study is essentially the only one of the five that supports the case of gender parity, though Lohman et al. found weak factorial variance between gender variables, meaning there were no significant differences between gender.
Correlations ranged from relatively low to medium (r = .17-.78). Interestingly, most of the researchers evaluated scores ranging from .11-.34, which are relatively weak correlations. Nonetheless, all the studies confirm the presence of the relationships with fervor, but barely emphasize the weakness of these associations, which seems rather deceptive. They might suggest these weak correlations are due to participants being low-risk, but only a couple studies incorporate this in their study’s limitations (Derefinko, DeWall, Metze, Walsh, & Lynam, 2011; Filson, Ulloa, Runfola, & Hokoda, 2010; Lohman, Neppl, Senia, & Schofield, 2013). However, Filson et al. also argue that since links were found in support of their hypotheses, these associations would only be higher in higher-risk individuals, which may be something of importance when moving forward (p. 411). However, it is important to realize that weak correlations and weak generalizability may mean weak findings.
Summary and Conclusion