16
ORIGINAL ARTICLE Conceptualization of Intimate Partner Violence: Exploring Gender Differences Using Concept Mapping Patricia OCampo 1 & Yu Janice Zhang 1 & Mairi Omand 2 & Alisa Velonis 3 & Michael Yonas 4 & A. Minh 1,5 & Ajitha Cyriac 1 & Farah Ahmad 6 & Janet Smylie 1 Published online: 7 June 2016 # Springer Science+Business Media New York 2016 Abstract While numerous studies have explored prevalence and determinants of intimate partner violence (IPV), one area that has yet to be sufficiently explored is whether men and women agree on the acts, behaviours, and attitudes that com- prise IPV. Through the use of concept mapping, we examined the similarities and differences in the conceptualization of IPV among a diverse sample of men and women. Although men and women conceptualized physical and sexual violence similarly, men tended to rate non-physical behaviors related to control as less important to the definition of IPV than women. Moreover, even when men and women used similar labeling language for grouping of items, the statements included in each cluster were substantially different for non-physically abusive items. A greater understanding of similarities and differences in the conceptualization of IPV by gender can help inform appropriate gender specific IPV intervention and prevention efforts. Keywords Partner abuse . Semi-quantitative methods . Men . Women . Measurement Over the past 30 years, intimate partner violence (IPV) has increasingly become recognized as a public health problem, both within national and international settings. Although esti- mates of the rates of IPV vary widely by setting and surveil- lance method (Alhabib et al. 2010), recent data from the United States suggests that one in three women and one in four men in that country have, at some point, experienced rape, physical assault, or were stalked by an intimate partner (Black et al. 2011). Furthermore, IPV has been associated with numerous physical and mental health problems, including injuries, gastrointestinal disorders, asthma, chronic pain, suicide attempts, substance use, sexually transmitted infec- tions, depression, anxiety, and symptoms of post-traumatic stress disorder (PTSD) (Black et al. 2011; Bonomi et al. 2006; Krug et al. 2002). In recent years, there has been increasing controversy among IPV researchers about gender specific factors and in- fluences in the perpetration, victimization, and understanding of domestic violence (Anderson and Umberson 2001; Dutton 2012; Johnson 2011). While much of this debate has centered on the question of whether women are as violent as men, one outcome has been increased discussion about the centrality of coercive and controlling behaviours to IPV. Stark (2007) de- scribes coercive behaviours as those involving force or threats to obtain a desired response, and control as structural forms of deprivation, exploitation, and command that compel obe- diencethrough the use of behaviours that limit options and deny resources (pp. 228229). It is the use of a pattern of coercive and controlling tactics, he claims, that differentiates fightsthat may turn violent from true domestic violence.Likewise, Johnson has found support for hypotheses that * Patricia OCampo o'[email protected] 1 Centre for Research on Inner City Health, St Michaels Hospital, Dalla Lana School of Public Health, University of Toronto, 30 Bond, Toronto, ON M5B1W8, Canada 2 Centre for Addiction and Mental Health, Toronto, ON, Canada 3 Community Health Sciences, University of Illinois at Chicago, School of Public Health, 1603 West Taylor St, Chicago 60612, Illinois 4 The Pittsburgh Foundation, Five PPG Place, Suite 250, Pittsburgh, PA 15222, USA 5 Department of Epidemiology, University of British Columbia, 2329 West Mall, Vancouver, BC V6T 1Z4, Canada 6 Faculty of Health, York University, 4700 Keele St, Toronto, ON M3J 1P3, Canada J Fam Viol (2017) 32:367382 DOI 10.1007/s10896-016-9830-2

Conceptualization of Intimate Partner Violence: Exploring Gender … · 2018-06-05 · ORIGINAL ARTICLE Conceptualization of Intimate Partner Violence: Exploring Gender Differences

  • Upload
    others

  • View
    0

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Conceptualization of Intimate Partner Violence: Exploring Gender … · 2018-06-05 · ORIGINAL ARTICLE Conceptualization of Intimate Partner Violence: Exploring Gender Differences

ORIGINAL ARTICLE

Conceptualization of Intimate Partner Violence: ExploringGender Differences Using Concept Mapping

Patricia O’Campo1 & Yu Janice Zhang1 & Mairi Omand2& Alisa Velonis3 &

Michael Yonas4 & A. Minh1,5& Ajitha Cyriac1 & Farah Ahmad6

& Janet Smylie1

Published online: 7 June 2016# Springer Science+Business Media New York 2016

Abstract While numerous studies have explored prevalenceand determinants of intimate partner violence (IPV), one areathat has yet to be sufficiently explored is whether men andwomen agree on the acts, behaviours, and attitudes that com-prise IPV. Through the use of concept mapping, we examinedthe similarities and differences in the conceptualization of IPVamong a diverse sample of men and women. Although menand women conceptualized physical and sexual violencesimilarly, men tended to rate non-physical behaviors relatedto control as less important to the definition of IPV thanwomen. Moreover, even when men and women used similarlabeling language for grouping of items, the statementsincluded in each cluster were substantially different fornon-physically abusive items. A greater understanding ofsimilarities and differences in the conceptualization of IPVby gender can help inform appropriate gender specific IPVintervention and prevention efforts.

Keywords Partner abuse .Semi-quantitativemethods .Men .

Women .Measurement

Over the past 30 years, intimate partner violence (IPV) hasincreasingly become recognized as a public health problem,both within national and international settings. Although esti-mates of the rates of IPV vary widely by setting and surveil-lance method (Alhabib et al. 2010), recent data from theUnited States suggests that one in three women and one infour men in that country have, at some point, experiencedrape, physical assault, or were stalked by an intimate partner(Black et al. 2011). Furthermore, IPV has been associatedwithnumerous physical and mental health problems, includinginjuries, gastrointestinal disorders, asthma, chronic pain,suicide attempts, substance use, sexually transmitted infec-tions, depression, anxiety, and symptoms of post-traumaticstress disorder (PTSD) (Black et al. 2011; Bonomi et al. 2006;Krug et al. 2002).

In recent years, there has been increasing controversyamong IPV researchers about gender specific factors and in-fluences in the perpetration, victimization, and understandingof domestic violence (Anderson and Umberson 2001; Dutton2012; Johnson 2011). While much of this debate has centeredon the question of whether women are as violent as men, oneoutcome has been increased discussion about the centrality ofcoercive and controlling behaviours to IPV. Stark (2007) de-scribes coercive behaviours as those involving force or threatsto obtain a desired response, and control as “structural formsof deprivation, exploitation, and command that compel obe-dience” through the use of behaviours that limit options anddeny resources (pp. 228–229). It is the use of a pattern ofcoercive and controlling tactics, he claims, that differentiates“fights” that may turn violent from true “domestic violence.”Likewise, Johnson has found support for hypotheses that

* Patricia O’Campoo'[email protected]

1 Centre for Research on Inner City Health, St Michael’s Hospital,Dalla Lana School of Public Health, University of Toronto,30 Bond, Toronto, ON M5B1W8, Canada

2 Centre for Addiction and Mental Health, Toronto, ON, Canada3 Community Health Sciences, University of Illinois at Chicago,

School of Public Health,1603 West Taylor St, Chicago 60612, Illinois

4 The Pittsburgh Foundation,Five PPG Place, Suite 250, Pittsburgh, PA 15222, USA

5 Department of Epidemiology, University of British Columbia,2329 West Mall, Vancouver, BC V6T 1Z4, Canada

6 Faculty of Health, York University, 4700 Keele St,Toronto, ON M3J 1P3, Canada

J Fam Viol (2017) 32:367–382DOI 10.1007/s10896-016-9830-2

Page 2: Conceptualization of Intimate Partner Violence: Exploring Gender … · 2018-06-05 · ORIGINAL ARTICLE Conceptualization of Intimate Partner Violence: Exploring Gender Differences

distinguish coercive/controlling violence from “situational”IPV (Johnson 2008; Johnson and Leone 2005). The literatureon men’s and women’s use of coercive control is mixed, withsome finding that women are the subjects of coercive controlmore often than men (Ansara and Hindin 2010; Tanha et al.2010) and others finding virtually no differences (Cho 2012).

It has been noted that compared to studies focusing ondeterminants of partner violence, few empirical studies havefocused their attention on capturing the views of the generalpublic on what behaviors constitute IPV (Makepeace 1986;Muehlenhard and Kimes 1999; Beyers et al. 2000; Carlsonand Worden 2005). In a random sample of 200 communityresidents, Carlson andWorden (2005) investigated gender dif-ferences and other demographic differences in attitudes aboutwhat kind of behaviours should be considered as domesticviolence (DV). Men were slightly less likely to view wivesslapping husbands as DV; both genders had similar percep-tions about whether husbands slapping wives, and boyfriendsor girlfriends slapping their partners constitute DV.

To fully understand the intricacies of IPV, we must build onthe current understanding of the similarities and differences inwomen’s and men’s perceptions of domestic violence. From atheory standpoint, much has been written about why men andwomen hold different attitudes toward perpetration and vic-timization of partner violence (Pease and Flood 2008; Flynnand Graham 2010). While attitudes vary widely by locality,men, more than women, may hold hostile or negative sexualattitudes toward women or identify with traditional “macho”gender role attitudes and may, therefore, be less likely thanwomen to view certain sexual or physical acts as being abu-sive (Heise 1998; O’Neil & Harway, 1997; Murnen et al.2002). Similarly, women, compared to men, are more likelyto express agreement with traditional gender roles or ‘self-silencing’ behaviours, impacting women’s likelihood of view-ing violent behaviours as abusive (Harris et al. 2005). Studiesof college students’ and community-dwelling participants’ re-actions to scenarios of physical IPV where the gender of theperpetrator and victim varied, found that both women andmenhad more sympathy for female victims and took scenarios ofmale battering of female victims most seriously (Harris andCook 1994; Seelau and Seelau 2005; Sorenson and Taylor2005; Stewart-Williams 2002). In addition, regardless of thegender of the victim, women reacted more strongly to thedescription of the battering incident and expressed greatersympathy for the victim than men, suggesting that womenmay identify more strongly with the victim (Harris andCook 1994). However, these studies generally focus on sce-narios of physical IPVand it is unclear how these attitudes andperceptions might differ for scenarios of non-physical IPV.

Yet, this inquiry is not merely theoretical in nature.Violence-supportive norms have been linked to rates of vio-lence (Heise 1998; Harris et al. 2005). Moreover, current pre-vention and intervention efforts have demonstrated only

limited success, possibly due to an incomplete understandingof the experiences of and motivations for the use of violenceagainst partners (Langhinrichsen-Rohling 2010) and howthose differences may vary by gender. If men and womendiffer in their conceptualizations of IPV, this suggests thatinterventions must begin with the foundational work ofcreating greater awareness and a common understandingabout the problem, especially for non-physical violence.Moreover, such awareness raising efforts may have to begender specific. Through the use of concept mapping, thisstudy examined the similarities and differences in the con-ceptualization of IPV among a sample of men and womenrecruited from an urban setting. Specifically, we wanted tounderstand:

a) How men and women relate non-physically violent be-haviors to the concept of intimate partner violence; and

b) How gender influences the perceived importance of thesebehaviors on the meaning of IPV and on the need forintervention.

Methods

We employed Kane and Trochim’s (2007) approach of con-cept mapping, a participatory research method that seeks toclarify the conceptual elements, domains, and underlyingframework of a phenomenon as experienced by a populationof interest (Trochim, 1985; Galvin 1989; Trochim, 1989). It isa structured process that integrates participants’ input on asingle topic and produces an interpretable pictorial view ofparticipants’ ideas, conceptualizations, and their internal relat-edness with an emphasis on engaging participants and includ-ing them in the interpretation of findings (Burke et al., 2005;Trochim, 1989). Information about recruitment and study pro-cedures are presented below but can also be found in greaterdetail elsewhere (Author reference).

Recruitment and Data Collection

We recruited our sample mainly by strategically placing re-cruitment flyers in organizations across the city which servepopulations of interest for a study on “relationship stress”(e.g., community based and social services organizations).Direct mention of IPV in the flyers was avoided so that poten-tial participants in abusive relationships would not be placedat risk of violence for seeking information about or participat-ing in such a study. Potential participants were screened overthe phone to determine eligibility (age 18 or older; comfort-able speaking, reading, and writing in English; and, availableto come to one or more of the group sessions) and informed ofthe exact nature of the study by trained personnel. For some

368 J Fam Viol (2017) 32:367–382

Page 3: Conceptualization of Intimate Partner Violence: Exploring Gender … · 2018-06-05 · ORIGINAL ARTICLE Conceptualization of Intimate Partner Violence: Exploring Gender Differences

groups, women participants brought along friends to partici-pate. Two rounds of recruitment were undertaken to achieveour intended sample size. Recruitment and data collectiontook place over an eight month period.

Concept mapping has the flexibility of including partici-pants in one or more phases of the study. Often, participantsdo or do not participate in one or another activity because oftime constraints such that they cannot make a scheduled groupsession or are not able to complete the online session prior tothe closing date. Our goal is to ensure that ample participants,in this case for both genders, take part in all three phases of thestudy.

Concept Mapping Methodology

There are three distinct participant-involved activities associ-ated with the concept mapping process: brainstorming; sortingand rating; and mapping and interpretation of the data(Trochim, 1989). Each set of activities was implemented inthree separate phases of our study. Brainstorming, sorting, andrating could be accomplished individually online or face toface as part of a group. All group sessions (4 men’s and 4women’s brainstorming sessions and 2 men’s and 2 women’ssorting and rating sessions and 1 men’s and 1 women’s inter-pretation sessions) were separated by gender in order to pro-vide an environment where both women and men felt com-fortable to speak freely.

Brainstorming Brainstorming activities involved free-listingof ideas and written responses through independent onlineparticipation and in-person groups. A total of eight group ses-sions were held each approximately 1.5 h in length.

The purpose of the brainstorming was to generate ideas thatparticipants considered to be descriptive of intimate partnerviolence. We used best practices to guide the generation ofour focal question (Kane and Trochim 2007) which wasintended to promote open-ended exploration of several dimen-sions of IPV: “What are the behaviours or attitudes that wouldmake up the part of the relationship characterized by conflict,abuse, excessive control, neglect or even violence?” The term“or even violence” was added after piloting the question asparticipants, perhaps because of the sensitivity of the topic,avoided discussion of abuse and violence without it.Participants were further prompted by more specific probesto ensure that a wide range of areas in relationship conflict wascovered (e.g. participants were asked to think about issues ofneglect, control, or conflict in relationships where partnersmay have tried to part ways.)

Approximately 870 statements were collected during allbrainstorming activities and this list was then reduced into afinal list of 71 unique items by the research team. Duplicateitems were eliminated, and similar items were grouped underone statement. The final list of items thus represents the

“universe of all possible notions of partner abuse” for theentire study sample that participated in brainstorming.

Sorting and Rating The sorting and rating activity is com-pleted independently by participants, and for this project themajority of participants completed the sorting and rating ac-tivity online. However, opportunities to complete this usingpaper and pencil were offered for those who did not haveaccess to a computer, did not feel comfortable using a com-puter, or just preferred to attend in person. The main differencebetween data obtained via the two modes of collection is thatwe had higher completion rates (100 %) for those attendingthe groups compared to those working online (72 %) as thoseworking online could complete their sorts and ratings over oneor more sittings and sometimes failed to complete the tasks.

The sorting activity facilitated participant input into theinterrelatedness of the 71 ideas identified during the brain-storming activity by asking participants to independently sortitems into conceptually similar groups that “made sense tothem.” Participants were then asked to label their groups inaccordance with the theme of the items. The following restric-tions were given to participants: individual statements couldnot be placed in more than one group; all statements could notbe placed in the same group; and the number of groups had tobe less than the number of items.

The rating activity provided the opportunity for partici-pants to rate each of the 71 items based on various ratingquestions using a 5-point Likert scale, (where 1 is the lowestscore and 5 the highest): 1) How important is each item inmaking up relationship violence, abuse, or severe conflict; 2)For each item, how strongly do you feel that the victim shouldget professional help; and 3) For each item, how strongly doyou feel the perpetrator should get professional help? Theliterature contains studies that ask participants about appropri-ate legal responses when victims experience certain behav-iours or acts (Dennison and Thompson 2011) but no studieshave asked about appropriate health professional responses aswe do here.

Mapping The data obtained from the sorting and ratingactivities were entered into Concept Systems GlobalMax© software. Two statistical steps were performed inorder to create a representation of participants’ perceptionsof relationships between items in a concept map. First,multidimensional scaling (MDS) (Kane and Trochim 2007)was performed on a similarity matrix of participants’ sorteditems. Multidimensional scaling arranges points, representingitems, on a spatial field such that the distance between pointsreflects the frequency with which items were sorted together.Thus, items that were sorted together by more people appearcloser to one another on the map. Second, hierarchical clusteranalysis, by means of Ward’s minimum variance method(Kane and Trochim 2007), was used to partition the MDS

J Fam Viol (2017) 32:367–382 369

Page 4: Conceptualization of Intimate Partner Violence: Exploring Gender … · 2018-06-05 · ORIGINAL ARTICLE Conceptualization of Intimate Partner Violence: Exploring Gender Differences

coordinates into distinct clusters of items. The cluster analysisprocess places similar items into non-overlapping clusters,representing conceptual domains, and the results are displayedas a cluster map.

In order to obtain gender-specific representations of theconcept map, the research team decided to create two distinctconcept maps and hold separate map interpretation sessionsfor male and female participants. Stress values, the scalingdiagnostic statistics used in multidimensional scaling to deter-mine goodness of fit of the map, were calculated for group-specific maps. A lower stress value implies a better fit andideal values fall within the range of 0.21 and 0.37 (Kane andTrochim 2007). The stress value for the female map was 0.28and for the male map was 0.34.

Prior to the group mapping sessions, members of the aca-demic study team used the Concept Systems software to ex-plore a variety of cluster map solutions for the men’s andwomen’s data separately. Concept Systems applies hierarchi-cal cluster analyses and multidimensional scaling to the dataand is able to use those findings to display maps comprised ofdifferent numbers of clusters. The research team examined arange of cluster solutions, for the male and female dataseparately, starting with a 15-cluster solution map and pro-gressively moving down to a 4-cluster solution map.Examining each map in terms of conceptual cohesionwithin clusters as a guide, the research team selected andprepared an 8 cluster solution map for female and a 7-cluster solution map for male data to bring to the genderspecific group interpretation sessions where participantshad the opportunity to make modifications to their finalcluster solution map. During the group discussion sessions,participants also were shown maps with different clusternumbers (e.g., cluster sizes from about 6 to 10).

One women’s (n = 11) and one men’s (n = 9) group washeld. Both groups were heterogeneous in terms of age, ethnic-ity, and socioeconomic background of the participants(Table 1). A female facilitator (PO) led the women’s groupand a male facilitator (MY) led the men’s group to ensureparticipants were comfortable sharing their opinions. The fa-cilitator helped participants evaluate and interpret the finalcluster solution maps to finalize the maps and to ensure theyreflected the participants’ perceptions of IPV. Emphasis wasplaced on viewing the map and confirming the final number ofclusters, ensuring all statements were in the appropriate clus-ters and labeling each cluster. Movement of items into moreappropriate clusters was permitted, leading to the redrawing ofcluster boundaries to incorporate the statement and thus thecreation of overlapping clusters.

The cluster labeling process was facilitated and guided byconsensus decision-making where input and agreement wassought for each final cluster name. Participants were also askedto examine the list of statements within each cluster and dis-cuss whether certain statements were inappropriately situated

and potentially a better fit within another cluster grouping.Each of these identified items to be relocated were discussedby the group and through consensus, a decision was madewhether to change its cluster designation. This process contin-ued until group consensus was achieved on all 71 items. Whilethere was a great deal of discussion with some opposing ideasexpressed during the concept mapping interpretation sessions,all disagreements were discussed and resolved openly and untilconsensus was reached. Following the extensive group discus-sion, the women’s and men’s group both chose a 7-clustersolution map. The results of both mapping sessions areoutlined below. The research team tracked comments and de-cisions made during the mapping groups using written notes.

Results

Demographic characteristics of the men and women who par-ticipated in our study are presented in Table 1. A greater

Table 1 Participant Demographics

Men (n = 26) Women (n = 45)

Ethnicity

Canadian 6 13

South Asian 15 12

East Asian 3 6

European 2 4

African 0 2

West Indian 0 4

Aboriginal 0 1

Other/Unknown 0 3

Age

20–29 1 12

30–39 6 11

40–49 14 16

50+ 5 6

Marital Status

Married/common-law 17 17

Divorced 2 2

Single 4 15

Separated 1 4

Widowed 0 3

No Response 2 4

Income

Under $20,000CAD 6 6

$21,000–35,000 4 7

$36,000 - $50,000 2 8

$51,000–100,000 3 9

Over $100,000CAD 10 8

No Response 1 7

370 J Fam Viol (2017) 32:367–382

Page 5: Conceptualization of Intimate Partner Violence: Exploring Gender … · 2018-06-05 · ORIGINAL ARTICLE Conceptualization of Intimate Partner Violence: Exploring Gender Differences

proportion of women fell into the 20–29 age range, were notmarried, and identified as Canadian compared to men.

Women’s Map

During the women’s participatory mapping and labeling exer-cise, a great deal of discussion took place that led to the orig-inal 8-cluster map being condensed to 7-clusters and numer-ous statements, 22 out of the 71, being shifted between clus-ters. Figure 1 represents the final cluster map that emergedfrom this iterative process.

Most of the labels that women used to define the clusterswere reflective of the language traditionally used to describe“domestic violence”: physical, sexual, and emotional abuse,control, and manipulation. In addition to clustering togethermore obvious forms of physical and sexual violence,women included actions such as destroying personal propertyand other behaviors that resulted in (or had the potential toresult in) physical harm, including interfering with a victim’saccess to a health care provider or denying a victim or childrenbasic needs.

Most of the non-physically violent statements fell into theclusters labeled Control, Emotional Abuse, and Manipulation.Whereas the Control cluster encompassed behaviors that ac-tively restrict a victim’s personal agency (such as restrictingthe victim’s access to work, school, friends, and family,restricting family finances, and controlling a victim’s physicalappearance), Emotional Abuse and Manipulation comprisedactions that enable a perpetrator to gain direct or indirect ad-vantages over a victim. Emotionally abusive behaviors wereclearly intended to cause the victim emotional harm and dam-age self-esteem (such asmaking a victim feel crazy, minimizingthe perpetrator’s actions, and turning others against the victim),while manipulative conduct exploited a victim’s feelingsand/or actions. Emotional blackmail (i.e., threatening suicide

or divorce), making someone feel that s/he cannot do any-thing right, and withholding information about a secret life-style all create a power imbalance that enables a perpetratorto indirectly control a victim’s sense of reality.

External and Cultural Influences incorporated statementsthat focused on the causes or mechanisms of abuse that wereconsidered outside of the relationship itself. Abusing a victimas a result of drug or alcohol use or a gambling addiction orpunishing a victim on “issues related to child gender” (e.g.,blaming the woman because the child is not a boy) refer toreasons why a perpetrator may abuse a victim. Similarly, usingcultural values to excuse violence or slanting religious viewsto encourage the abuse illustrates how cultural forces can beused as tools of abuse. Finally, Victim Responses to Abusecontained all of the statements related to behaviours takenby victims, such as destroying the personal property of a per-petrator, provoking the use of violence, and abusing the per-petrator in response to abuse.

Men’s Map

Like the women’s session, the men’s participatory mappingand labeling exercise also consisted of extensive iterative dis-cussion about the computer-generated map and relationship ofstatements. The men were presented with a seven-cluster map,which did not change, although some statements, 6 out of thefull set of 71, were moved to alternate clusters following ex-tensive discussion and consensus, as men developed the over-all themes of the clusters (Fig. 2). Like the women, the menorganized descriptions of physical and sexual violence intoseparate categories, although they included the majority ofvictim-driven behaviors (e.g., victim abusing the perpetratorin response to the abuse, victim provoking the perpetratorto use violence) within these clusters rather than creating aseparate grouping.

Fig. 1 Concept map generatedfrom women’s group

J Fam Viol (2017) 32:367–382 371

Page 6: Conceptualization of Intimate Partner Violence: Exploring Gender … · 2018-06-05 · ORIGINAL ARTICLE Conceptualization of Intimate Partner Violence: Exploring Gender Differences

Unlike women, men engaged a very different frameworkfor organizing the non-sexually and physically violent behav-iors. The men created a cluster labeled Financial Abuse,which focused on a perpetrator’s use of financial control togain and maintain power over a victim (e.g., blocking a vic-tim’s access to work or school, restricting activities, and con-trolling finances). A cluster labeled Emotional appeared tofocus on the use of attacks on emotional well-being as a wayof gaining an upper hand (e.g., public humiliation, encourag-ing children to take part in abusive behaviors, and displayingpower over a victim). Also included in this cluster were state-ments referring to the mental health of both perpetrators andvictims (e.g., the perpetrator abusing the victim because ofdrugs or alcohol use or as a result of his/her mental illness).

The largest cluster was labeled Family/Culture-OrientedAbuse, which centered on the perpetrator’s use of familialand cultural dynamics as a means of controlling the victimor justifying that control. For example, actions such askeeping the victim and her/his children separated, turningother people against the victim, denying the victim orchildren basic necessities, and denying wrongdoing towardthe victim in front of family or friends are ways in whicha perpetrator can use a victim’s relationships with familyas a tool of control.

The cluster entitledControl/Personal/Intimate/Psychologicaldescribed the use of humiliation as a means of establishingcontrol (e.g., the perpetrator inappropriately blaming the victimor publically disclosing details of their sex life); at first, theresearch team was challenged to understand how these relatedto the label chosen for the cluster, but upon further reflection,many of these items involve the use of very personal and inti-mate aspects of the relationship.

The final cluster, Psychological Abuse appears to sharesome thematic constructs with Emotional, but a deeper look

suggests that these behaviours imply a level of intimacy that isnot necessarily part of the statements Emotional. For example,emotionally blackmailing a victim by threatening suicide ordivorce, stalking, making a victim believe s/he is crazy or cannever do anything right, and controlling a victim’s appearanceare to some extent dependent on the emotional relationshipbetween the victim and the perpetrator; it is less likely thatsomeone in a casual relationship would have the power toblock a victim’s access to health care, sabotage a victim’shousework, or cause extreme duress through hurtful com-ments or criticism. The lack of clear distinction between someof these clusters may, in part, reflect the fact that the men’sdata were less cohesive than the data women had to work withgiven that the stress values for men’s data (0.34) were higherthan that for women (0.28).

Comparison of Cluster Contents and Orientationsby Gender

Although the maps created by men and women shared severalorganizational themes, noticeable differences emerged aswell. Both men and women grouped the items that describedphysical and sexual violence into their own clusters andlabeled them as such, but each group took different ap-proaches to the organization of non-physical and victim-driven actions. Even when men and women used similarlabeling language for clusters, the statements included ineach cluster were substantially different. For example, womencreated a cluster labeled simply as control, which includeditems that directly limited a victim’s actions or agency; theseitems were scattered among clusters in the men’s map. Forexample “perpetrator controlling victim’s appearance” was inthe women’s Control cluster but in the men’s PsychologicalAbuse cluster, and “perpetrator controlling victim’s important

Fig. 2 Concept map generatedfrom men’s group

372 J Fam Viol (2017) 32:367–382

Page 7: Conceptualization of Intimate Partner Violence: Exploring Gender … · 2018-06-05 · ORIGINAL ARTICLE Conceptualization of Intimate Partner Violence: Exploring Gender Differences

documents (passports, credit cards) was in women’s Controlcluster but in the men’s Family/Culture-Oriented Abuse clus-ter. Among men, although the map used the word controlwithin the title of one cluster, it was one of several words anddid not include the same items as found in the women’s controlcluster (e.g.,”perpetrator ignoring victim” and “perpetratorinappropriately blaming victim” were in women’s EmotionalAbuse cluster). Finally, women’s creation of a separate cate-gory for victim-driven actions and men’s creation of a uniquecluster for financial abuse represented the other substantialdifferences.

Ratings

To gain a deeper understanding of how men and women con-ceptualize the construct of IPV, participants were asked to rateeach of the 71 items on a scale of 1 to 5 (where 1 is the lowestscore and 5 the highest) according to three questions listed inthe methods section above. First, we calculated the averagescore for each item (on a scale of 1 to 5) by gender to deter-mine how important men and women believed each statementwas to the question (high, moderate, or low) (Table 2). Wealso analyzed the similarity of men and women’s scores ofimportance using go-zone plots, a graphical depiction ofthe scores indicating, in the shaded areas, where there ishigh agreement by gender (see Fig. 3 for an example of ago-zone plot).

In general, both men and women rated all but a few state-ments as moderately or highly important to all four ratingquestions, although the average for women’s scores tendedto be higher across the board. The only statement thathad low importance to both genders across all rating ques-tions was #16, “Victim ignoring perpetrator.” Additionally,although men rated item #65, “Victim criticizing perpetrator,”as moderately important to the construct of domestic violence(compared to women, who rated it as low), both men andwomen rated this statement as having low importance acrossthe remaining rating questions. The go-zone analysesalso illustrate high correlation between men’s and women’saverage statement rating (correlation coefficients range from0.84–0.89).

The Construct of Abuse Both men and women ratedphysically/sexually violent acts, threats, or outcomes as theitems most important to the meaning of relationship violence.Items 14 (“Perpetrator physical abusing victim”), 19(“Perpetrator infecting victim with sexually transmitted infec-tions”), 42 (“Perpetrator using a weapon to intimidate or scarevictim”), 60 (Perpetrator using a weapon to harm victim”),66 (“Perpetrator forcing victim into sexual acts”), and 68(“Perpetrator injuring victim’s breasts or genitals”) wereamong the statements with strong agreement and highscores across gender.

Although men and women were roughly in agreementabout the importance of physically violent behaviours to theconstruct of IPV, several items describing non-violent actions(or with non-physical ramifications) were among those withthe largest discrepancies between men’s and women’s rating.Items such as 4 (“Perpetrator inappropriately blamingvictim”), 7 (“Victim destroying perpetrator’s personalproperty”), 8 (“Perpetrator slanting cultural, religious,and moral values to encourage abuse of victim”), 12(“Perpetrator making victim feel crazy”), 21 (“Perpetratorusing aggressive behaviours intended to scare victim”), and61 (“Perpetrator cursing and name-calling victim”) wereamong those with the largest gaps between men’s andwomen’s average score, with women rating them as moreimportant to the construct compared to men. The correlationcoefficient between men’s and women’s ratings on this itemwas 0.99 (N = 63 due to some non-reporting of gender on ourdemographic survey).

Should the Victim Seek Professional Help? For the mostpart, both men and women considered the same physicalbehaviours that were central to the construct of IPV werealso ones that signal a need for professional help for avictim. Men and women differed the most on their opinionsabout the need for victims to seek help for several non-physical behaviours, most notably items 68 (“Perpetrator fre-quently becoming jealous of victim”) and 18 (“perpetratorstalking victim”), both of which women rated higher thanmen. The correlation coefficient between men’s and women’sratings on this item was 0.96 (N = 63).

Should the Perpetrator Seek Professional Help?Interestingly enough, both men and women tended to rateitems as more indicative of a perpetrator needing to seek pro-fessional help compared to a victim, and the go-zone showedfew areas of great difference. Item 11 (“perpetrator controllingand restricting daily activities of victim”) had the largest gapbetween men’s and women’s score (0.7), with women more,compared to men, reporting that this behavior warranted pro-fessional help for the perpetrator than men. The correlationcoefficient between men’s and women’s ratings on this itemwas 0.99 (N = 61).

Discussion

Through the use of concept mapping, this research highlightsboth similarities and differences in how men and womenperceive the construct of IPV. After a larger number ofmen and women generated a list of statements describingIPV and then categorized those statements as “conceptuallysimilar,” smaller gender-segregated subgroups were taskedwith determining if the clusters created by men and women

J Fam Viol (2017) 32:367–382 373

Page 8: Conceptualization of Intimate Partner Violence: Exploring Gender … · 2018-06-05 · ORIGINAL ARTICLE Conceptualization of Intimate Partner Violence: Exploring Gender Differences

Table 2 Final Brainstorm Statements about acts, behaviours and attitudes that comprise intimate partner violence with cluster and ratings results

Statement Women’s Cluster Men’s Cluster Level of Importance to Rating Questions

Def. of IPV Victim SeekHelp

Perpetra-torSeek Help

Stigma forVictim

1 Perpetrator forcing victim to workfor pay

Control Financial Abuse Mod M = Low Mod M = Low

W = Mod W = Mod

2 Perpetrator punishing victim onissues related to child gender(e.g., blaming women for nothaving boy child or forcingchild gender preference)

External & Cultural Family/Cultural Abuse M = Mod Mod M = Mod ModW = High W = High

3 Perpetrator abusing victim as aresult of a gambling addiction

External & Cultural Physical Violence M = Mod Mod High ModW = High

4 Perpetrator inappropriatelyblaming victim

Emotional Abuse Control/Per. M = Low M = Low Mod M = Low

Intimate/Psyc W = Mod W = Mod W = Mod

5 Perpetrator punishing victim fornot having sex

Sexual Abuse Sexual Abuse High Mod M = Mod ModW = High

6 Perpetrator insisting on a dowryfrom victim or victim’s familyprior to or during marriage

External & Cultural Financial Abuse Mod Mod Mod Mod

7 Victim destroying perpetratorspersonal property

Victim Response Physical Violence Mod Mod Mod M = Moda

W = Low

8 Perpetrator slanting cultural,religious and moral values toencourage abuse of victim

External & Cultural Family/Cultural Abuse M = Mod Mod High ModW = High

9 Perpetrator forcibly sleep deprivingvictim

Physical Abuse Physical Violence M = Mod M = Mod High ModW = High W = High

10 Perpetrator manipulating and lyingto victim

Manipula-tion Family/Cultural Abuse Mod Mod M = Mod M = Low

W = High W = Mod

11 Perpetrator controlling andrestricting daily activitiesof victim (e.g. when to dogrocery shopping, haircut,banking)

Control Financial Abuse M = Mod Mod M = Mod ModW = High W = High

12 Perpetrator making victim feel thatthey are crazy

Emotional Abuse Psychological Abuse M = Mod Mod M = Mod Men = Mod

W = High W = High W = High

13 Perpetrator interfering or blockingvictim’s access to health careproviders

Physical Abuse Psychological Abuse High High High Mod

14 Perpetrator physically abusingvictim (e.g.: beating, slapping,pushing, spitting)

Physical Abuse Physical Violence High High High High

15 Perpetrator controlling victim’scommunications (e.g. emailsand phone calls)

Control Family/Cultural Abuse M = Mod Mod M = Mod M = Mod

W = High W = High W = High

16 Victim ignoring perpetrator Victim Response Control/Per. Low Low Low LowIntimate/Psyc

17 Perpetrator encouraging family/friends to engage in abusivebehaviours/language towardsvictim

Manipula-tion Psychological Abuse M = Mod M = Mod M = Mod M = Mod

W = High W = High W = High W = High

18 Perpetrator stalking victim Control Psychological Abuse M = Mod M = Mod M = Mod M = Mod

W = High W = High W = High W = High

19 Perpetrator infecting victim withsexually transmitted infections

Sexual Abuse Physical Violence High High High High

20 Perpetrator using their culturalvalues to excuse abuse orviolence

External & Cultural Family/Cultural Abuse M = Mod Mod M = Mod M = Mod

W = High W = High W = High

374 J Fam Viol (2017) 32:367–382

Page 9: Conceptualization of Intimate Partner Violence: Exploring Gender … · 2018-06-05 · ORIGINAL ARTICLE Conceptualization of Intimate Partner Violence: Exploring Gender Differences

Table 2 (continued)

Statement Women’s Cluster Men’s Cluster Level of Importance to Rating Questions

Def. of IPV Victim SeekHelp

Perpetra-torSeek Help

Stigma forVictim

21 Perpetrator using aggressivebehaviours intended to scarevictim (e.g.: punching a wall)

Physical Abuse Emotional M = Mod M = Mod High M = Mod

W = High W = High W = High

22 Perpetrator publically denying anywrongdoing toward victim (e.g.:in front of family, friends etc.)

Emotional Abuse Family/Cultural Abuse Mod Mod Mod M = Mod

W = High

23 Perpetrator keeping victim fromenjoying sex

Sexual Abuse Sexual Abuse Mod Mod M = Mod ModW = High

24 Perpetrator making unwantedsexually explicit comments tovictim

Sexual Abuse Sexual Abuse Mod Mod M = Mod M = Mod

W = High W = High

25 Victim provoking perpetrator touse violence

Victim Response Sexual Abuse Mod Mod Mod Mod

26 Perpetrator publically humiliatingvictim

Emotional Abuse Emotional M = Mod Mod M = Mod M = Mod

W = High W = High W = High

27 Perpetrator neglecting victimwhen they are sick

Physical Abuse Emotional Mod Mod Mod Mod

28 Perpetrator allowing externalparties (e.g., business colleagues,extended family) to make orinfluence major family decisions(e.g., marriage, economicconcerns) against victim’s wishes

Control Family/Cultural Abuse Mod Mod M = Mod M = Mod

W = High W = High

29 Perpetrator destroying victim’spersonal property

Physical Abuse Physical Violence M = Mod M = Mod High M = Mod

W = High W = High W = High

30 Perpetrator screaming and yellingat victim

Emotional Abuse Emotional Mod Mod M = Mod ModW = High

31 Perpetrator abusing victim as aresult of victim’s mental illness(e.g.: depression, mood disorder)

External & Cultural Emotional High High High M = Mod

W = High

32 Perpetrator preventing victim fromseeing a healthcare provider ofthe opposite gender

Control Family/Cultural Abuse Mod Mod M = Mod ModW = High

33 Perpetrator abusing victim as aresult of alcohol and/or drug use

External & Cultural Emotional High M = Mod High M = Mod

W = High W = High

34 Perpetrator controlling victim’ssocial contact (e.g.: victimcannot visit friends)

Control Family/Cultural Abuse M = Mod M = Mod M = Mod M = Mod

W = High W = High W = High W = High

35 Perpetrator keeping victim andchildren separated

Control Family/Cultural Abuse High M = Mod High HighW = High

36 Perpetrator making hurtfulcomments about physicalappearance of victim

Emotional Abuse Psychological Abuse Mod Mod Mod M = Mod

W = High

37 Victim screaming and yelling atperpetrator

Victim Response Sexual Abuse Mod Mod M = Low M = Low

W = Mod W = Mod

38 Perpetrator making sexist andracist remarks about victim

Emotional Abuse Family/Cultural Abuse Mod Mod M = Mod M = Mod

W = High W = High

39 Perpetrator turning other people(e.g.: children, family, friends)against victim

Emotional Abuse Family/Cultural Abuse M = Mod M = Mod M = Mod M = Mod

W = High W = High W = High W = High

40 Victim abusing perpetrator inresponse to abuse

Victim Response Physical Violence Mod Mod Mod M = Low

W = Mod

41 Perpetrator controlling victim’simmigration activities(e.g.: threatening deportation)

Control Family/Cultural Abuse M = Mod High M = Mod M = Mod

W = High W = High W = High

J Fam Viol (2017) 32:367–382 375

Page 10: Conceptualization of Intimate Partner Violence: Exploring Gender … · 2018-06-05 · ORIGINAL ARTICLE Conceptualization of Intimate Partner Violence: Exploring Gender Differences

Table 2 (continued)

Statement Women’s Cluster Men’s Cluster Level of Importance to Rating Questions

Def. of IPV Victim SeekHelp

Perpetra-torSeek Help

Stigma forVictim

42 Perpetrator using a weapon tointimidate or scare victim (e.g.knife, baseball bat)

Physical Abuse Physical Violence High High High High

43 Perpetrator sabotaging victim’shousework (e.g. deliberately noteating home cooked meal)

Manipula-tion Psychological Abuse Mod Mod Mod Mod

44 Perpetrator making victim feelthey can never do anything rightor are ever good enough

Manipula-tion Psychological Abuse Mod Mod M = Mod M = Mod

W = High W = High

45 Perpetrator publically disclosingdetails of sex life with victim toshow their power

Emotional Abuse Control/Per. M = Mod Mod M = Mod M = Mod

Intimate/Psyc W = High W = High W = High

46 Perpetrator ignoring victim Emotional Abuse Control/Per. Mod Low Mod M = Low

Intimate/Psyc W = Mod

47 Perpetrator abusing victim as a resultof perpetrators mental illness(e.g.: depression, mood disorder)

External & Cultural Emotional M = Mod M = Mod M = Mod M = Mod

W = High W = High W = High W = High

48 Perpetrator controlling victim’simportant documents (e.g.:passport, credit cards)

Control Family/Cultural Abuse M = Mod M = Mod High M = Mod

W = High W = High W = High

49 Perpetrator denying victim orchildren basic necessities(e.g.: deny children clothing)

Physical Abuse Family/Cultural Abuse High High High High

50 Perpetrator controlling andrestricting family finances

Control Financial Abuse Mod Mod M = Mod ModW = High

51 Perpetrator emotionallyblackmailing victim(e.g.: perpetrator threateningsuicide or divorce)

Manipula-tion Psychological Abuse M = Mod M = Mod M = Mod M = Mod

W = High W = High W = High W = High

52 Perpetrator denying to the victimany wrongdoing within theirrelationship

Manipula-tion Family/Cultural Abuse Mod Mod Mod Mod

53 Perpetrator encouraging childrento take part in violence towardsvictim (e.g., encouraging kids toact dismissive and demeaningtoward the victim)

Manipula-tion Emotional High High High High

54 Perpetrator maintaining a secretlifestyle and/or withholdinginformation about lifestyle fromvictim

Manipula-tion Family/Cultural Abuse Mod Mod M = Mod ModW = High

55 Perpetrator demonstrating publicdisplays of power over victim(e.g.: silencing, grabbing victimin public)

Physical Abuse Emotional M = Mod M = Mod M = Mod M = Mod

W = High W = High W = High W = High

56 Perpetrator controlling victim’sphysical appearance (e.g. victimtold what to wear)

Control Psychological Abuse Mod Mod Mod Mod

57 Perpetrator imposing religiousbeliefs on victim and children

External & Cultural Family/Cultural Abuse Mod Mod Mod M = Mod

W = High

58 Perpetrator criticizing victim (e.g.:bullying, belittling, demeaning,humiliating, ridiculing, constantput downs)

Emotional Abuse Psychological Abuse M = Mod M = Mod M = Mod M = Mod

W = High W = High W = High W = High

59 Perpetrator controlling sexualactivity with victim (e.g.: howoften, contraception)

Sexual Abuse Sexual Abuse M = Mod M = Mod M = Mod M = Mod

W = High W = High W = High W = High

376 J Fam Viol (2017) 32:367–382

Page 11: Conceptualization of Intimate Partner Violence: Exploring Gender … · 2018-06-05 · ORIGINAL ARTICLE Conceptualization of Intimate Partner Violence: Exploring Gender Differences

made sense and if so, how they should be labeled. While bothgenders tended to agree on the specific behaviours that com-prise physical and/or sexual forms of IPV, when asked toconsider statements that describe less physically aggressivebehaviours, participants constructed very different conceptualmaps, suggesting that each group views these behavioursthrough different lenses. This is consistent with Carlson andWorden’s (2005) findings that men and women generallyagreed that physical aggression should be labeled as domesticviolence, but had substantially less consensus about non-physical acts.

One of the most noticeable distinctions can be seen in howwomen and men viewed the use of non-physical behaviorsthat restrict a victim’s personal autonomy by dictating whats/he can or cannot do (e.g., controlling the victim’s socialcontacts, keeping the victim and children separated, blockingaccess to education or work). Women tended to cluster thesestatements together under the classification of control, where-as men labeled most of these same statements as either family/cultural-oriented abuse or financial abuse (e.g., Perpetratorcontrolling victim’s communications or Perpetrator allowing

external parties to make or influence major family decisionsagainst victim’s wishes). One interpretation of this distinctionis that when men viewed these behaviors, they focused on theabusive strategy (such as using family members or usingmoney to control the victim) rather than of the result orintent of the act (to control the victim). Women, on theother hand, considered the purpose of the behaviors: control,manipulation, etc.

While it is possible that men did not consider the behaviorsto be controlling, this distinction is consistent with what hasbeen shown for physical IPV, from the perspective of theperpetrator as noted earlier. Traditionally, men in numerouscultures are socialized to equate masculinity with controland aggression, and even those who staunchly reject theacceptability of violence against women may find them-selves more prone to think about such behaviors from theviewpoint of the actor, rather than the recipient. Women,on the other hand, viewed these same statements from theperspective of the impact they had on a victim; additionally,women separated victim-perpetrated acts from the rest of thebehaviors, suggesting that these were somehow different in

Table 2 (continued)

Statement Women’s Cluster Men’s Cluster Level of Importance to Rating Questions

Def. of IPV Victim SeekHelp

Perpetra-torSeek Help

Stigma forVictim

60 Perpetrator using a weapon toharm victim

Physical Abuse Physical Violence High High High M = Mod

W = High

61 Perpetrator cursing and namecalling victim

Emotional Abuse Psychological Abuse M = Mod Mod Mod M = Mod

W = High W = High

62 Perpetrator forcing victim toconsume alcohol and/or drugs

Physical Abuse Physical Violence High High High M = Mod

W = High

63 Perpetrator making scenes that putdown victim at social events

Manipula-tion Emotional Mod Mod M = Mod M = Mod

W = High W = High

64 Perpetrator restricting or blockingvictim’s access to education orwork

Control Financial Abuse M = Mod M = Mod M = Mod M = Mod

W = High W = High W = High W = High

65 Victim criticizing perpetrator Victim Response Control/Per. M = Mod Low Low LowIntimate/Psyc W = Low

66 Perpetrator forcing victim intosexual acts (e.g. unwanted:sodomy, viewing pornography,oral sex, sex with animals)

Sexual Abuse Sexual Abuse High High High High

67 Perpetrator treating victim likethey are their own personalservant

Control Emotional High M = Mod M = Mod M = Mod

W = High W = High W = High

68 Perpetrator frequently becomingjealous of victim

Manipula-tion Control/Per. Mod M = Low Mod M = Low

W = Mod W = ModIntimate/Psyc

69 Perpetrator injuring victim’sbreasts or genitals

Sexual Abuse Sexual Abuse High High High High

70 Perpetrator accusing victim ofhaving an affair

Emotional Abuse Control/Per. Mod Mod M = Mod ModW = HighIntimate/Psyc

71 Perpetrator making victim cry Emotional Abuse Emotional Mod Mod Mod Mod

J Fam Viol (2017) 32:367–382 377

Page 12: Conceptualization of Intimate Partner Violence: Exploring Gender … · 2018-06-05 · ORIGINAL ARTICLE Conceptualization of Intimate Partner Violence: Exploring Gender Differences

nature. Our findings are somewhat consistent with studies intwo diverse literatures: psychology of anger and aggressionand feminist economics. For the former, men were found torepresent aggression “as a matter of management” of behav-iors and not “as a breakdown in control.”Whereas for women,notions of control were central to motivations of anger andaggression (Campbell & Muncer 1987). Alternatively, fromthe feminist economics literature, it has been noted thatwomen’s traditional lack of equal access to and control oversocietal resources, imposed by what, Folbre (1994) calls‘structures of constraint’ has had greater negative impacts up-on women than men (Folbre 1994). Thus, given their histori-cal position, women have a heightened awareness of issuesrelated to control and domination. This distinction in howmenand women view non-physical violence can also be seen inhowmen and women rated the items’ importance to the mean-ing of relationship abuse. Both men and women rated theitems describing physical and sexual violence as being highlyimportant to the construct of IPV. It is not likely that thosedistinctions can be explained by the differential compositionof the two samples in terms of marital status, age or income asprior analyses on these data indicated that there was littlevariation in rating items’ importance to the meaning of rela-tionship abuse based upon those particular demographic

characteristics (Author reference). However, men tended torate non-physical behaviors related to control as less importantthan women. Similar gender-related perspectives have beenseen in other research, as well. Stanley and colleagues foundmale focus group participants hesitant to accept a “broad def-inition” of domestic violence, in part out of concern that ordi-nary arguments could be construed as violent (Stanley et al.2012). After finding that male college students rated a list ofsimilar behaviors as less controlling than female students,Ehrensaft and Vivian (1999) suggested “it may be that menview these behaviours as less controlling because they haveminimal impact upon them, whereas for women, theirpartner’s attempts to control them may be more successful”(p. 262). One challenge with these types of studies, includingours, where men and women are asked to weigh in on or ratethe same questions about violent behaviours, is that it is notclear whether men’s reports are that of a perpetrator or ofbeing male, and similarly, whether women are providing theperspective of being a victim or of a woman. Thus, if the menand women in our study were reporting more as perpetratorsand victims rather thanmen and women, it is possible that truegender differences are smaller than what we report here. Thisis an area to investigate in future research as we were not ableto, from the data that we collected, tease out these nuanced

83.2 4.713.05

4.88

Men

Wom

en

4.25

8

11

18

20

24

2630

3245

50

57

58

1

5

5

5

71

r = .84

Fig. 3 Go-Zone plot illustrating men’s vs. women’s ratings on the importance that victim’s seek help

378 J Fam Viol (2017) 32:367–382

Page 13: Conceptualization of Intimate Partner Violence: Exploring Gender … · 2018-06-05 · ORIGINAL ARTICLE Conceptualization of Intimate Partner Violence: Exploring Gender Differences

differences. Finally, it is interesting to note that in developingand labeling their conceptualizations of IPV, both men andwomen highlighted several elements that Stark (2007) usedto describe coercion (physical and sexual force or threats), andcontrol (the structured use of manipulation, and the exploita-tion of both internal (emotional) and external (cultural) influ-ences). However, women emphasized the intention of thesebehaviors, and thus more clearly addressed the notion thatthese actions are intended to demand obedience and prescriptthe recipient’s activities.

The gender differences observed in this study may haveimportant implications for the prevention and interventionon IPV. With regards to reporting IPV, for example, inter-couple agreement of IPV has been moderate in the literature(Chan 2011), and women have generally been described asmore likely to disclose IPV and are more likely to seek helpthan men (Hamby 2014; Tsui 2014). There are a number ofhypothesized explanations for this discrepancy: social desir-ability, which has been related to underreporting by men ofboth perpetration and experiences of non-physical forms ofviolence (Arnocky and Vaillancourt 2014; Freeman et al.2014); adherence to of males to ideologies of masculinityand power, which has been related to decreased help seeking(Wu et al. 2011); and insufficient measurement of context,motives, and consequences of IPV by existing measures,which is a source of bias for detection of IPV (Chan 2011).However, our findings support a complementary hypothesis:that these differences arise from gendered conceptualizationsof IPV – what constitutes IPV, what concepts are most centralto the definition, and in particular, the role of control andcoercion. We also found that women are more likely thanmen to endorse the need for help seeking for both perpetratorsand victims in relation to controlling and jealous behaviours.These findings emphasize the need for policy makers to attendto gendered representations of IPV in screening, reporting,help seeking, and importantly, education. For example, whileuniversal screening of IPV continues to be important for pre-vention and clinic-based intervention, small changes to thewording of screening tools has affected the relative detectionof IPV in men and women (Hamby 2014; Hamby 2014).Whereas men have been found to report lower rates of IPVwith survey items that ask them about IPV in context withnon-intimate violence than with items asking about IPV inisolation, women have reported higher rates of IPV, implyingthat IPV may be more salient than other forms of violence forwomen than for men (Hamby and Turner 2012). Screeningstrategies that explicitly acknowledge gender-biases in the un-derstanding of IPV may therefore improve the reporting ofIPV. At the very least, the reporting of IPV may be improvedby better understanding the different ways that people concep-tualize this concept.

Our findings add to the small but growing literature ongender differences in the conceptualization of IPVand at best

can begin to generate some hypotheses about where thesedifferences are the greatest. Further research is needed refinethe conceptual domains identified in our study. A multi-method approach with large sample could enhance the gener-alizability of these findings. One potential limitation inherentin this inquiry is the gender of the research team being pre-dominately female, with one exception (MY). In an effort toaddress this bias, we discussed the conversation that tookplace among men during the participatory mapping exercise.In addition to confirming our assumption that men interpretedthe term “perpetrator” asmale and “victim” as female, themenwho participated in that process openly addressed the state-ments as they related to the roles that they, as men, wereexpected to adhere: financial planners, decision-makers, andauthoritarian figures. Men are forced to play a balancing act,conforming to traditional images of masculinity without abus-ing the power that gave them; thus, these actions represented afailure to achieve that balance. Given this context, severalelements of the men’s map make sense, especially the separa-tion of “financial abuse” (an arena of traditional male control)from other non-physical behaviors and the focus on the formthat the abuse takes rather than on the results. A related limi-tation is that some of our findings around the labeling of clus-ters may reflect the use of language within the group settingand not necessarily reflect underlying gender differences.While research process could not enable us to explore thispossibility, however, this should be an area pursued for futureresearch.

One of the strongest elements of this investigation is the useof the concept mapping process is its inclusion of multipleapproaches to research (qualitative and quantitative, individu-al and group input); that said, it is also subject to the biasesinherent in each method. One limitation of this study is thatwhile our sample of 71 participants was larger than the aver-age concept mapping sample of between 30 to 50 participants,as is the case with many qualitative projects, it is based on aconvenience sample. In addition to the general concerns aboutbias that convenience samples bring, the high proportion ofparticipants with post-secondary education and with East andSouth Asian backgrounds and the relatively small sample ofmen and women who identified as Aboriginal may haveshaped the findings. Furthermore, the use of a smaller, self-selected group of participants in the participatory mappingprocess may have resulted in final cluster maps that did notfully reflect the intentions of the larger groups of men andwomen; to minimize this risk, and facilitators with experiencein concept mapping led the final participatory mapping pro-cess as a way of ensuring that a wide range of participantsviews were represented. Demographic differences betweenmen and women were noted in our sample with women beingyounger and more likely to be single than men. These demo-graphic differences by gender may have impacted the sortingof our data but we have no way assessing whether this is the

J Fam Viol (2017) 32:367–382 379

Page 14: Conceptualization of Intimate Partner Violence: Exploring Gender … · 2018-06-05 · ORIGINAL ARTICLE Conceptualization of Intimate Partner Violence: Exploring Gender Differences

case. However, this did not seem to impact the fourareas of rating as correlations by gender were all foundto be close to 1.0.

Our study also highlights the importance of involvingmen’s voices and perceptions in research on conceptualizingand defining IPV. Traditional notions of masculinity suggestthat men should be strong, aggressive, and in control (O’Neil1981); these images conflict with the idea of victimization,and may serve as a barrier to participation in this topic areaof research if men believe they will be asked about their ex-periences of abuse or assault (Drijber, Reijnders & Ceelen,2013; Tsui, Monit et al., 2010). On the other hand, becausemost IPVresearch that has includedmale participants focus onthem as perpetrators rather than as recipients or even con-cerned citizens, “everyday” men may believe that their viewswill be discounted, distorted, or met with hostility. We hadhoped to have a larger sample of men but it was particularlychallenging to recruit them into our study. Similarly, Velonis(2013), despite the involvement of male researchers, had a lessthan 10 % positive response rate among men asked to partic-ipate in interviews as a follow-up to a quantitative study; thiscompared to a 65 % positive response rate among similarwomen. Concept mapping may be an effective approach, asit allows men to express their beliefs and opinions anony-mously as well as in a controlled group setting with othermen, rather than in one-on-one interviews.

In spite of this limitation, the participants in this studyincluded a level of diversity not often seen in IPV research,and the inclusion of men in both the individual and groupphases is an additional strength. Concept mapping is a partic-ipatory research process, allowing the participants to providesubstantial input into the findings and reducing the potentialfor researcher-driven biases and misinterpretations. Finally,the research team itself constituted a multidisciplinary set ofinvestigators, all of whom brought unique ideas and interpre-tations to the process. Although several members had exten-sive experience with IPV research, others brought clinical,psychological, and sociological backgrounds to the table,and significant effort was made to be inclusive of all typesof expertise.

Our findings have relevance to the design of health andlegal interventions and programs. First, there is a need to en-sure that a broad range of acts, behaviors, and attitudes areincluded when identifying perpetrators and victims of IPVandnot just an emphasis on physical abuse. To close the gendergap in understanding the nature of IPV, especially acts andbehaviours that comprise non-physical aggression, there maybe a need for more general education and awareness-raisingabout specific domains of violence, in particular, controllingbehaviors and the central role they play in violent or abusiverelationships. This could take place in training efforts for pro-fessionals and service providers involved in violence preven-tion and treatment programs broadly (e.g., social workers, law

enforcement, health care) as well as general educational cam-paigns for the general public about what constitutes IPV (e.g.,schools) by providing a gender analysis to not only the im-pacts of IPV on men and women but also on the differentialconceptualization of IPV by gender. Moreover, in such train-ings, gender-sensitive techniques are needed to engage menand women for IPV concerns. Developing such sensitivity ishighly pertinent considering the social stigma attached to IPVand communication difficulties reported by patients and pro-fessionals (Kimberg 2008; Ghandour et al. 2015). In 2013, areview of training programs by theWorld Health Organizationfound inadequate attention is placed on communication andclinical skills for front line professionals (WHO 2013).

Conclusions

Although much of this paper has focused on the differencesbetween men and women’s conceptualizations of IPV, therewere numerous similarities, as well. Both men and womenbelieved that physical and sexual violence are central to theconcept of relationship violence and are indications that vic-tims and perpetrators need to seek help. Moreover, both gen-ders agreed that non-physically abusive behaviors were atleast moderately important to the construct of relationshipviolence, suggesting that two decades of efforts by domesticviolence advocates across North America to expand our think-ing about relationship abuse has had an effect. Finally, as weaddressed in a previous article (Author reference), these find-ings speak to the need to include variables that reflect thislarger concept of IPV in our population-level surveillance,our academic research, and our clinical screening processes.

Through our use of concept mapping, we have contributedto the small but growing body of research on the conceptual-ization of IPV. Specifically, we explore how gender influencesour perceptions of what IPV is and of the role that non-physically violent but controlling behaviors play in relation-ship conflict. Concept mapping is an underused methodologyin this field, but is particularly useful for this investigation inthat it allows participants themselves to generate the state-ments used to describe IPV, to determine the relationshipsbetween specific behaviors, and then to interpret the results.This approach provides a deeper understanding of how menand women outside of the domestic violence advocacy andresearch community view IPVand the behaviors they believecomprise this problem.

The findings from this semi-qualitative investigation arepreliminary, but generate numerous questions and hypothesesthat deserve further attention not only in larger research stud-ies but also in the design of programs and interventions. Forexample, if men perceive controlling behaviors as less-centralto their understanding of IPV, why? Is physical violence easierto recognize than controlling behaviors, or do men accept

380 J Fam Viol (2017) 32:367–382

Page 15: Conceptualization of Intimate Partner Violence: Exploring Gender … · 2018-06-05 · ORIGINAL ARTICLE Conceptualization of Intimate Partner Violence: Exploring Gender Differences

these behaviors as a normal part of interpersonal relation-ships? Another area ripe for further exploration is whetherthe perceived gender of the “perpetrator” and “victim” influ-ences men’s and/or women’s understanding of IPV; this infor-mation could be particularly useful in the design of battererintervention programs, victim services, and primary preven-tion efforts.

References

Alhabib, S., Nur, U., & Jones, R. (2010). Domestic violence againstwomen: Systematic review of prevalence studies. Journal ofFamily Violence, 25, 369–382.

Anderson, K. L., & Umberson, D. (2001). Gendering violence:Masulinity and power in Men’s accounts of domestic violence.Gender & Society, 15(3), 358–380.

Ansara, D. L., & Hindin, M. J. (2010). Exploring gender differences inthe patterns of intimate partner violence in Canada: A latent classapproach. Journal of Epidemiology and Community Health, 64(10),849–854. doi:10.1136/jech.2009.095208.

Arnocky, S., & Vaillancourt, T. (2014). Sex differences in response tovictimization by an intimate partner: More stigmatization and lesshelp-seeking among males. Journal of Aggression, Maltreatment &Trauma, 23(7), 705–724. doi:10.1080/10926771.2014.933465.

Beyers, J., Leonard, J., Mays, V., & Rosen, L. (2000). Gender differencesin the perception of courtship abuse. Journal of InterpersonalViolence, 15(5), 451–466.

Black, M. C., Basile, K. C., Breiding, M. J., Smith, S. G., Walters, M. L.,Merrick, M. T., et al. (2011). The National Intimate Partner andsexual violence survey (NISVS): 2010 summary report. Atlanta:National Center for Injury Prevention and Control, Centers forDisease Control and Prevention.

Bonomi, A. E., Thompson, R. S., Anderson, M., Reid, R. J., Carrell, D.,Dimer, J. A., et al. (2006). Intimate partner violence and Women’sphysical, mental, and social functioning. American Journal ofPreventive Medicine, 30(6), 458–466.

Burke, J., P. O'Campo, G. Peak, A. Gielen, K. McDonnell and W.Trochim (2005). “An introduction to concept mapping as a partici-patory putlic health research method.” Qualitative Health Research,15(10), 1392–1410.

Campbell, A. and S. Muncer (1987). “Models of anger and aggression inthe social talk of women and men.” Journal for the Theory of SocialBehaviour 17(4), 489–511.

Carlson, B. E., & Worden, A. P. (2005). Attitudes and beliefs aboutdomestic violence: Results of a public opinion survey: I. Definitionsof domestic violence, criminal domestic violence, and prevalence.Journal of Interpersonal Violence, 20(10), 1197–1218.

Chan, K. L. (2011). Gender differences in self-reports of intimate partnerviolence: A review. Aggression and Violent Behavior, 16(2), 167–175. doi:10.1016/j.avb.2011.02.008.

Cho, H. (2012). Examining gender differences in the nature and contextof intimate partner violence. Journal of Interpersonal Violence,27(13), 2665–2684. doi:10.1177/0886260512436391.

Dennison S., Thompson C. (2011) Intimate partner violence: the effect ofgender and contextual factors on community perceptions of harm,and suggested victim and criminal justice responses. Violence andVictims, 26,(3), 347–363.

Drijber, B. C., Reijnders, U. J., & Ceelen, M. (2013). Male victims ofdomestic violence. Journal of Family Violence, 28(1), 173–178. doi:10.1007/s10896-012-9482-9

Dutton, D. G. (2012). The case against the role of gender in intimatepartner violence. Aggression and Violent Behavior, 17(1), 99–104.doi:10.1016/j.avb.2011.09.002.

Ehrensaft, M. K., & Vivian, D. (1999). Is partner aggression related toappraisals of coercive control by a partner? Journal of FamilyViolence, 14(3), 251–266.

Flynn, A., & Graham, K. (2010). “why did it happen?” a review andconceptual framework for research on perpetrators’ and victims’explanations for intimate partner violence. Aggression and ViolentBehavior, 14(3), 239–251.

Folbre, N. (1994).Who takes Care of the Kids? Gender and the structuresof constraint. London: Routledge.

Freeman, A. J., Schumacher, J. A., &Coffey, S. F. (2014). Social desirabilityand partner agreement of Men’s reporting of intimate partner violencein substance abuse treatment settings. Journal of InterpersonalViolence, 30(4), 565–579. doi:10.1177/0886260514535263.

Galvin, P.F. (1989). Concept mapping for planning and evaluation of aBig Brother/Big Sister program. Evaluation and Program Planning,12(1), 53–58.

Ghandour, R., Camplbell, J., & Lloyd, J. (2015). Screening and counsel-ing for IPV: A vision for the future. Journal of Women’s Health,24(1), 57–61.

Hamby, S. (2014). Psychology of violence self-report measures that donot produce gender parity in intimate partner violence: A multi-study investigation self-report measures that do not produce genderparity in intimate partner violence: A multi-study investigation.Psychology of Violence. Oct.

Hamby, S., & Turner, H. (2012). Measuring teen dating violence in malesand females: Insights from the National Survey of Children’s expo-sure to violence. Psychology of Violence, 3(4), 323–339. doi:10.1037/a0029706.

Harris, R. J., & Cook, C. A. (1994). Attributions about spouse abuse: Itmatters who the batterers and victims are. Sex Roles, 30, 553–565.

Harris, R., Firestone, J., & Vega, W. (2005). The interaction of country oforigin, acculturation and gender role ideology on wife abuse. SocialScience Quarterly, 86(2), 465–483.

Heise, L. (1998). Violence against women: An integrated, ecologicalframework, violence against women. June, 4(3), 262–283.

Johnson, M. P. (2008). A typology of domestic violence: intimate terror-ism, violent resistance, and situational couple violence. Boston:Northeastern University Press.

Johnson, M. P. (2011). Gender and types of intimate partner violence: Aresponse to an anti-feminist literature review. Aggression andViolent Behavior, 16(4), 289–296. doi:10.1016/j.avb.2011.04.006.

Johnson, M. P., & Leone, J. M. (2005). The differential effects of intimateterrorism and situational couple violence: Findings from theNational Violence against Women Survey. Journal of FamilyIssues, 26(3), 322–349. doi:10.1177/0192513x04270345.

Kane, M., & Trochim, W. M. (2007). Concept mapping for planning andevaluation. Thousand Oaks: Sage.

Kimberg, L. (2008). Addressing intimate partner violence with male pa-tients: A review and introduction of pilot guidelines. Journal ofGeneral Internal Medicine, 23(12), 2071–2078.

Krug, E. G., Dahlberg, L., Mercy, J. A., Zwi, A. B., & Lozano, R. (2002).World report on violence and health. Geneva: World HealthOrganization.

Langhinrichsen-Rohling, J. (2010). Controversies involving gender andintimate partner violence in the United States. Sex Roles, 62(3–4),179–193. doi:10.1007/s11199-009-9628-2.

Makepeace, J. (1986). Gender differences in courtship violence victimi-zation. Family Relations, 35(3), 383–388.

Muehlenhard, C., & Kimes, L. (1999). The social construction of vio-lence: The case of sexual and domestic violence. Personality andSocial Psychology Review, 3(3), 234–245.

Murnen, S. K., Wright, C., & Kaluzny, G. (2002). If boys will be boys,then girls will be victims? A meta-analytic review of the research

J Fam Viol (2017) 32:367–382 381

Page 16: Conceptualization of Intimate Partner Violence: Exploring Gender … · 2018-06-05 · ORIGINAL ARTICLE Conceptualization of Intimate Partner Violence: Exploring Gender Differences

that relates masculine ideology to sexual aggression. Sex Roles,46(11–12), 359–375.

O’Neil, J. M. (1981). Patterns of gender role conflict adn strain: Sexismand fear of Feminity in Men’s lives. The Personnel and GuidanceJournal, 60(4), 203–210.

O’Neil, James, and Michel Harway. (1997). A multivariate modelexplaining men’s violence against women. Violence AgainstWomen, 3(2), 182–203.

Pease, B., & Flood, M. (2008). Rethinking the significance of attitudes inpreventing men’s violence against women. Australian Journal ofSocial Issues, 43(4), 547–561.

Seelau, S., & Seelau, E. (2005). Gender-role stereotypes and perceptionsof heterosexual, gay and lesbian domestic violence. Journal ofFamily Violence, 20(6), 363–371.

Sorenson, S. B., & Taylor, C. A. (2005). Female agression toward maleintimate partners: An examination of social norms in a community-based sample. Psychology of Women Quarterly, 29(1), 78–96.

Stanley, N., Fell, B., Miller, P., Thomson, G., &Watson, J. (2012). Men’stalk: Men’s understanding of violence against women and motiva-tions for change. Violence Against Women, 18(11), 1300–1318.

Stark, E. (2007). Coercive control: how men entrap women in personallife. New York: Oxford University Press.

Stewart-Williams, S. (2002). Gender, the perception of aggression, andthe overestimation of gender bias. Sex Roles, 46, 177–189.

Tanha, M., Beck, C. J. A., Figueredo, A. J., & Raghavan, C. (2010).Sex differences in intimate partner violence and the use of

coercive control as a motivational factor for intimate partner vi-olence. Journal of Interpersonal Violence, 25(10), 1836–1854.doi:10.1177/0886260509354501.

Trochim, W. (1985). Pattern matching, validity, and conceptualization inprogram evaluation. Evaluation Review, 9, 575–604.

Trochim, W. (1989). “An introduction to concept mapping for programplanning and evaluation.”Evaluation and ProgramPlanning ,12,1–16.

Tsui, V. (2014). Male victims of intimate partner abuse: Use andhelpfulness of services. Social Work, 59(2), 121–130. doi:10.1093/sw/swu007.

Tsui, V., Cheung, M., Leung, P. (2010). Help-seeking among male vic-tims of partner abuse: men's hard times. Journal of CommunityPsychology, 38,(6), 1520–6629

Velonis, A. J. (2013). Are All Hits the Same? AMixed Methods Investigationof Gender and Violence in Heterosexual relationships. (Ph.D.),University of ColoradoDenver, Denver, CO. (UMINumber: 3562673).

World Health Organization. (2013) Responding to intimate partner vio-lence and sexual violence against women: WHO clinical and policyguidelines. Available: http://apps.who.int/iris/bitstream/10665/85240/1/9789241548595_eng.pdf?ua=1.

Wu, E., El-Bassel, N., Gilbert, L., O’Connor, M., & Seewald, R. (2011).Traditional male ideology and service system involvement amongdrug-involved men who perpetrate intimate partner violence: alongitudinal study. Journal of Interpersonal Violence, 26(8),1646–1663. doi:10.1177/0886260510370594.

382 J Fam Viol (2017) 32:367–382