International Journal of Clinical and Health Psychology 22 (2022) 100276

International Journal
of Clinical and Health Psychology
[Link]/ijchp

ORIGINAL ARTICLE

Dating Violence Questionnaire for Victimization and

Perpetration (DVQ-VP): An Interdependence Analysis of
Self-Reports
Luis Rodríguez-Francoa, Joel Juarros-Basterretxeab,*, Susana Paíno-Quesadac,
Juan Herrerod, Francisco Javier Rodríguez-Díazd

a
Departamento de Personalidad, Evaluacion y Tratamiento Psicologicos, Universidad de Sevilla, Spain
b
Departamento de Psicología y Sociología, Universidad de Zaragoza, Spain
c
Departamento de Psicología Clínica, Experimental y Social, Universidad de Huelva, Spain
d
Departamento de Psicología, Universidad de Oviedo, Spain

Received 12 March 2021; accepted 24 August 2021

Available online 3 December 2021

KEYWORDS Abstract Background/Objective: Although measurement instruments for intimate partner vio-
Intimate partner vio- lence (IPV) are available, their validity considering the interdependence of victimization and
lence perpetration self-reports based on dyadic reports has not been tested. The aim was to test the
DVQ validity and reliability of a new version of the Dating Violence Questionnaire (DVQ R) that
Dyadic data includes the interdependence of victimization and perpetration self-reports using current couple
Interdependence information. Method: Participants were young adults comprising 616 current heterosexual cou-
Descriptive study ples. Each dyad member responded to the victimization and perpetration versions of the DVQ-R
independently from their partner. Results: The victimization-perpetration interdependence
model based on dyadic data showed a good fit to the data and was invariant across sexes. All the
factors were significantly correlated with each other and were reliable. Conclusions: The DVQ is
a valid and reliable measurement instrument for the independent assessment of IPV perpetration
and victimization in adolescent and young adult populations and an interdependent measure of
IPV victimization and perpetration. The DVQ VP is invariant across sexes, which makes the
results obtained for males and females comparable. These results show the relevance of consid-
ering perpetration and victimization together and emphasize the necessity to be cautious
regarding the excessive reliability of individual self-reported perpetration or victimization to
obtain more precise knowledge.
© 2021 Asociación Española de Psicología Conductual. Published by Elsevier España, S.L.U. This is
an open access article under the CC BY-NC-ND license ([Link]
by-nc-nd/4.0/).

* Corresponding author: Universidad de Zaragoza: Facultad de Educacion, Calle de Pedro Cerbuna, 12, 50009 Zaragoza, Spain
E-mail address: juarrosbasterretxea.j@[Link] (J. Juarros-Basterretxea).

[Link]
1697-2600/© 2021 Asociación Española de Psicología Conductual. Published by Elsevier España, S.L.U. This is an open access article under the
CC BY-NC-ND license ([Link]
L. Rodríguez-Franco, J. Juarros-Basterretxea, S. Paíno-Quesada et al.

lisis
Dating Violence Questionnaire for Victimization and Perpetration (DVQ-VP): un ana
PALABRAS CLAVE de la interdependencia de los autoinformes en parejas
Violencia la pareja;
DVQ; Resumen Antecedentes/Objetivo: Aunque existen instrumentos de medida de la violencia en
dicos;
Datos dia la pareja (IPV), no se ha evaluado su validez considerando la interdependencia entre los autoin-
Interdependencia; formes de victimizacio n y perpetracion con datos dia
dicos. El objetivo fue evaluar la validez y
Estudio descriptivo fiabilidad del Cuestionario de Violencia entre Novios (CuViNo) incluyendo la interdependencia
entre los autoinformes de victimizacio n y perpetracion de los miembros de parejas actuales.
Metodo: Seiscientas diecise
is parejas heterosexuales de adultos jovenes participaron en el estu-
dio. Cada participante respondio de manera independiente a las versiones de victimizacio n y
perpetracion del CuViNo. Resultados: El modelo de interdependencia victimizacio n-
perpetracion basado en datos dia dicos mostro un buen ajuste a los datos e invarianza entre
sexos. Todos los factores correlacionaron significativamente y fueron fiables. Conclusiones: El
CuViNo es un instrumento va lido y fiable para la medicio
n independiente de perpetracio n y vic-
n de IPV en adolescentes y adultos-jo
timizacio venes, pero tambie n para la medicio
n interdepen-
diente de ambas. El CuViNo tambie n es invariante entre sexos, lo que permite comparar los
resultados de hombres y mujeres. Estos resultados muestran la relevancia de tener en cuenta la
interdependencia entre victimizacio n y perpetracio
n, así como de cuidar la excesiva confianza
en los autoinformes individuales centrados en la perpetracio n o la victimizacio
n a la hora de
alcanzar un conocimiento preciso.
© 2021 Asociación Española de Psicología Conductual. Published by Elsevier España, S.L.U. This is
an open access article under the CC BY-NC-ND license ([Link]
by-nc-nd/4.0/).

Intimate partner violence (IPV) is a widespread social and psychological violence), important differences were found
health problem affecting millions of people worldwide. It is in the inclusion of males and females and its roles. In their
necessary to consider all the potential resources, starting review, 6.25% (n = 5) of the studies only included males. Of
from the precise measure of the phenomena for primary pre- these, 60% (n = 3) addressed males as perpetrators, and 40%
vention, to better understand and address IPV. Along with (n = 2) addressed males victims. However, no study included
interviews, psychometric tests are the basic instrument both victimization and perpetration. Fifty percent (n = 40)
used by psychologists to measure, analyze, and understand of the studies only included female participants. Of these,
human behavior; and psychometric test are widely used in 97.5% (n = 39) addressed females as victims, and 2.5% (n = 1)
the applied research. In this regard, different measurement addressed females as victims and perpetrators. However,
instruments have been developed and are available in the none of the studies included females only as perpetrators.
scientific literature to obtain an accurate measure of IPV. Finally, 43.75% of the studies included males and females.
Some of the instruments are specifically oriented to young Of These, 17.14% (n = 6) only measured perpetration, 25.71%
populations (Exner-Cortens et al., 2016a; Jennings et al., (n = 9) only measured victimization, and 57.14% (n = 20)
2017; Riesgo-Gonza lez et al., 2019), and the IPV among measured victimization and perpetration. Of the twenty
these populations is generally called dating violence (DV). studies measuring perpetration and victimization, 75%
Although the definitions of DV vary considerably in some (n = 15) did not assign a specific role of perpetrator or vic-
cases (Duval et al., 2020; Exner-Cortens et al., 2016a; Mar- timization to males or females, and the other 25% (n = 5)
cos et al., 2020; Sjo€din et al., 2017; Smith et al., 2015), assigned the perpetrator role to males and the victim role to
here, DV is considered as a synonym of IPV in adolescent and females. In summary, only 18.75% (n = 15) of the studies con-
young populations that are not cohabiting or married (for a sidered the perpetration and victimization of both males
similar approach, see Rodríguez-Díaz et al., 2017; Rodrí- and females.
guez-Franco et al., 2007, 2010, 2017). As demonstrated by Lo pez-Cepero et al. (2015), the mea-
From a primary prevention perspective, exactly estimat- surement of IPV using behavioral instruments generally
ing the prevalence of DV is pivotal not only because of its tended to study individual self-reports of victimization or
immediate effect on young people involved in the violent perpetration. To overcome these limitations, different mea-
dyad but also because of the long-term consequences (Lin & surement instruments (e.g., the Revised Conflict Tactics
Chiao, 2020). The probability of IPV increases when violent Scale [CTS-2] or the Conflict in Adolescent Dating Relation-
and abusive patterns occur in adolescent and youth couple ships Inventory [CADRI], among others) allow researchers to
(Exner-Cortens et al., 2017; Wincentak et al., 2017); and as connect the perpetration and victimization reports of the
mentioned above, different behavioral tests that can be same informant using analogous items for perpetration
used to measure IPV are available in the literature. Never- (e.g., “I slapped my partner”) and victimization (e.g., “My
theless, considerable variability exists across instruments. partner slapped me”). Nevertheless, there is a tendency to
For example, Lo pez-Cepero et al. (2015) reviewed 54 behav- validate the perpetration and victimization scales sepa-
ioral instruments and observed that beyond the usual varia- rately and not consider the potential interdependence
tions in the structure of the types of violence measured and between perpetration and victimization self-reports. The
that could be justified by specific aims (e.g., measuring only relevance of considering DV perpetration and victimization

2
International Journal of Clinical and Health Psychology 22 (2022) 100276

together has been recently emphasized by different individual reports and the focus on victimization. Based on
researchers (Herrero et al., 2020; Juarros-Basterretxea et current reports of victimization and perpetration for both
al., in press; Park & Kim, 2019), indicating that higher levels members of couples, the aim of the current research was
of perpetration are related to levels of victimization in both twofold. The first aim was to test the measurement models
males and females. Considering this, it is crucial to have of the DVQ-R in its original version for victimization and
available validated measurement instruments including per- a version for perpetration (DVQ-RP) independently, as has
petration-victimization interdependence that can be used traditionally been conducted. The second aim was to test
to generate new empirical evidence by considering this the measurement model based on the interdependency
potentially influential element. approach for the Dating Violence Questionnaire for Victimi-
The interdependence between perpetration and victimi- zation and Perpetration (DVQ-VP), which assumes interde-
zation self-reports can be attributed to different reasons or pendence between members’ reports of victimization and
situations (Juarros-Basterretxea et al., in press), such as the perpetration.
mutual and bidirectional IPV found in different researches
(Exner-Cortens et al., 2016a; Hadersty & Ogolsky, 2020;
Rubio-Garay et al., 2017) and the consequential overlapping Method
victim-offender characteristics of the dynamics of a violent
couple in which both couple members are victims and perpe- Participants
trators, the perpetration of reactive violence as retaliation
or self-defense, explaining the perpetration of violence a The sample was composed of 616 Spanish heterosexual cou-
consequence of previous victimization (Park & Kim, 2019), ples (1,258 participants) of young adults between 18 and
or the tendency to justify one’s own perpetration by making 26 years of age (M = 21.07, SD = 2.29). The females
levels of victimization congruent (“I battered my partner (M = 20.57, SD = 2.11) were younger than the males
because he/she battered me”) (Herrero et al., 2020). (M = 21.58, SD = 2.35) (t(1215.96)= -7.99, p .001, Hedges
Although the mentioned cases are different in nature, they g = .45). A total of 35.5% of the participants had secondary
share the necessity to consider the interdependence studies, 63.1% had university studies, and only a minority of
between perpetration and victimization self-reports to 1.2% had primary studies. A significant small association
obtain a more precise understanding of DV. between academic level and sex was found (x22 = 16.63, p
To the best of our knowledge and contrary to our expect- .001, Cramers V = .12) with a higher proportion of females
ations, the validity of the victimization-perpetration model possessing university students (69%) and a higher proportion
based on dyadic reports has not been tested. Therefore, the of males possessing secondary education (41.1%). The pro-
model could assume the interdependence of dyad members’ portions of males and females who had only primary studies
scores and test the validity of the dyadic DV measurement were similar. Only people who were engaged in an intimate
model. relationship at the time of the assessment and his/her part-
The Dating Violence Questionnaire (DVQ, also known as ner were considered for the present study. The relationship
CuViNo due to its Spanish name Cuestionario de Violencia length varied from 1 month to 118 months (M = 28.49,
entre Novios) was originally developed by Rodríguez-Franco SD = 24.71).
et al. (2007) to measure IPV victimization in adolescent and
young populations. This first version was composed of 42 Instruments
items distributed in eight factors: Detachment (7 items),
Humiliation (7 items), Coercion (6 items), Emotional punish- Dating Violence Questionnaire-R for Victimization (DVQ-R).
ment (3 items), Physical (5 items), Sexual (6 items), Gender- This instrument was used to measure dating violence victim-
based violence (5 items), and Instrumental (3 items). Never- ization at the hands of the current partner. The DVQ-R is
theless, the original scale was considered to be too long for composed of 20 items measuring five different forms of dat-
massive application, and a reduced version (DVQ-R) of 20 ing violence victimization: physical (i.e., has beaten you),
items was developed to facilitate its application (Rodríguez- sexual (i.e., insists on touching you in ways and places that
Díaz et al., 2017). This alternative reduced instrument also you do not like and do not want), humiliation (i.e., criticizes
implicates a more parsimonious model of five factors com- you, underestimates the way you are, or humiliates your
posed of four items each: Detachment, Humiliation, Coer- self-esteem), detachment (i.e., does not recognize any
cion, Physical, and Sexual. Thus, the DVQ-R is based on the responsibility regarding both of you), and coercion (i.e., has
three big types of IPV of psychological (detachment, humili- physically kept you). Each dimension was measured by four
ation, and coercion), physical, and sexual IPV. items using a five-point Likert scale ranging from 0 (never)
The DVQ model has been adapted to different countries to 4 (all the time).
and cultural contexts and shown good psychometric proper- Dating Violence Questionnaire for Perpetration (DVQ-RP):
ties. Specifically, the DVQ model proposed for Spanish sam- The DVQ-R items were adapted to measure aggression
ples has been replicated in Spain, Me xico and Argentina against one’s partner (DVQ-RP). The DVQ-RP is composed of
(Rodríguez-Franco et al., 2010); Bolivia (Alfaro-Urquiola, 20 items measuring five different forms of dating violence
2020); Chile (Lara & Lo pez-Cepero, 2018); Ecuador (Cher- victimization: physical (i.e., you have beaten your partner),
rez-Santos et al., in press); Italy (Presaghi et al., 2015); and sexual (i.e., you insist on touching your partner in ways and
the United States (Lo pez-Cepero et al., 2016). Despite the places which she/he does not like and does not want),
demonstrated validity and reliability of the DVQ-R, it was humiliation (i.e., you criticize your partner, underestimate
originally proposed to measure participants’ DV victimiza- the way she/he is, or humiliate her/his self-esteem),
tion only and thus is limited by the excessive reliability on detachment (i.e., you do not recognize any responsibility

3
L. Rodríguez-Franco, J. Juarros-Basterretxea, S. Paíno-Quesada et al.

regarding both of you), and coercion (i.e., you have physi- estimation was used for estimations in CFAs and multigroup
cally kept your partner). Each of the five dimensions of dat- analyses. Finally, the reliability of the scales was estimated
ing violence aggression was measured by four items using a by the ordinal omega. As indicated by several researchers
five-point Likert scale ranging from 0 (never) to 4 (all the (see, for example, Elosua & Zumbo, 2008), when the item
time). response scale is ordinal, the reliability should be estimated
based on the polychoric correlation matrix due to the ten-
Procedure dency of coefficients based on the covariance matrix to
underestimate the real reliability.
Following the Code of Ethics of the World Medical Associa- The Mplus 8.6 software (Muthe n & Muthe n, 1998-2021)
tion (Declaration of Helsinki), informed consent signed by was used for the CFA, multigroup analyses and reliability
each participant was obtained before responding to the self- estimations.
reports. Anonymity and confidentiality were ensured: both
members of the couple responded to the questionnaires at
the same time separately, and no information was given to Results
the other member of the couple. To guarantee this anonym-
ity, once both participants answered the questionnaire, they Response frequency per item category analysis
were kept in an envelope that was closed in front of partici-
pants. After that, the questionnaires were encoded to iden- The analysis of responses revealed a misrepresentation of
tify the couples (e.g., 1A and 1B). Finally, all the analyses the third (habitually) and fourth (all of the time) response
were conducted using the complete sample and never con- categories. More specifically, the proportion of participants
ducted for individual responses. The researcher provided who stated they experienced at least one type of victimiza-
their contact information to respond to any possible discom- tion habitually or all the time varied between 0.1% and
fort or doubts associated with the study. The only inclusion 5.2%. Regarding perpetration, between 0.1% and 3.8% of the
criterion for sample selection was being involved in a het- participants said that they perpetrated at least one type of
erosexual relationship at the time of the evaluation. aggression habitually or all the time. Considering that the
second response option was frequently and these results,
Data analysis the responses of the second, third and fourth categories
were grouped into three response categories: 0 for never, 1
First, considering the lack of representativeness regarding for sometimes, and 2 for frequently. These are indicative of
the nature and frequency more severe forms of DV in com- a frequency-based increase in severity.
munity samples, the potential lack of representation of the
responses to item categories was analyzed using frequency Measurement model for DVQ RV
analysis. To accomplish this, IBM SPSS Statistics 26 was used.
Second, three confirmatory factor analyses (CFAs) were con- The dating violence victimization model was calculated
ducted to test the model fit to the data of the (1) DVQ RV, based on the original five correlated factor structure. In this
(2) DVQ RP, and (3) combined Dating Violence Question- model, all the items significantly loaded to their correspond-
naire for Victimization and Perpetration (DVQ VP). ing factors. The model fit was good: x2(160) = 345.24, p
Although the DVQ-R demonstrated good psychometric prop- .001, CFI = .973, RMSEA = .031 and 90% CI [.026, .035].
erties in previous research (Rodríguez-Díaz et al., 2017), it
never has been used in the way used by this research (both Measurement model for DVQ RP
members answer the same questions, and participants also
had to respond to her/his own DV perpetration). Therefore, The dating violence perpetration model was calculated
the factor structure was also tested to ensure the replicabil- based on the five correlated factor structure of the dating
ity of the DVQ-R in this condition. The model fit to the data violence victimization model. In this model, all the items
was measured through the chi-squared statistic (p > .05 for significantly loaded to their corresponding factors. The
good fit), CFI (values .95 for good fit) and RMSEA (values model fit was good: x2(160) = 312.38, p .001, CFI = .967,
.05 for good fit) (Hu & Bentler, 1999). Third, after obtaining RMSEA = .028 and 90% CI [.023, .032].
the model that best fit the data, the configural, metric, and Standardized estimates for the measurement model of
scalar invariance across sexes (male and female) of the the DVQ RV (out of brackets) and DVQ RP (in brackets) are
DVQ VP was tested (Putnick & Bornstein, 2016; Vandenberg displayed in Figure 1.
& Lance, 2000). Configural invariance represents the invari-
ance of the form of a model, and it means that the organiza- Measurement model for DVQ VP
tion of the tested constructs is supported for both sexes.
Metric invariance represents the contributions of the items The dating victimization perpetration model was calcu-
to the latent construct, and it is obtained if these contribu- lated considering the previous two models. This model, as
tions are similar for both sexes. Finally, scalar invariance shown in Figure 2, is composed of ten correlated factors:
means that differences in the latent construct capture all five factors for victimization and five factors for perpetra-
mean differences in the shared variance of the items. The d tion. Thus, this model includes the correlations among
Dx2 test and its associated probability, DCFI, and DRMSEA dimensions of victimization such as the DVQ RV (intramodel
were considered to test the invariance across groups (Rut- correlations), the correlations among the dimensions of
kowski & Svetina, 2017; Svetina et al., 2019). The Mean and perpetration such as DVQ RP (intramodel correlations),
Variance adjusted Weighted Least Square (WLSMV) and the correlations between the five dimensions of

4
International Journal of Clinical and Health Psychology 22 (2022) 100276

Figure 1 Standardized parameter estimates for the measurement model of DVQ-RV and DVQ-RP.

Figure 2 Standardized parameter estimates for the measurement model of DVQ-VP.

5
L. Rodríguez-Franco, J. Juarros-Basterretxea, S. Paíno-Quesada et al.

Table 1 Intramodel and intermodel correlations of the measurement model for DVQ-VP.
1 2 3 4 5 6 7 8 9 10
Victimization
1 Physical 1
1 Sexual .61 1
1 Humiliation .67 .67 1
1 Detachment .47 .57 .76 1
1 Coercion .72 .68 .79 .74 1
Perpetration
1 Physical .72 .57 .57 .41 .62 1
1 Sexual .57 .61 .58 .41 .61 .64 1
1 Humiliation .55 .51 .62 .56 .65 .70 .71 1
1 Detachment .42 .50 .65 .74 .72 .52 .58 .79 1
1 Coercion .55 .60 .69 .75 .73 .70 .69 .82 .79 1
Note. p .001 in all the cases

Table 2 Configural, metric, and scalar invariance tests.

x2 (df) Dx2 (Ddf) CFI DCFI RMSEA (CI) DRMSEA
Configural 1673.73*** (1350) - .980 - .020 (.016, .023)
Metric 1704.76*** (1380) 46.98* (30) .980 .000 .020 (.016, .023) .000
Scalar 1722.88*** (1410) 34.67 (30) .981 .001 .019 (.016, .022) .001
*
p .05, ** p .01, and ***p .001

victimization and the five dimensions of perpetration (inter- presented in Table 2. The results displayed in Table 2 show
model correlations) (see Table 1). In addition, the error that the model is invariant across sexes. The configural
terms of equivalent items (i.e., item 1 of victimization [V1] model showed a good fit (x2(1350) = 1673.73, p .001,
and item 1 of perpetration [P1]) were freely estimated. The CFI = .980, RMSEA = .020 and 90% CI [.016, .023]), confirming
model fit to the data was good: x2(675) = 1018.83, p .001, that the organization of the tested constructs is supported
CFI = .980, RMSEA = .020 and 90% CI [.018, .023]. in both groups. The metric model also showed a good fit
The correlations between the ten factors (five factors of (x2(1380) = 1704.76, p .001, CFI = .980, RMSEA = .020 and
victimization and five factors of perpetration) are displayed 90% CI [.016, .023]); thus, the factor loadings are equivalent
in Table 1. The intramodel correlations varied between .477 across groups. Finally, the scalar model also showed a good
and .793 for victimization and between .527 and .793 for fit (x2(1410) = 1722.88, p .001, CFI = .981, RMSEA = .019 and
perpetration. The intermodel correlations varied from .410 90% CI [.016, .022]), so any differences found between
to .750. All the correlations were statistically significant (p groups in the latent construct captured all mean differen-
.001). ces.
Finally, Table 3 displays the reliabilities of the ten factors
Invariance across sexes for the DVQ-VP model (five factors of perpetration and five factors of victimiza-
tion). All the reliabilities varied between .78 and .92, show-
Once the model fit to the data was tested and its goodness ing good reliability. A breakdown of the descriptive
was confirmed, the invariance of the model across sexes information of the ten different factors is displayed in
(male and female) was tested at the configural, metric, and Table 3. The results show that psychological forms of DV
scalar levels; and the fit indices and model comparisons are (detachment, coercion, and humiliation), followed by

Table 3 Internal consistency and descriptive information for DVQ-VP scales.

vPerpetration M (SD) vVictimization M (SD)
Physical .88 0.72 (1.05) .90 0.86 (1.19)
Sexual .92 0.35 (0.93) .91 0.52 (1.17)
Humiliation .83 0.74 (1.02) .85 1.35 (0.97)
Detachment .78 1.21 (1.41) .83 1.45 (1.77)
Coercion .80 1.05 (1.40) .81 1.24 (1.64)

6
International Journal of Clinical and Health Psychology 22 (2022) 100276

physical and sexual violence, respectively, were the most and Cherrez-Santos et al. (in press) used the Mean and Vari-
common forms of perpetration and victimization. ance adjusted Weighted Least Square (WLSMV) estimator, as
in the current study, which is considered the more appropri-
ate estimator for modeling categorical or ordered data (see
Discussion Brown, 2006). In a similar fashion, the perpetration version
of the DVQ-R consisting of analogous items referring to per-
The current research tested the measurement models of petration instead of victimization also showed good validity.
DVQ RV/DVQ RP independently and tested the measure- As this measurement instrument has never been used in this
ment model based on the interdependence approach for the way, there is no previous empirical evidence on independent
Dating Violence Questionnaire for Victimization and Perpe- samples to which we can compare our results. Nonetheless,
tration (DVQ VP), which includes both dyad members’ self- the results obtained in the current research are congruent
reports of IPV victimization and perpetration. For this pur- with previous victimization model results, as usually occurs
pose, the complete data of 616 heterosexual couples of with other measurement instruments with analogous items
young adults were used. for perpetration and victimization (e.g., CTS-2 or CADRI).
The original version (DVQ-R) included a five-point Likert Although different measurement instruments allow
response scale ranging from 0 for never to 4 for all of the researchers to evaluate self-reported perpetration and vic-
time. The results obtained in the current research showed timization with analogous items, the validity has been tradi-
that some of these categories (3 for habitually and 4 for all tionally tested considering them as two independent scales
of the time) were systematically misrepresented. The lack and, thus, without considering the potential interdepen-
of representation of the higher categories of responses when dence between self-reports of victimization and self-reports
measuring violence is not a new issue, and it is related to of perpetration. As shown in recent research, own victimiza-
the nature of the community sample. The probability of tion self-reports are predicted by own perpetration self-
choosing higher categories that represent higher frequencies reports, so a significant part of victimization reported by
and are interpreted as more severe violence is lower in com- participants is explained by their own perpetration (Herrero
munity samples than in other samples that tend to show et al., 2020; Juarros-Basterretxea et al., in press).
more severe and frequent forms of violence (e.g., peniten- Following these recent results, the main aim of the cur-
tiary samples) (Cunha & Gonçalves, 2017). In this regard, rent research was to test the measurement model based on
collapsing some category responses was considered appro- the interdependency approach for the Dating Violence Ques-
priate since it (a) did not imply a significant loss of informa- tionnaire for Victimization and Perpetration (DVQ-VP). From
tion because it was misrepresented and (b) increased the this perspective, which considered the interdependence
simplicity and interpretability. At this point, it is important between participants’ reports of victimization and perpetra-
to note that collapsing categories 3 and 4 into category 2 is tion, a victimization-perpetration measurement model was
coherent with the nature of the scale because those tested and showed a good fit to the data. As for the perpe-
respondents who answer 3 and 4 would respond 2 in the case tration measurement model, there is no previous empirical
of using only a trichotomic response scale as the options research that permits a comparison, but the fit of the inter-
refer to the increasing frequency of a behavior. This type of dependent model is congruent with previous results
modification can be observed in other research where fewer obtained for independent models of victimization and per-
category responses were used compared to the original ver- petration.
sion (e.g., the CTS; see Johnson et al., 2020; Lin et al., Additionally, the DVQ VP is invariant across sexes; thus,
2020; Mun ~oz-Rivas et al., 2007; Schnurr et al., 2010; Smith the perpetration victimization model is equal for males
et al., 2015). Thus, this modification reduces the range of and females, improving the generalizability of the model.
the frequency but still allows researchers to differentiate The configural, metric and scalar invariance allow research-
categories according to the nature of DV (CDC, 2020). The ers to precisely interpret latent mean differences (Putnick &
obtained results add to previous research and suggest that Bornstein, 2016; Vanderberg & Lance, 2000), demonstrating
using fewer category responses for items is a more suitable that the models have the same form, that the items contrib-
option for future research. ute similarly to the latent construct for males and females,
The obtained results show that the DVQ is a reliable and and that differences in the latent construct capture all
valid measurement instrument for the independent assess- mean differences.
ment of DV victimization and perpetration and an interde- These results show the relevance of considering perpe-
pendent measure of victimization and perpetration. tration and victimization together to obtain more precise
Congruent with previous research (Alfaro-Urquiola, 2020; knowledge of the DV phenomena and thus the necessity to
Cherrez-Santos et al., in press; Lara & Lo pez-Cepero, 2018; develop valid measurement instruments that consider this
pez-Cepero et al., 2016; Presaghi et al., 2015; Rodríguez-
Lo interdependence to address the diversity of scenarios that
Díaz et al., 2017; Rodríguez-Franco et al., 2010), the validity can explain it: (1) the existence of mutual violence (victim-
of the DVQ model for victimization was demonstrated. Fur- offender overlap) and being engaged in a violent couple
thermore, in the current research, the free estimation of dynamic in which both couple members are victims and per-
certain parameters included in the original version (see petrators; (2) the perpetration of reactive violence as retal-
Rodríguez-Díaz et al., 2017) was not required, in line with iation or self-defense, explaining violence perpetration as a
more recent research (Cherrez-Santos et al., in press). The consequence of the previous victimization; and (3) the
difference in these results is attributable to the different upwards victimization bias or the tendency to find justifica-
estimators used in each research. For example, Rodríguez- tion for one’s own perpetration biasing the levels of victimi-
Díaz et al. (2017) used the maximum likelihood estimation; zation and increasing them. From this theoretical

7
L. Rodríguez-Franco, J. Juarros-Basterretxea, S. Paíno-Quesada et al.

approach, the proposed model has important practical measurement instrument potentially applicable to young
implications in measuring and understanding DV and empha- samples in the community to conduct the primary detec-
sizes the necessity to be cautious regarding the excessive tion of DV perpetration and victimization.
reliability of individual self-reports of only perpetration or
victimization.
The current research possesses strengths and potential
Funding
limitations. This study extends the knowledge on a valid and
reliable assessment instrument by supporting the results Support for this research was provided by the Spanish
obtained in previous research (Rodríguez-Díaz et al., 2017) Ministry of Science and Innovation (Reference PID2020-
and is, to our knowledge, the first study reporting evidence 114736GB-I00).
on the reliability (internal consistency) and validity (factor
structure and invariance) of a DV/IPV measurement instru-
ment considering the interdependence between one’s own
perpetration and victimization reports. The inclusion of the Appendix A. Supplementary data
interdependent nature of DV victimization and perpetration
(Herrero et al., 2020; Juarros-Basterretxea et al., in press) Supplementary data associated with this article can be
is an innovative approach and contributes to advancing the found, in the online version, at [Link]
field by presenting a useful and brief DV perpetra- ijchp.2021.100276.
tion victimization model formed using empirical evidence
misrepresented in much research (Exner-Cortens et al.,
2016b). Nonetheless, the DVQ-VP possesses the inherent lim- References
itations of behavioral measurement instruments (Hardesty &
Ogolski, 2020), and it would be beneficial for future research Alfaro-Urquiola, A. (2020). Validation of Dating Violence Question-
to use it together with other instruments and data collection naire-R (DVQ-R) in a sample of young people from La Paz. Ajayu,
methods and to test its convergent and divergent validity 18, 102–120.
considering different levels of correlates of IPV (Dardis Brown, T. (2006). Confirmatory factor analysis for applied research.
Guildford.
et al., 2015; Hammock et al., 2017; Herrero et al., 2016,
Cantos, A. L., Kosson, D. S., Goldstein, D. A., & O’Leary, K. D (2019).
2018; Juarros-Basterretxea et al., 2018, 2019, 2020). Like- Treatment impact on recidivism of family only vs. generally vio-
wise, it would be convenient to include other information lent partner violence perpetrators. International Journal of
sources and methods to assess real injury (Vilarin ~o et al., Clinical and Health Psychology, 19, 171–180. [Link]
2018) and to verify the victimization and perpetration (Gan- 10.1016/[Link].2019.05.002.
cedo et al., 2021). Cherrez-Santos, A., V., Alulema-Sanchez, S., & Juarros-Basterret-
Furthermore, the sample of the current research also xea, J. (in press). Psychometric properties of the Dating Violence
represents an important strength. First, all the participants Questionnaire-R in women from Ecuador. Revista Electro nica de
must be engaged in an intimate relationship at the time of Metodología Aplicada.
CDC. (2020). Preventing intimate partner violence. . [Link]
the study, which makes the data more reliable than those of
gov/violenceprevention/intimatepartnerviolence/[Link].
other studies based on previous or last relationships because
Cunha, O., & Gonçalves, R. A. (2017). Male perpetrators of intimate
of the lower memory recall bias (Exner-Cortens et al., partner violence: A comparison between incarcerated and com-
2016b). Second, the study’s sample size exceeds the sample munity offenders. International Journal of Offender Therapy
sizes of other studies (Exner-Cortens et al., 2016a), espe- and Comparative Criminology, 62, 3260–3277. [Link]
cially considering that the sample was composed of current 10.1177/0306624x17741070.
couples. Nevertheless, it is also true that the sample was Dardis, C. M., Dixon, K. J., Edwards, K. M., & Turchik, J. A. (2015).
not representative, so any generalization of the data must An examination of the factors related to dating violence perpe-
be done cautiously. For instance, generalized and specific tration among young men and women and associated theoretical
perpetrators have not been included (Cantos et al., 2019). explanations: A review of literature. Trauma, Violence & Abuse,
16, 136–152. [Link]
Similarly, even though the number of participating couples
Duval, A., Lanning, B. A., & Patterson, M. S. (2020). A systematic
was high, they were all heterosexual couples, and future
review of dating violence risk factors among undergraduate col-
research should replicate these results considering more lege students. Trauma, Violence, & Abuse, 21, 567–585. https://
diversity among couples (Edwards et al., 2015; Harden et [Link]/10.1177/1524838018782207.
al., 2020; Laskey et al., 2019; Lin et al., 2020; Martínez- Edwards, K. M., Sylaska, K. M., & Neal, A. M. (2015). Intimate part-
Gomez et al., 2021; Peitzmeier et al., 2020; Ramiro-Sanchez ner violence among sexual minority populations: A critical
et al., 2018; Rojas-Solís et al., 2019). review of the literature and agenda for future research. Psychol-
In conclusion, the DVQ VP is the only DV measure- ogy of Violence, 5, 112–121. [Link]
ment instrument considering the interdependence of per- Elosua, P., & Zumbo, B. D. (2008). Reliability coefficients for cate-
petration and victimization self-reports that has been gorical response scales. Psicothema, 20, 896–901.
Exner-Cortens, Eckenrode, J., Bunge, J., & Rothman, E (2017).
validated, allowing its use regardless of the sex of the
Revictimization after adolescent dating violence in a matched,
participants. The inclusion of the interdependence of
national sample of youth. Journal of Adolescent Health, 60, 176–
self-reported scores is a methodological improvement in 183. [Link]
the DV measure that better measures and understands Exner-Cortens, D., Gill, L., & Eckenrode, J. (2016a). Measurement
DV and the complex link between victimization and per- of adolescent dating violence: A comprehensive review (Part 1,
petration. In line with its predecessor, the DVQ R, the behaviors). Aggression and Violent Behavior, 27, 64–78. https://
DVQ-VP is a valid, reliable, and relatively short [Link]/10.1016/[Link].2016.02.007.

8
International Journal of Clinical and Health Psychology 22 (2022) 100276

Exner-Cortens, D., Gill, L., & Eckenrode, J. (2016b). Measurement men. The European Journal of Psychology Applied to Legal Con-
of adolescent dating violence: A comprehensive review (Part 2, text, 11, 61–69. [Link]
attitudes). Aggression and Violent Behavior, 27, 93–106. Lara, L., & Lopez-Cepero, J. (2018). Psychometric properties of the
[Link] Dating Violence Questionnaire: Reviewing the evidence in Chil-
Gancedo, Y., Farin ~a, F., Seijo, D., Vilarin
~o, M., & Arce, R. (2021). ean Youths. Journal of Interpersonal Violence. [Link]
Reality Monitoring: A meta-analytical review for forensic prac- 10.1177/0886260518760612 Advanced online publication..
tice. European Journal of Psychology Applied to Legal Context, Laskey, P., Bates, E. A., & Taylor, J. C. (2019). A systematic lit-
13, 99–110. [Link] erature review of intimate partner violence victimisation: An
Hammock, G. S., Richardson, D. S., Lamm, K. B., Taylor, E., & inclusive review across gender and sexuality. Aggression and
Verlaque, L. (2017). The effect of gender of perpetrator and vic- Violent Behavior, 47, 1–11. [Link]
tim on perceptions of psychological and physical intimate part- avb.2019.02.014.
ner aggression. Journal of Family Violence, 32, 357–365. Lin, W. H., & Chiao, C. (2020). Adverse adolescence experiences,
[Link] feeling lonely across life stages and loneliness in adulthood.
Harden, J., McAllister, P., Spencer, C. M., & Stith, S. M. (2020). The International Journal of Clinical and Health Psychology, 20, 243–
dark side of the rainbow: Queer women’s experiences of inti- 252. [Link]
mate partner violence. Trauma, Violence, & Abuse. [Link] Lin, K., Hu, R., Wang, X., & Xue, J. (2020). Female same-sex bidi-
org/10.1177/1524838020933869 Advanced online publication. rectional intimate partner violence in China. Journal of Interper-
Hardesty, J. L., & Ogolsky, B. G. (2020). A socioecological perspec- sonal Violence. [Link]
tive on intimate partner violence research: A decade in review. Advanced online publication..
Journal of Marriage and Family, 82, 454–477. [Link] pez-Cepero, J., Fabelo, H. E., Rodríguez-Franco, L., &
Lo
10.1111/jomf.12652. Rodríguez-Díaz, F. J. (2016). The Dating Violence Question-
Herrero, J., Rejano-Herna ndez, L., Rodríguez-Franco, L., naire: Validation of the Cuestionario de Violencia de Novios
Juarros-Basterretxea, J., & Rodríguez-Díaz, F. J. (2020). The using a college sample from the United States. Violence and
Actor-Partner Interdependence Model in the study of aggression Victims, 31, 438–456. [Link]
and victimization within couples: an empirical examination in D-14-00077.
361 dyads. Psychosocial Intervention, 29, 165–174. [Link] pez-Cepero, J., Rodríguez-Franco, L., & Rodríguez-Díaz, F. J. (2015).
Lo
org/10.5093/pi2020a12. Measuring intimate partner abuse. A review of behavioral assess-
Herrero, J., Torres, A., Fernandez-Suarez, A., & ment tools. Revista Iberoamericana de Diagno stico y Evaluacio
n
Rodríguez-Díaz, F. J. (2016). Generalists versus specialists: Psicologica, 2, 37–50.
Toward a typology of batterers in prison. The European Journal Marcos, V., Gancedo, Y., Castro, B., & Selaya, A. (2020). Dating vio-
of Psychology Applied to Legal Context, 8, 19–26. [Link] lence victimization, perceived gravity in dating violence behav-
org/10.1016/[Link].2015.09.002. iors, sexism, romantic love myths and emotional dependence
Herrero, J., Torres, A., & Rodríguez-Díaz, F. J. (2018). Child abuse, between female and male adolescents. Revista Iberoamericana
risk in male partner selection and intimate partner violence vic- de Psicología y Salud, 11, 132–145. [Link]
timization of women of the European Union. Prevention Science, rips.2020.02.040.
19, 1102–1112. [Link] Martínez-Go mez, J. A., Bolívar-Sua rez, Y., Rey-Anacona, C. A.,
Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in Ramírez-Ortiz, L. C., Lizarazo-Ojeda, A. M., & Yanez-
covariance structure analysis: Conventional criteria versus new Perun~ un
~i, L. Y. (2021). Traditional schemes of sexual gender
alternatives. Structural Equation Modeling: A Multidisciplinary roles, power in relationships and dating violence. Revista Iberoa-
Journal, 6, 1–55. [Link] mericana de Psicología y Salud, 12, 1–16. [Link]
Jennings, W. G., Okeem, C., Piquero, A. R., Sellers, C. S., 10.23923/[Link].2021.01.041.
Theobald, D., & Farrington, D. P. (2017). Dating and intimate Mun~oz-Rivas, M. J., Gran ~a Gomez, J. L., O’Leary, K. D., &
partner violence among Young persons ages 15-30: Evidence Gonza lez Lozano, P. (2007). Physical and psychological aggres-
from a systematic review. Aggression and Violent Behavior, 33, sion in dating relationships in Spanish university students. Psico-
107–125. [Link] thema, 19, 102–107.
Johnson, W. L., Taylor, B. G., Mumford, E. A., & Liu, W. (2020). Muthen, L., & Muthe n, B. (1998). Mplus User’s Guide. Seventh edi-
Dyadiccorrelatesoftheperpetrationofpsychologicalaggressiona- tion. Author.
mongintimatepartners. Psychology of Violence, 10, 422–431. Park, S., & Kim, S. H. (2019). Who Are the Victims and Who Are the
[Link] Perpetrators in Dating Violence? Sharing the Role of Victim and
Juarros-Basterretxea, J., Herrero, J., Escoda-Mene ndez, P., & Perpetrator. Trauma, Violence, & Abuse, 20, 732–741. https://
Rodríguez-Díaz, F. J. (2020). Cluster B personality traits and psy- [Link]/10.1177/1524838017730648.
chological intimate partner violence: Considering the media- Peitzmeier, S. M., Malik, M., Kattari, S. K., Marrow, E.,
tional role of alcohol. Journal of Interpersonal Violence. Stephenson, R., Age nor, M., & Reisner, S. L. (2020). Intimate
[Link] Advanced online partner violence in transgender populations: Systematic review
publication. and meta-analysis of prevalence and correlates. American Jour-
Juarros-Basterretxea, J., Herrero, J., Ferna ndez-Sua rez, A., nal of Public Health. [Link]
Perez, B., & Rodríguez-Díaz, F. J. (2018). Are generalist batter- 305774 Advanced online publication.
ers different from generally extra-family violent men? A study Presaghi, F., Manca, M., Rodríguez-Franco, L., & Curcio, G. (2015). A
among imprisoned male violent offenders. The European Journal questionnaire for the assessment of violent behaviors in young
of Psychology Applied to Legal Context, 10, 8–14. [Link] couples: the Italian version of Dating Violence Questionnaire
org/10.5093/ejpalc2018v10n1a1. (DVQ). PLOS One, 10. [Link]
Juarros-Basterretxea, J., Ocampo, N. Y., Herrero, J., & Rodríguez- pone.0126089 Article e0126089.
Díaz, F. J. (in press). Dyadic analysis of emotional intimate part- Putnick, D. L., & Bornstein, M. H. (2016). Measurement invariance
ner violence: An empirical examination among 242 dyads. The conventions and reporting: The state of the art and future direc-
European Journal of Psychology Applied to Legal Context. tions for psychological research. Developmental Review, 41, 71–
Juarros-Basterretxea, J., Overall, N., Herrero, J., & 90. [Link]
Rodríguez-Díaz, F. J. (2019). Considering the effect of sexism on Ramiro-Sa nchez, T., Ramiro, M. T., Bermu dez, M. P., &
psychological intimate partner violence: A study with imprisoned Buela-Casal, G. (2018). Sexism in adolescent relationships: A

9
L. Rodríguez-Franco, J. Juarros-Basterretxea, S. Paíno-Quesada et al.

systematic review. Psychosocial Intervention, 27, 123–132. Rutkowski, L., & Svetina, D. (2017). Measurement invariance in
[Link] international surveys: Categorical indicators and fit measure
Riesgo Gonza lez, N., Ferna ndez-Suarez, A., Herrero, J., performance. Applied Measurement in Education, 30, 39–51.
Rejano-Herna ndez, L., Rodríguez-Franco, L., Paino-Quesada, S. G., [Link]
& Rodríguez-Díaz, F. J. (2019). Concordancia en la percepcio n de €din, A. K., Wallinius, M., Billstedt, E., Hofvander, B., &
Sjo
conductas violentas en parejas adolescentes. Terapia psicolo gica, Nilsson, T. (2017). Dating violence compared to other types of
37, 154–165. [Link] violence: Similar offenders but different victims. The European
Rodríguez-Díaz, F. J., Herrero, J., Rodríguez-Franco, L., Journal of Psychology Applied to Legal Context, 9, 83–91.
Bringas-Molleda, C., Paíno-Quesada, S. G., & [Link]
Perez-Sanchez, B. (2017). Validation of Dating Violence Question- Smith, J., Mulford, C., Latzman, N. E., Tharp, A. T., Niolon, P. H., &
narie-R (DVQ-R). International Journal of Clinical and Health Psy- Blachman-Demner, D. (2015). Tacking stock of behavioral meas-
chology, 16, 77–84. [Link] ures of adolescent dating violence. Journal of Aggression, Mal-
Rodríguez-Franco, L., Antun ~a, M. A., Rodríguez-Díaz, F. J., treatment & Trauma, 24, 674–692. [Link]
Herrero, F. J., & Nieves, V. E. (2007). Violencia de ge nero en 10926771.2015.1049767.
relaciones de pareja durante la adolescencia: ana lisis diferencial Schnurr, M. P., Lohman, B. J., & Kaura, S. A. (2010). Variation in late
del Cuestionario de Violencia entre Novios (CuViNo). In R. Arce, adolescents' reports of dating violence perpetration: a dyadic
~a, E. Alfaro, C. Civera, F. Tortosa (Eds.), Psicología juríd-
F. Farin analysis. Violence and Victims, 25, 84–100. [Link]
ica. Violencia y víctimas (pp. 137 147). Council of Valencia. 10.1891/0886-6708.25.1.84.
Rodríguez-Franco, L., Gracia, C., Juarros-Basterretxea, J., Svetina, D., Rutkowski, L., & Rutkowski, D. (2019). Multiple-
Fernandez-Sua rez, A., & Rodríguez-Díaz, F. J. (2017). Generalist group invariance with categorical outcomes using updated
and specialist batterers in teen and young dating violence: Impli- guidelines: An illustration using Mplusand the lavaan/sem-
cations for development of prevention programs. Accio n Psi- Tools Packages. Structural Equation Modeling: A Multidisci-
cologica, 14, 1–16. [Link] plinary Journal, 27, 111–130. [Link]
Rodríguez-Franco, L., Lo pez-Cepero, J., Rodríguez-Díaz, F. J., 10705511.2019.1602776.
Bringas, C., Antun ~a, A., & Estrada, C. (2010). Validacio n del Vandenberg, R. J., & Lance, C. E. (2000). A review and synthesis of
Cuestionario de Violencia entre Novios (CUVINO) en jo venes his- the measurement invariance literature: Suggestions, practices,
panohablantes: Ana lisis de resultados en Espan ~a, Me xico y and recommendations for organizational research. Organiza-
Argentina. Anuario de Psicología Clínica y de la Salud /Annuary tional Research Methods, 3, 4–70. [Link]
of Clinical and Health Psychology, 6, 45–53. 109442810031002.
Rojas-Solís, J. L., Guzma n-Pimentel, M., Jime nez-Castro, M. P., ~o, M., Amado, B. G., Va
Vilarin zquez, M. J., & Arce, R. (2018). Psy-
Martínez-Ruíz, L., & Flores-Herna ndez, B. G. (2019). Violence chological harm in women victims of intimate partner violence:
towards men in the heterosexual couple: A review. Ciencia y Socie- Epidemiology and quantification of injury in mental health
dad, 44, 57–70. [Link] markers. Psychosocial Intervention, 27, 145–152. [Link]
Rubio-Garay, F., Lo pez-Gonza lez, M. A., Carrasco, M. A., & org/10.5093/pi2018a23.
Amor, P. J. (2017). The prevalence of dating violence: A system- Wincentak, K., Connolly, J., & Card, N. (2017). Teen dating vio-
atic review. Psychologist Papers, 38, 135–147. [Link] lence: A meta-analytic review of prevalence rates. Psychology
10.23923/pap.psicol2017.2831. of Violence, 7, 224–241. [Link]

10