Psychometric Properties of the Violence Scale (EV) in Colombian Men

Abstract

Intimate partner violence is a public health and human rights issue that must be addressed and measured to inform care and prevention efforts. The Violence Scale (EV) is used to assess different types of violence; however, there is no evidence of its validity and reliability in Colombia, nor are there specific studies on men. Objective: To evaluate the metric properties of the Partner Violence Scale (EV, by its Spanish acronym) in men within the Colombian community context. Methodology: Content validation was carried out with expert judges and victims of partner violence. Subsequently, the EV was tested on a sample of 1293 Colombian men who had not reported/denounced partner violence. Confirmatory factor analyses were performed using rival models to evaluate the best-fitting factor structure, along with mean variance extracted analysis as an indicator of convergent validity and reliability analysis. Results: Every tested structure for organizing intimate partner violence showed adequate fit indicators (CFI > 0.95; TLI > 0.95). However, the lowest level of error was evident in the five-factor structure (psychological, physical, severe physical, economic, and sexual violence; RMSEA = 0.009; SRMR = 0.080). All items were loaded by the factor to which they belong, and all loadings were significant (p < 0.005). The extracted mean variances exceeded 40%, and there were high reliability indicators for the total test (>0.94) and for each factor (>0.70). Conclusion: The Violence Scale(EV) provides adequate evidence of validity and reliability to be used as a tool for assessing violent behaviors experienced by Colombian men. Its use is recommended as an instrument for highlighting violence that has been normalized due to the socio-cultural barriers that men face.

1. Introduction

Intimate partner violence (IPV) has been recognized as a social, public health, and human rights issue that affects people from all social and economic backgrounds (Ramasubramani et al. 2024; Jónasdóttir et al. 2024; Arroyave López et al. 2022; World Health Organization 2026; Young et al. 2021; Walker et al. 2020; García-Díaz et al. 2018; Heise et al. 1999; Fischbach and Herbert 1997). However, as Jahin et al. (2024) and Modise (2024) emphasize, the historical framing of intimate partner violence as a phenomenon predominantly affecting women has contributed to the marginalization of male victims. This neglect is rooted in social, cultural, and psychological determinants that are frequently underestimated due to entrenched gender stereotypes, which construct men primarily as aggressors rather than potential victims. Dominant representations of masculinity further obscure the acknowledgment of male victimization, reinforcing perceptions of male authority and strength. Within patriarchal contexts, these dynamics often engender feelings of guilt and shame among male survivors, thereby impeding disclosure and access to support (Micklitz et al. 2025; Armas-Vargas 2020; Walker et al. 2020; Rojas-Solís et al. 2019; Alegría del Ángel and Rodríguez Barraza 2015; Hines and Douglas 2010).

According to Olatunji et al. (2023), men are generally assumed to be the aggressors, which may explain the underreporting of cases of IPV against them. Furthermore, men who have suffered violence in childhood are more likely to repeat patterns of victimization or perpetration in adulthood.

Armas-Vargas (2020) notes that distorted ideas about gender roles and violent practices against partners are not exclusive to men towards women and therefore proposes conducting studies that consult both women and men from the community. Underreporting is a problem that affects men and women globally, which is explained by the normalization of violent behaviors anchored in culture (Akl Moanac et al. 2022; Támara Patiño 2021; García-Díaz et al. 2018; Alegría del Ángel and Rodríguez Barraza 2015; Ministerio de Salud y Protección Social 2015).

In this context and considering the repercussions of IPV on health, the economy, and individual, family, and social development, there is a compelling need for interventions focused on prevention and assistance, as well as the quantification and assessment of the phenomenon (Kuskoff and Parsell 2024; Expósito-Álvarez et al. 2023; Támara Patiño 2021; García-Díaz et al. 2018).

IPV is defined as a set of behaviors perpetrated by one of the members of a current or former couple that causes psychological, physical, sexual, or economic harm through manipulation, psychological abuse, physical aggression, economic control, and sexual constraint, among other tactics (Wu et al. 2020; Alegría del Ángel and Rodríguez Barraza 2015). IPV can be categorized into four main forms: psychological, physical, economic, and sexual violence. Psychological violence includes coercive, manipulative, degrading, and intimidating behaviors intended to control, dominate, or emotionally harm a partner (Acosta Vélez 2016; Alegría del Ángel and Rodríguez Barraza 2015). Physical violence encompasses acts that inflict bodily harm or injury and may jeopardize the victim’s physical integrity or survival (Caballero Guzmán et al. 2024; Acosta Vélez 2016; Alegría del Ángel and Rodríguez Barraza 2015; Corte Constitucional (CC) 2014). Economic violence refers to actions that limit a partner’s access to, control over, or use of financial resources and assets, thereby reinforcing dependency and unequal power dynamics within the relationship (Alegría del Ángel and Rodríguez Barraza 2015; Corte Constitucional (CC) 2014). Sexual violence involves the use of coercion, manipulation, threats, or physical force by a current or former partner to compel participation in unwanted sexual activities (Acosta Vélez 2016; Alegría del Ángel and Rodríguez Barraza 2015).

Regarding the prevalence in Colombia, the Ministerio de Salud y Protección Social (2015) reports that psychological violence affects 74.4% of men, compared to 64.1% of women. It also reports that 90.1% of men and 76.4% of women have not sought help. The INMLCF (National Institute of Legal Medicine and Forensic Sciences) has reported annual prevalence rates of IPV against men ranging from 14% to 16% (Arroyave López et al. 2022; Támara Patiño 2021; Acosta Vélez 2016).

Smith et al. (2021) report that detecting victimization by violence and preventing male violence before it occurs or escalates are important for reducing the risk of health consequences, which justifies the need for screening and assessment tools. Tarzia et al. (2024) indicate that only a few rigorous instruments capture the full range of violent tactics or behaviors, implying a lack of robust quantitative data for their evaluation and for achieving a deeper understanding. Currently, there is a wide variety of instruments available to assess violence. In the case of Colombia, the literature shows a wealth of studies on violence against women; however, no research was found on the validation of measurement scales to assess IPV against men.

Although scales are used to assess intimate partner violence, e.g., Straus’ Conflict Tactic Scale (CTS) (Straus, 1979), the Index of Spouse Abuse (ISA) by Hudson and McIntosh (1981), the Severity of Violence Against Women Scale (SVAWS) by Marshall (1992), and the Male Victims of Domestic Abuse Scale (MVDAS) by Hines and Douglas (2010), these scales do not cover the various dimensions of violence described above. Therefore, they do not correspond to the ways in which official statistics classify them, or they focus on violence against women and are not equally applicable to women and men.

Valdez-Santiago et al. (2006) developed and validated the Violence Scale and the Severity Index (EV) for use in the Mexican population. The instrument was originally designed to assess psychological, physical, sexual, and economic violence, and was based on the ISA and SVAWS. Exploratory factor analysis identified a four-factor structure comprising psychological violence, physical violence, severe physical violence, and sexual violence. The scale demonstrated excellent internal consistency, with a Cronbach’s alpha coefficient of 0.99. However, items intended to measure economic violence did not load onto the expected factor and were therefore excluded from the final version of the instrument. The resulting four-factor model accounted for 62.2% of the total variance.

In Colombia, a validation study was conducted exclusively with women from community samples who had not reported episodes of intimate violence (Acuña-Arango et al. forthcoming). This research identified a five-factor structure (psychological, physical, severe physical, economic, and sexual violence) that was consistent with the original model. Items related to life-threatening violence (severe physical violence) constituted a distinct, new dimension. Furthermore, and in contrast to the findings of the Mexican study, this investigation retained the dimension of economic violence.

Some abusive or violent practices toward partners are not always understood as violence; therefore, they need to be identified and evaluated to inform actions to prevent abusive behaviors that can escalate. Likewise, assessing the prevalence of IPV provides necessary input for evidence-based care and intervention programs. Given the lack of literature on validated instruments in Colombia, this study is justified in its objective to adapt and evaluate the evidence of validity and reliability of the Mexican EV (Valdez-Santiago et al. 2006) within the Colombian context using male samples. Having an instrument that allows for the timely detection and classification of IPV against men facilitates progress in understanding the broader phenomenon of IPV in the country.

2. Materials and Methods

Type of study

This research is a quantitative, instrumental study (Montero and León 2007) designed to analyze the psychometric properties of the Intimate Violence Scale (EV, by its Spanish acronym) (Valdez-Santiago et al. 2006) in Colombian men.

Participants

A total of 1293 Colombian men between 18 and 65 years of age (mean = 24.94, SD = 4.29) participated in the study. Participants were recruited using a non-probability convenience and snowball sampling strategy (Espinoza Tamez et al. 2018). The inclusion criteria required that participants be over 18 years of age, born and residing in Colombia, and in a cohabitating or non-cohabitating relationship for more than two years at the time of participation. The participants had never reported intimate partner violence, did not identify themselves as victims, and were not listed in official victim registries. In accordance with ethical standards, the research instrument commenced with an informed consent form. Although participation was voluntary, if the questions prompted reflection, discomfort, or a desire to seek psychological services, email addresses were provided for participants to contact us. Detailed demographic characteristics of the participants are presented in

Table 1. Sociodemographic characteristics of participants.

Educational Level	f	%	% Valid
No education	6	0.5	0.5
Elementary school	41	3.2	3.2
High School	291	22.5	22.5
Technical/Technologist	250	19.3	19.3
University	316	24.4	24.4
Postgraduate	389	30.1	30.1
Occupation	f	%	% valid
Worker	404	31.2	77.0
Unemployed	45	3.5	8.6
Retired	8	0.6	1.5
Student	68	5.3	13.0
No data	768	59.4
Socioeconomic Level	f	%	% Valid
Low	346	26.8	26.8
Medium	647	50.0	50.0
High	300	23.2	23.2
Marital status	f	%	% valid
Single	290	22.4	29.3
Married	365	28.2	36.9
Common-law	324	25.1	32.8
Divorced	6	0.5	0.6
Widowed	4	0.3	0.4
No data	304	23.5

Note: These demographic data provide information about the participants in the study. No analyses were conducted based on occupation or marital status due to missing data.

.

Instruments

The Intimate Partner Violence Scale (EV, Valdez-Santiago et al. 2006) consists of 27 items scored on a four-point Likert scale (1 = never; 4 = always), which assesses various types of violence (psychological, physical, economic, and sexual). For this study, the version previously adapted in Colombia for use with women was utilized (Acuña-Arango et al. forthcoming), which included two new items designed to capture economic violence, resulting in a total of 29 items.

Procedure

This study adheres to the guidelines of the American Educational Research Association, the American Psychological Association, and the National Council on Measurement in Education (2018). To gather evidence of content validity, an expert judge validation process was conducted to assess the suitability of the questions for the Colombian context and to verify the categorization of items within each IPV dimension.

The scale was evaluated by eight judges divided into two groups. The first group consisted of four professional experts (men and women specializing in victim services, couple counselors, a semantics specialist, and academics), and the second group comprised four experts by experience (men who were victims of IPV).

To evaluate the severity of the different forms of violence, the judges were asked to rate each question across two dimensions: (1) emotional harm and (2) physical harm, using a scale from 1 to 10, where 1 corresponded to the lowest level of harm and 10 to the highest. The scores assigned by the judges, victims, and professional experts yielded averages above 6.0 for all forms of violence, except for the following items: Has she become jealous or suspicious of your friends? And has she looked down on him or her as a parent because of his lower economic status? These items received average scores below 5.0; however, they were retained in the quantitative analysis phase to understand how these questions behaved in the factor analysis.

In addition, the four professional expert judges were asked to rate the clarity, pertinence, relevance, and adequacy of each item on a four-point Likert scale as follows: (1) Strongly disagree: The item does not meet the evaluation criterion, contains significant shortcomings, and requires a complete rewrite. (2) Disagree: The item does not sufficiently meet the evaluation criteria; it has significant weaknesses that require adjustments. (3) Agree: The item meets the evaluation criterion, although it could be slightly improved; it is relevant and functional. (4) Strongly agree: The item fully meets the criterion being assessed, and its wording is appropriate, precise, and consistent with the purpose of the instrument. The raters assigned scores between 3 and 4 to all items across the various forms of violence.

After adjusting based on this evaluation, the questionnaire was deployed in the Google Forms link distributed through a “snowball” sampling method among the researchers’ networks, who shared it with other potential participants (Espinoza Tamez et al. 2018). The instrument commenced with the informed consent form. Once data collection was completed, a database was generated, and statistical analyses were performed.

As evidence of validity from the internal structure, after verifying sample size assumptions with the Kaiser-Meyer-Olkin (KMO) test, correlation matrix adequacy with Bartlett’s sphericity test, and non-collinearity (correlations less than 0.90), a confirmatory factor analysis was performed, replicating the structures obtained in previous studies (four factors and five factors), using a diagonalized weighted least squares method, with polychoric matrices, latent variable typification, and robust analysis to compensate for the lack of multivariate normality and given the ordinal nature of the items (Lloret-Segura et al. 2014). Adjustment indicators greater than 0.95, and error indices equal to or less than 0.08 were expected.

In evidence of validity from convergence, extracted variance indices were obtained with expected values of 0.4 or higher. Reliability analyses were also performed: Cronbach’s alpha, McDonald’s omega, Guttman’s lambda 2, and split-half coefficient for the general test and for factors, as well as simulations when eliminating items, item-item correlations, and item-test correlations. All analyses were performed using JASP statistical software (version 0.95).

3. Results

Table 2. Goodness-of-fit indices and error of the tested factorial models.

Indicator	Model 1: Conceptual (Four Factors)	Model 2: Mexican Model	Model 3: Colombian Women
Comparative Fit Index (CFI)	0.997	0.997	0.998
Tucker–Lewis Index (TLI)	0.997	0.996	0.998
Bentler-Bonett Non-Normalized Fit Index (NNFI)	0.997	0.996	0.998
Bentler-Bonett Normalized Fit Index (NFI)	0.977	0.976	0.978
Bollen Relative Fit Index (RFI)	0.975	0.973	0.976
Bollen Incremental Fit Index (IFI)	0.997	0.997	0.998
Goodness-of-Fit Index (GFI)	0.981	0.981	0.982
Root Mean Square Error of Approximation (RMSEA)	0.010	0.011	0.009
RMSEA 90% CI lower bound	0.002	0.003	0.000
RMSEA 90% CI upper bound	0.015	0.016	0.014
Standardized Root Mean Square Error (SRMR)	0.091	0.087	0.080

presents the goodness-of-fit and error indicators from the confirmatory factor analyses. The evaluated models replicate distributions based on conceptual and legal classifications (Acosta Vélez 2016; Corte Constitucional (CC) 2014), as well as the factor structures from previous validation studies (Valdez-Santiago et al. 2006) and the analysis of competing models. These rival models were structured as follows: Model 1, a four-dimensional framework based on local conceptual and legal definitions (psychological, physical, economic, and sexual violence); Model 2, the four-dimensional structure originally designed and validated in Mexico (psychological, physical, severe physical, and sexual violence); and Model 3, a five-factor structure previously validated with Colombian women (psychological, physical, severe physical, economic, and sexual violence).

As shown in Table 2, the goodness-of-fit indices are adequate for the three models analyzed, although in the first two models, one error indicator exceeds the expected value (SRMR < 0.08). From this, it can be inferred that the five-factor structure (psychological, physical, severe physical, economic, and sexual violence) is the dimensionality with the best fit, and that it also includes all the dimensionalities in their entirety.

Table 3. Factor loadings of the items in each dimension for each model.

Model 1: Conceptual Model				Model 2: Mexican Model				Model 3: Colombian Women
Factor	Item	Factor Loading	p	Factor	Item	Factor Loading	p	Factor	Item	Factor Loading	p
PS V	PS V1 Has he/she put you down or belittled you?	0.567	<0.001	PS V	PS V1 Has he/she put you down or belittled you?	0.565	<0.001	PS V	PS V1 Has he/she put you down or belittled you?	0.570	<0.001
	PS V2 Has he/she belittled or humiliated you in front of others?	0.491	<0.001		PS V2 Has he/she belittled or humiliated you in front of others?	0.487	<0.001		PS V2 Has he/she belittled or humiliated you in front of others?	0.493	<0.001
	PS V3 Has he/she insulted you?	0.549	<0.001		PS V3 Has he/she insulted you?	0.553	<0.001		PS V3 Has he/she insulted you?	0.552	<0.001
	PS V4 Has he/she become jealous or suspicious of your friends?	0.503	<0.001		PS V4 Has he/she become jealous or suspicious of your friends?	0.510	<0.001		PS V4 Has he/she become jealous or suspicious of your friends?	0.505	<0.001
	PS V5 Has he/she told you things like you are unattractive or ugly?	0.342	<0.001		PS V5 Has he/she said things like you are unattractive or ugly?	0.335	<0.001		PS V5 Has he/she said things like you are unattractive or ugly?	0.342	<0.001
	PS V6 Has he/she reacted by hitting a wall or piece of furniture in an argument?	0.482	<0.001		PS V6 Has he/she reacted by hitting a wall or piece of furniture in an argument?	0.487	<0.001		PS V6 Has he/she reacted by hitting a wall or piece of furniture in an argument?	0.483	<0.001
	PS V7 Has he/she destroyed any of your things?	0.415	<0.001		PS V7 Has he/she destroyed any of your things?	0.414	<0.001		PS V7 Has he/she destroyed any of your things?	0.416	<0.001
	PS V8 Has he/she threatened to hit you?	0.403	<0.001		PS V8 Has he/she threatened to hit you?	0.411	<0.001		PS V8 Has he/she threatened to hit you?	0.406	<0.001
	PS V9 Has he/she threatened you with a knife, a machete, or a pocket knife?	0.114	<0.001		PS V11 Has he/she made you feel afraid?	0.479	<0.001		PS V11 Has he/she made you feel afraid?	0.477	<0.001
	PS V10 Has he/she threatened you with a gun?	0.041	<0.001		PHY V 13 Has he/she deliberately pushed you?	0.518	<0.001		PHY V 13 Has he/she deliberately pushed you?	0.508	<0.001
	PS V11 Has he/she made you feel afraid?	0.476	<0.001		PS V12 Has he/she threatened to kill you or kill himself/herself?	0.210	<0.001		PS V12 Has he/she threatened to kill you or kill himself/herself?	0.215	<0.001
PHY V	PHY V 13 Has he/she pushed you on purpose?	0.539	<0.001	PHY V	PS V9 Has he/she threatened you with a knife, a machete, or a pocket knife?	0.123	<0.001	PHY V	PS V9 Has he/she threatened you with a knife, a machete, or a pocket knife?	0.123	<0.001
	PHY V 14 Has he/she shaken you, pulled you, or pulled you?	0.532	<0.001		PHY V 14 Has he/she shaken, pulled, or pulled you?	0.540	<0.001		PHY V 14 Has he/she shaken, pulled, or pulled you?	0.532	<0.001
	PHY V 15 Has he/she twisted your arm?	0.232	<0.001		PHY V 15 Has he/she twisted your arm?	0.228	<0.001		PHY V 15 Has he/she twisted your arm?	0.234	<0.001
	PHY V 16 Has he hit you with his hand or fist?	0.494	<0.001		PHY V 16 Has he/she hit you with his/her hand or fist?	0.500	<0.001		PHY V 16 Has he/she hit you with his hand or fist?	0.494	<0.001
	PHY V 17 Has he/she kicked you?	0.193	<0.001		PHY V 17 Has he/she kicked you?	0.193	<0.001		PHY V 17 Has he/she kicked you?	0.195	<0.001
	PHY V 18 Has he/she hit you with a stick, belt, or other household object?	0.411	<0.001		PHY V 18 Has he/she hit you with a stick, belt, or other household object?	0.414	<0.001		PHY V 18 Has he/she hit you with a stick, belt, or other household object?	0.412	<0.001
	PHY V 19 Has he/she burned you with a cigarette or other substance?	0.026	0.003		PHY V 20 Has he/she tried to choke or suffocate you?	0.139	<0.001		PHY V 20 Has he/she tried to choke or suffocate you?	0.143	<0.001
	PHY V 20 Has he/she tried to choke or suffocate you?	0.142	<0.001		PHY V 21 Has he/she attacked you with a knife, blade, or machete?	0.069	<0.001		PHY V 21 Has he/she attacked you with a knife, a machete, or a knife?	0.071	<0.001
	PHY V 21 Has he/she attacked you with a knife, blade, or machete?	0.070	<0.001	PHY VS	PS V10 Has he/she threatened you with a gun?	0.141	<0.001	EV	EV 23 Has he/she controlled you by withholding money or taking it away from you?	0.531	<0.001
	PHY V 22 Has he/she shot you with a gun?	0.013	0.064		PHY V 19 Has he/she burned you with a cigarette or other substance?	0.090	0.002		EV 24 Has he/she taken or used your belongings against your will?	0.400	<0.001
EV	EV 23 Has he/she controlled you by withholding money or taking it away from you?	0.531	<0.001		PHY V 22 Has he/she shot you with a gun?	0.045	0.037		EV 28 Has he/she belittled you because of your lower economic status?	0.403	<0.001
	EV 24 Has he/she taken or used your belongings against your will?	0.400	<0.001	SV	SV 25 Has he/she demanded sex?	0.394	<0.001		EV 29 Has he/she belittled you in his role as a parent because of your lower economic status?	0.377	<0.001
	EV 28 Has he/she looked down on you because of your lower economic status?	0.403	<0.001		SV 26 Has he/she threatened to leave you with other men (women) if you don’t agree to have sex with him/her?	0.393	<0.001	PHY SV	PS V10 Has he/she threatened you with a gun?	0.145	<0.001
	EV 29 Has he/she looked down on you in your role as a parent because of your lower economic status?	0.377	<0.001		SV 27 Has he/she used physical force to have sex?	0.111	<0.001		PHY V 19 Has he/she burned you with a cigarette or some other substance?	0.087	0.002
SV	SV 25 Has he/she demanded sexual relations with you?	0.393	<0.001						PHY V 22 Has he/she fired a gun?	0.047	0.033
	SV 26 Has he/she threatened to leave you with other men (women) if you don’t agree to have sexual relations with him/her?	0.385	<0.001					SV	SV 25 Has he/she demanded sexual relations with you?	0.393	<0.001
	SV 27 Has he/she used physical force to have sexual relations?	0.116	<0.001						SV 26 Has he/she threatened to leave you with other men (women) if you don’t agree to have sexual relations with him/her?	0.386	<0.001
									SV 27 Has he/she used physical force to have sexual relations with you?	0.116	<0.001

Note: PS V = psychological violence; PHY V = physical violence; PHY SV = severe physical violence; EV = economic violence; SV = sexual violence.

shows the loadings of the items for each estimated factor.

Table 3 presents the items loaded by the factor to which they belong, and shows that all loadings are significant (p < 0.005), although in all models, the loadings are low in the items representing severe violence; higher loadings were found in model 3. Reviewing the frequency of violent events, the rates are low, which explains why they have lower loading factors.

Table 4. Average variance extracted from each model.

Model 1: Conceptual Model		Model 2: Mexican Model		Model 3: Colombian Women
Factor	AVE	Factor	AVE	Factor	AVE
Psychological	0.489	Psychological	0.516	Psychological	0.515
Physical	0.598	Physical	0.565	Physical	0.557
Economic	0.574	Severe Physical	0.425	Economic	0.574
Sexual	0.597	Sexual	0.608	Severe Physical	0.440
				Sexual	0.598

shows the mean variance extracted indicators in each model tested.

Regarding the mean variances for each model (Table 4), in all cases, the values exceed 40% and, in general, they are better in the five-factor model (model 3 that was previously tested in women). Upon reviewing the extracted variances, factor loadings, and factorial fit indicators, and subsequently examining the comprehensive model, which incorporates all dimensions from the preceding models—thus contributing to content validity (psychological, physical, severe physical, economic, and sexual violence)—it is concluded that the optimal structure is that of Model 3. Based on this model, the reliability analyses shown in

Table 5. Reliability indicators for the five-factor model.

	Ω Coefficient	α Coefficient	λ2 Coefficient	Coefficient of Division into Halves
General Scale	0.955	0.946	0.956	0.969
Psychological violence	0.921	0.915	0.920	0.928
	ω coefficient (if item deleted)	α coefficient (if item deleted)	λ2 coefficient (if item deleted)	Item-rest correlation
PS V1 Has he/she put you down or belittled you?	0.908	0.902	0.907	0.768
PS V2 Has he/she belittled or humiliated you in front of others?	0.910	0.903	0.909	0.757
PS V3 Has he/she insulted you?	0.908	0.902	0.907	0.778
PS V4 Has he/she become jealous or suspicious of your friends?	0.918	0.912	0.917	0.632
PS V5 Has he/she told you things like you are unattractive or ugly?	0.918	0.912	0.917	0.576
PS V6 Has he/she reacted by hitting a wall or piece of furniture in an argument?	0.913	0.906	0.912	0.706
PS V7 Has he/she destroyed any of your things?	0.915	0.908	0.914	0.673
PS V8 Has he/she threatened to hit you?	0.920	0.914	0.919	0.560
PS V11 Has he/she made you feel afraid?	0.912	0.905	0.911	0.730
PS V12 Has he/she threatened to kill you or kill himself/herself?	0.919	0.916	0.919	0.495
PHY V 13 Has he/she deliberately pushed you?	0.910	0.903	0.909	0.762
	Ω coefficient	α coefficient	λ2 coefficient	Coefficient of division into halves
Physical violence	0.885	0.843	0.877	0.773
	ω coefficient (if item deleted)	α coefficient (if item deleted)	λ2 coefficient (if item deleted)	Item-rest correlation
PS V 9 Has he/she threatened you with a knife, a machete, or a pocket knife?	0.883	0.837	0.872	0.490
PHY V 14 Has he/she shaken you, pulled you, or pulled you?	0.845	0.806	0.836	0.738
PHY V 15 Has he/she twisted your arm?	0.879	0.825	0.867	0.584
PHY V 16 Has he hit you with his hand or fist?	0.835	0.800	0.830	0.760
PHY V 17 Has he/she kicked you?	0.878	0.825	0.866	0.596
PHY V 18 Has he/she hit you with a stick, belt, or other household object?	0.855	0.806	0.844	0.713
PHY V 20 Has he/she tried to choke or suffocate you?	0.882	0.836	0.872	0.498
PHY V 21 Has he/she attacked you with a knife, blade, or machete?	0.883	0.846	0.875	0.395
	Ω coefficient	α coefficient	λ2 coefficient	Coefficient of division into halves
Severe physical violence	0.833	0.786	0.809	-
	ω coefficient (if item deleted)	α coefficient (if item deleted)	λ2 coefficient (if item deleted)	Item-rest correlation
PS V10 Has he/she threatened you with a gun?	-	0.624	0.624	0.717
PHY V 19 Has he/she burned you with a cigarette or some other substance?	-	0.836	0.836	0.507
PHY V 22 Has he/she fired a gun?	-	0.593	0.593	0.729
	Ω coefficient	α coefficient	λ2 coefficient	Coefficient of division into halves
Economical violence	0.847	0.851	0.855	-
	ω coefficient (if item deleted)	α coefficient (if item deleted)	λ2 coefficient (if item deleted)	Item-rest correlation
EV 23 Has he/she controlled you by withholding money or taking it away from you?	0.854	0.831	0.841	0.653
EV 24 Has he/she taken or used your belongings against your will?	0.834	0.825	0.828	0.659
EV 28 Has he/she belittled you because of your lower economic status?	0.804	0.795	0.798	0.725
EV 29 Has he/she belittled you in his role as a parent because of your lower economic status?	0.798	0.790	0.792	0.741
	Ω coefficient	α coefficient	λ2 coefficient	Coefficient of division into halves
Sexual violence	0.770	0.710	0.736	-
	ω coefficient (if item deleted)	α coefficient (if item deleted)	λ2 coefficient (if item deleted)	Item-rest correlation
SV 25 Has he/she demanded sexual relations with you?	-	0.517	0.517	0.629
SV 26 Has he/she threatened to leave you with other men (women) if you don’t agree to have sexual relations with him/her?	-	0.518	0.518	0.607
SV 27 Has he/she used physical force to have sexual relations with you?	-	0.737	0.737	0.452

were conducted.

As shown in Table 5, all overall and subfactor reliability indicators ranged from acceptable to excellent. Item-deletion simulations indicated that the removal of any single item negatively affects the reliability of the respective factor, and the item-total correlations exceeded 0.30. Taken together, these findings demonstrate that the scale possesses robust evidence of reliability and internal consistency.

4. Discussion

The objective of this study was to adapt the Violence Scale (VS) originally developed by Valdez-Santiago et al. (2006) and to examine its validity and reliability within the Colombian context using male samples. The overarching aim was to provide a robust instrument capable of promptly identifying intimate partner violence against men, thereby enhancing its visibility, scholarly understanding, and clinical response in the country. This endeavor is particularly relevant given that, in the case of male victims, existing literature has tended to emphasize barriers to accessing services rather than acknowledging the phenomenon as a pressing social issue. Such neglect is closely tied to the historical reality that women have been disproportionately affected by intimate partner violence, a circumstance that has inadvertently contributed to the marginalization of male victims.

The findings yielded a five-dimensional factor structure (comprising psychological, physical, severe physical, economic, and sexual violence) that comprehensively covers the various types and forms of violence described in the theoretical framework (Acosta Vélez 2016; Corte Constitucional (CC) 2014), thereby providing strong evidence of content validity. This model demonstrated excellent goodness-of-fit indices and low error indicators, alongside adequate factor loadings across all items, which indicates appropriate evidence of validity from the internal structure. Furthermore, mean variance extracted indices greater than 0.40 in each dimension, which refers to evidence of convergent validity and reliability with adequate and high values (generally greater than 0.90, and for four factors, values greater than 0.80, and for the remaining factors, values greater than 0.70). In turn, simulations of item deletion negatively affect reliability indicators, and significant correlations were also found between the items of each factor, which demonstrate adequate internal consistency. Taken together, this study provides adequate preliminary evidence regarding the validity and reliability of the Violence Scale (EV, Valdez-Santiago et al. 2006) for assessing intimate partner violence against men in this sample of men in the Colombian context.

These results align with both Valdez-Santiago et al. (2006) and Acuña-Arango et al. (forthcoming) by successfully differentiating between physical violence in relation to non-fatal injuries and those in which the victim’s life is placed at risk. Conversely, the findings differ from the EV by Valdez-Santiago et al. (2006) in that the Mexican study failed to highlight economic violence, an issue that was successfully addressed by the study by Acuña-Arango et al. (forthcoming) with women and successfully replicated in the present study with a sample of men through the inclusion of two newly designed items. Consequently, the Mexican study by Valdez-Santiago et al. (2006) yielded a four-factor model, and the Colombian study by Acuña-Arango et al. (forthcoming) obtained five.

One difference between these studies lies in how economic violence behaved. In the Colombian context, these behaviors cleanly loaded onto their own distinct factor, whereas in the Mexican validation, they grouped unexpectedly with sexual violence. In this regard, Marshall’s (1992) theoretical contributions are highly relevant, as they demonstrate that the specific behavioral manifestations of a type of violence can vary significantly depending on the unique sociocultural context in which individuals develop.

Another distinguishing feature across these studies is that Valdez-Santiago et al. (2006) conducted their research with a sample of women who had reported violence. In contrast, both Colombian studies utilized community samples of men and women who were included specifically because they did not perceive themselves as victims of intimate partner violence and, consequently, had no prior record of filing reports or complaints. This leads us to reflect on the normalization and justification of violence based on cultural factors in men and women. This study helps shed light on the phenomenon and raise awareness of the problem.

Although some items displayed low factor loadings, they remained statistically significant. Content analysis revealed that these items capture severe forms of physical violence, which are inherently characterized by low frequency and restricted variance, thereby explaining the lower magnitude of the loadings. Nevertheless, the global fit indices of the CFA, coupled with adequate reliability and internal consistency estimates, support the model’s acceptability. Additionally, their retention is conceptually warranted given their theoretical contribution to the measurement of severe physical violence. It is important to clarify this on two levels: psychometric and social. At the psychometric level, the findings confirm that these items remain valid, reliable, and sufficiently sensitive for assessing severe physical violence in community settings where low rates of occurrence are expected; but, on the social level, the situation is alarming because, although it was not the direct objective of the study, significant prevalences of severe physical violence were found in these community samples. This unexpected finding highlights the profound normalization of violence, emphasizing the urgent need to develop measures that track the frequency of severe violence in non-clinical contexts and to promote targeted prevention strategies given the high stakes of these behaviors. Future research should replicate these findings with larger and more diverse samples to strengthen external validity.

Although the rates of violence among men are low, it is important to highlight their victimization by their partners in order to explore new avenues for analyzing this phenomenon, as proposed by Christiansen (2014), Rojas-Solís et al. (2019), and Armas-Vargas (2020), and, consequently, to avoid biases that view men solely as perpetrators of intimate partner violence, given that historically, in most cases, the victim is a woman.

In this regard, this study contributes by providing a scale—with adequate evidence of initial validity and reliability—to assess intimate partner violence against men, with contributions in two key areas. On the one hand, given that the scale can be applied in a unified manner in social services by technical and care personnel in the field, it facilitates early warning and, consequently, timely intervention. On the other hand, it opens up opportunities to conduct studies in specific sociocultural contexts that support public policy lines that encourage joint action between the state, civil society, and families, allowing for an understanding of the phenomenon from the sociocultural perspective of gender stereotypes that some authors have recommended in recent years (Ramasubramani et al. 2024; Jónasdóttir et al. 2024; Rojas-Solís et al. 2019). In response to IPV being recognized as a global health problem, advances have been made in theoretical and methodological developments, care strategies, and the design and implementation of public policies. Progress has also been made in actions that promote cultural change toward more symmetrical relationships between men and women. However, research on intimate partner violence against men helps inform the development of prevention strategies that benefit both genders, even though, historically, intimate partner violence has primarily affected women (Walker et al. 2020; Rojas-Solís et al. 2019; Alegría del Ángel and Rodríguez Barraza 2015).

Another critical contribution of this scale is that its use in state services addressing intimate partner violence opens up an opportunity to establish differentiated objectives in programs for perpetrators (men and women) that can enhance the effective use of investment resources aimed at educational and clinical interventions, as other authors have suggested (Expósito-Álvarez et al. 2023; Akl Moanac et al. 2022; Heise et al. 1999). Finally, as demonstrated in this study, there is now a scale available that allows for the assessment of intimate partner violence in both men and women in specific contexts, which can be replicated in other Latin American countries.

The results of this study suggest that future research should address the cultural, socioeconomic, and family aspects that may influence the normalization and tolerance of IPV (Expósito-Álvarez et al. 2023; Arroyave López et al. 2022; Akl Moanac et al. 2022), within specific sociocultural contexts, as well as within samples of same-sex couples. These findings contribute to the broader literature on gender while providing an empirical rationale for intersectoral policy interventions. This aligns with Khofi et al. (2025), who argue that comprehensive actions involving the social welfare, education, and health sectors, among others, are essential to effectively address violence and promote justice across genders.

One limitation of the study was the sample selection process. Because certain cultural constructs surrounding masculinity often hinder men’s participation in this type of research, the reliance on non-probability sampling introduced inherent selection biases. Similarly, geographic and demographic limitations were present; participants were recruited predominantly from the central region of Colombia and its major urban centers, resulting in an overrepresentation of urban perspectives. Furthermore, significant missing data regarding participants’ marital status and occupation introduced additional constraints on the generalizability of these findings. Furthermore, although participants were informed that they would have access to psychological support via email once they had completed the questionnaire, one limitation was the lack of localized information on accessible care centers, especially in the country’s major urban centers.

While these results are acceptable for an initial validation study, future replication efforts should expand the sample criteria. Researchers should ideally utilize probabilistic or stratified sampling methods to ensure greater representation across diverse regions, age brackets, educational levels, socioeconomic statuses, and marital structures, with a particular emphasis on including rural populations. Additionally, future studies should conduct factorial invariance analyses across these demographic variables—an aspect not developed in the present study—as well as longitudinal internal consistency assessments using test–retest indicators.

A clear age imbalance was also observed, as most of the participants were young adults; therefore, future research must replicate these results with older populations. Finally, while the Average Variance Extracted (AVE) metrics provided sound evidence of internal convergence, studies are needed to evaluate convergent validity using external measures. These could include concurrent validation against other violence questionnaires, instruments assessing related constructs such as aggression or impulsivity, or known-groups divergent validity testing between men who formally report violence and those who do not.

5. Conclusions

Intimate partner violence is often viewed as a gender-based issue that primarily affects women, which frequently leads to male victims being overlooked: violence by women against men is neither sufficiently understood nor researched. This makes it difficult for men to report incidents and for awareness of this phenomenon to grow and can result in a lack of assistance and support for one gender. We need to better understand this phenomenon by taking both male and female victims into account to ensure that intimate partner violence is addressed from a comprehensive and balanced perspective, for both women and men.

This study offers a scale that makes it possible to assess intimate partner violence and gather information on the characteristics and support needs of male victims, while also providing a foundation and guidance for the design of strategies, support services, and care, thereby contributing to the recognition of the phenomenon as a social and human rights issue.