Brazilian Portuguese Version of Domestic Violence Myth Scale Passes First Psychometric Tests

Domestic violence against women remains one of the most pervasive public health problems in the world, and the beliefs that quietly sustain it are often harder to measure than the violence itself. Myths about abuse—ideas that victims provoke their attackers, that abuse only happens in certain families, or that victims could easily leave—are not fringe opinions. Decades of research show that these legitimizing beliefs shape how bystanders respond, how juries decide, how police investigate, and even how survivors understand their own experiences. A new study published in BMC Psychology has now taken a significant step toward making those beliefs measurable in Brazil, one of the most populous and culturally diverse countries in the world, by culturally adapting and psychometrically evaluating the Domestic Violence Myth Acceptance Scale, known in the literature as the DVMAS, for Brazilian Portuguese speakers.

The research team, led by Eduardo Azevedo Rebouças of the Federal University of Amazonas together with Marck de Souza Torres, Adriane de Almeida Santos, and Daniela Castro dos Reis of the Federal University of Pará, carried out a two-part investigation. The first part was a rigorous cross-cultural adaptation of the original instrument, and the second was a large-scale validation study involving 504 Brazilian adults drawn from every region of the country. The distinction matters: simply translating a questionnaire word for word rarely preserves its meaning, because myths about domestic violence are embedded in cultural narratives about gender, family, and responsibility that differ sharply between societies. A scale that works in one language can silently fail in another if idioms, social assumptions, or response conventions do not transfer.

To avoid that trap, the researchers followed a structured adaptation process spanning six phases. Independent translations of the scale were produced, then evaluated by a panel of expert judges who rated each item for clarity, relevance, and cultural appropriateness using content validity indices at both the item and scale levels. A pilot test with members of the target population followed, allowing the team to identify wording that sounded foreign or confusing to ordinary Brazilian respondents. Finally, the adapted version was back-translated into the original language and approved by the original author, a safeguard that ensures the adapted items still capture the psychological construct the scale was designed to measure. Only after this painstaking sequence did the team move to the validation stage.

The validation sample was broad and demographically varied: 504 participants aged 18 to 80, with an average age of 30, including 52 percent women, 46.4 percent men, and 1.6 percent nonbinary individuals, recruited through an online form using convenience sampling. Crucially, respondents came from all five regions of Brazil, from the Amazonian north to the southern states, giving the instrument a genuinely national footing rather than a narrow urban or regional one. The participants completed the adapted scale, and the researchers then subjected the data to a battery of modern psychometric techniques that go well beyond the classical toolkit.

At the heart of the analysis was confirmatory factor analysis, a statistical method that tests whether the pattern of responses matches a hypothesized structure—in this case, a single underlying dimension of myth acceptance. The researchers used a robust weighted least squares estimator, appropriate for ordinal response data, and the unifactorial model with residual covariances between semantically similar items showed excellent fit: a comparative fit index of 0.971 and a Tucker-Lewis index of 0.966, both comfortably above conventional thresholds, along with a root mean square error of approximation of 0.066 and a standardized root mean residual of 0.051. In practical terms, the scale behaves as if it is measuring one coherent attitude rather than a grab bag of unrelated opinions.

Internal consistency, the degree to which items reliably tap the same construct, was equally strong, with both Cronbach’s alpha and McDonald’s omega reaching 0.90, a level generally considered excellent for a research instrument. But the team did not stop there. Recognizing that high internal consistency can sometimes mask multidimensionality, they deployed network analysis, including exploratory graph analysis and unique variable analysis, alongside a bifactor model. These complementary approaches converged on what psychometricians call essential unidimensionality: the explained common variance was 0.782 and the omega hierarchical statistic reached 0.937, indicating that a single dominant factor explains nearly all of the reliable variance. The residual covariances between certain items reflected local dependence—items that were semantically overlapping tended to correlate more strongly—not the existence of distinct subdimensions. This is an important technical finding, because it tells future researchers they can confidently use a total score without worrying that they are averaging across fundamentally different constructs.

The study also examined measurement invariance across genders, asking whether the scale functions the same way for men and women, a prerequisite for meaningful group comparisons. The results supported partial metric and partial scalar invariance, meaning that most items operate equivalently across genders even if a few do not. With that caution in place, the researchers compared groups and found a pattern consistent with international literature: men endorsed significantly more myths than women, with mean scores of 2.44 versus 1.90 and an effect size corresponding to a correlation of −0.30. Within the male sample, married or cohabiting heterosexual men showed the highest levels of myth endorsement, a demographic signal that could prove valuable for targeting prevention programs.

Perhaps the most sobering finding was what did not change. Even among participants who reported direct experience of violence, myth acceptance remained high. The authors interpret this as evidence that these beliefs operate as cultural repertoires—shared scripts absorbed from family, media, and social environments—that are resistant to change even when personal experience contradicts them. This resilience underscores why measurement tools matter: interventions designed to reduce victim-blaming need a reliable way to assess baseline attitudes and to evaluate whether educational programs, awareness campaigns, or judicial training actually shift them. Without a validated scale, such evaluations rest on anecdote rather than evidence.

The researchers are careful about the limits of their work, and their caution is itself instructive. Because convergent and discriminant validity were not assessed, and because gender invariance was only partial, they describe the Brazilian version as culturally adapted with initial validity evidence rather than fully validated. That framing is honest and appropriate for a first national adaptation, and it sets a clear agenda for follow-up studies, including tests of convergent validity against related constructs such as ambivalent sexism, longitudinal stability, and invariance across other demographic boundaries. The work was supported by Brazilian funding agencies, including the Amazonas State Research Support Foundation and the National Council for Scientific and Technological Development, reflecting a growing investment in violence-prevention research in Latin America.

For Brazil, where millions of women report domestic violence and where the Maria da Penha Law and a national network of services have made gender-based violence a central policy issue, the arrival of a psychometrically sound Portuguese-language measure of myth acceptance fills a real gap. It gives researchers a common yardstick, gives intervention designers an outcome measure, and gives policymakers a way to track whether public attitudes are shifting over time. More broadly, the study is a model of how cross-cultural scale adaptation should be done: transparent phases, expert judgment, back-translation, robust modern statistics, and honest reporting of what remains unproven. As the authors note, the instrument is already promising for research and for evaluating interventions aimed at domestic and gender-based violence—and it arrives at a moment when the fight against those myths has never been more urgent.

Subject of Research: Cross-cultural adaptation and psychometric validation of the Domestic Violence Myth Acceptance Scale in Brazil

Article Title: The domestic violence myth acceptance scale in Brazil: cross-cultural adaptation and initial construct-validity evidence

Article References: Rebouças, E. A., de Souza Torres, M., de Almeida Santos, A., & dos Reis, D. C. (2026). The domestic violence myth acceptance scale in Brazil: cross-cultural adaptation and initial construct-validity evidence. BMC Psychology. https://doi.org/10.1186/s40359-026-05518-x

Image Credits: AI Generated

DOI: 10.1186/s40359-026-05518-x

Keywords: domestic violence, myth acceptance, psychometrics, scale validation, Brazil, cross-cultural adaptation, confirmatory factor analysis, measurement invariance, gender-based violence, network analysis, BMC Psychology, public health