Behind Closed Doors
Intimate Partner Violence in Times of Crisis
This data visualization report explores national trends in Intimate Partner Violence (IPV), highlighting how rates have changed over time and across populations. The project draws on publicly available data and presents it in an interactive Tableau dashboard designed to make complex information accessible and engaging. While completing this report as part of a broader research paper on IPV, I honed my data visualization skills—refining my ability to clean, structure, and present social data in a way that supports both exploration and insight. This work not only deepened my understanding of the social dynamics surrounding IPV but also strengthened my practice in transforming raw data into meaningful, user-centered narratives.
Outcome Abstract
This project investigates the prevalence, patterns, and social implications of Intimate Partner Violence (IPV) using data visualization methods. By collecting and analyzing public datasets from organizations such as the CDC and WHO, we aimed to identify trends in IPV incidents across various demographics, geographic areas, and periods. Utilizing tools like Tableau, we created both interactive and static visualizations that highlight significant insights, including the relationships between IPV rates and socioeconomic factors. The results illustrate how intentional design choices—such as appropriate chart types, color palettes, and contextual framing—can enhance comprehension and strengthen advocacy efforts. This project improved our technical skills in data visualization and emphasized the importance of presenting sensitive data ethically to inform policy and raise awareness.
Introduction
For over thirty-five years, intimate partner violence (IPV) has emerged as a significant public health and social justice concern worldwide. IPV is defined as physical, sexual, or psychological harm inflicted by a current or former partner or spouse, transcending geographic, cultural, and socioeconomic boundaries. Since the late 20th century, increased awareness, legal reforms, and advocacy efforts have heightened attention to IPV, prompting researchers and policymakers to monitor its prevalence and impact closely. However, the trajectory of IPV rates over the past thirty years presents a complex narrative, marked by advancements and persistent challenges. These charts illustrate historical patterns of IPV prevalence, highlight key shifts influenced by social, legal, and cultural changes, and evaluate the implications for current prevention and intervention strategies.
Previous Work
To provide an understanding of our previous work, our group investigated various datasets from different sources early on. This included data from the Bureau of Justice Statistics, Centers for Disease Control and Prevention, National IPV Survey, and more. These sources laid the groundwork for how we would structure our findings to conduct a well-rounded exploration of various demographics that experienced a form of assault under the pretense of IPV. In our exploratory analysis, we outlined our four guiding questions that informed the initial stages of our process. The questions addressed are the following:
How did rates of IPV change during major crises? Examples may include the COVID-19 pandemic, the Great Recession of 2008, or other crises.
Are there any demographic differences in IPV rates that became more pronounced during a crisis?
How has the severity changed over time during the 21st century, and do these changes align with known national crises?
Is there a correlation between rising IPV incidence rates and mortality rates during specific crisis periods?
Our investigation revealed that during specific periods of great strife, there was a significant increase in IPV, likely driven by factors such as social isolation, economic hardship, and heightened psychological stress. An analysis of demographic patterns by gender and race indicated that specific populations were disproportionately affected, underscoring the importance of targeted prevention and intervention strategies. Additionally, the correlation between rising IPV cases and increased mortality during crisis periods suggests that such times not only elevate the incidence of IPV but also the severity of its outcomes. Our visualization of our analysis of the detailed results are available in Appendix. Our group considered how statistical data on IPV and mortality may not always align with timelines, emphasizing the need for deeper analysis to understand the full context. These various thoughts led to further brainstorming about our project on intimate partner violence (IPV).
This interactive Tableau visualization presents the IPV (Intimate Partner Violence) rate over time, allowing users to explore changes in IPV prevalence across different years. By engaging with the chart, viewers can observe trends and identify periods of increase or decrease, which complements the analysis in my research paper. The data visualization provides an accessible way to understand the scope of IPV, offering key insights into factors that may influence violence rates and highlighting opportunities for targeted prevention and policy efforts.
Design process including milestones, with ample sketches, redesigns/iterations
The design process for developing Tableau visualizations on Intimate Partner Violence (IPV) began with identifying relevant and reliable datasets spanning the past 35 years or so. We brainstormed different approaches to setting up the data to intrigue potential target users by sketching out various bar and line graphs (see Appendix). In turn, the data was cleaned and standardized using Excel, ensuring consistency in categories such as gender, year, and age groups. Before importing the data into Tableau, we focused on restructuring it into a clean format, which enabled more flexible filtering and chart creation.
Once the dataset was prepared, we explored multiple chart types in Tableau to determine the most effective visual representations of key trends. Time-series line charts displayed changes in IPV rates over time, while bar charts facilitated comparisons across demographic groups such as age and gender. Throughout the design process, we prioritized interactivity by utilizing filters, parameters, and tooltips to enable users to explore the data dynamically. We also applied color palettes intentionally to distinguish categories while maintaining accessibility and visual clarity.
We adhered to key user-centered design principles to ensure the dashboards were informative and user-friendly. This included iteratively refining the layout based on peer feedback and usability testing. We aimed to strike a balance between aesthetic simplicity and analytical depth, employing clear labels, concise titles, and contextual annotations to help users interpret the data without requiring external explanations. Ultimately, the Tableau dashboards sought to translate complex, longitudinal IPV data into accessible and engaging visual narratives that could support public awareness and policy discussions.
Highlights of the design process
The process of exploring this project involved understanding, identifying, gathering, and visualizing data that informs the assessment of IPV rates during crisis periods, such as economic downturns, pandemics, and natural disasters. This phase was crucial for interpreting a concept and developing it into a visually driven analysis.
Challenges
One of the main challenges was finding data that accurately represented our focus. Initially, our idea focused on gender-based violence; however, after researching datasets on this topic, we found that there was an overwhelming amount of information. There are many types of gender-based violence and numerous ways to present it; we had to refine our scope. As a result, we narrowed our topic to intimate partner violence and found greater success.
Despite our efforts, we encountered other challenges. One major obstacle was the temporary removal of key websites and datasets due to executive orders from the current administration. These resources were only recently reinstated following a federal judge’s ruling. Changes in administration often lead to shifts in agency funding and data collection priorities, which can impact the accuracy and availability of information, as we saw firsthand. While external agencies strive to report as accurately as possible, inconsistent reporting standards across regions and years pose significant challenges, particularly in creating multivariate visualizations where comparability is crucial. Additionally, given the sensitive nature of intimate partner violence (IPV), we were careful to ensure that our data selection respected both privacy concerns and ethical standards, also representing and cleaning the data to show it to the best of our ability.
Insights
Throughout the process, several key insights emerged. Most notably, our group identified a strong correlation between spikes in IPV rates and periods of societal crisis, such as the 2008 financial crash, Hurricane Katrina, and the COVID-19 lockdown. For instance, in 2005, when Hurricane Katrina struck, a noticeable increase in IPV was observed among younger women (ages 18–24). We hypothesize that this may be tied to housing insecurity, as the storm disproportionately impacted urban centers where younger populations typically reside. During the 2008 financial crash, we noted a significant rise in IPV among women aged 25–34. This group is more likely to be in committed or marital relationships, and we believe the spike may be linked to increased household stress and financial strain affecting intimate dynamics. Across the data, one of the most vulnerable groups consistently identified was separated women aged 25–34. This aligns with existing research indicating that separation is often the most dangerous period in abusive relationships, and the data reinforced this understanding. A key pattern that emerged was what Emily referred to as “the wave”—a marked decline in IPV rates from the 1990s to the early 2000s, followed by a gradual increase starting around 2005, a sharp spike in 2008 during the recession, and a renewed surge during the COVID-19 pandemic. This trend suggests that while IPV had once been on the decline, recent societal crises have reversed that progress.
Usability Assessments and Changes
Initially, we leaned toward using more complex charts, believing they would present the data more comprehensively. However, user feedback revealed that simpler visualizations were far more effective. Users preferred to examine each piece of information individually, taking time to absorb and reflect on what the data represented. Across the board, participants expressed that they left the testing sessions with a significantly more profound understanding of intimate partner violence.
User evaluation(s) with methods, results, and discussion
Our group conducted a user-centered review with a small group of users to evaluate the effectiveness and usability of Tableau dashboards for visualizing trends in Intimate Partner Violence (IPV) over the past 35 years.
Most users found the visualizations intuitive and informative, with particular praise for the interactive features such as filters by gender, region, and year. Several users noted that the line and bar charts helped them easily identify patterns over time. However, some participants suggested improvements, such as more transparent labeling of axes, adding definitions for specific terms, and including context or annotations to highlight policy or societal changes that may have influenced the data.
While the dashboards succeeded in making complex data more accessible, the feedback emphasized the need for thoughtful design choices to support non-expert users. Based on these insights, we revised the charts to enhance labeling, added tooltips with definitions, and included brief narrative summaries to facilitate interpretation and understanding. This user-centered approach ensured the final visualizations were accurate, engaging, and actionable for the audiences most invested in addressing IPV.
This visualization explores the disproportionate impact of Intimate Partner Violence (IPV) on women across different racial and ethnic groups. Using data segmented by race, the chart reveals significant disparities in IPV rates, highlighting that women of color—particularly Black and Native American women—experience IPV at markedly higher rates than their white counterparts. By visualizing this data, the chart draws attention to the systemic inequities and intersectional factors contributing to vulnerability, underscoring the need for targeted prevention strategies, support services, and policy reform. This work aims to make the data both accessible and actionable for stakeholders in public health, advocacy, and social services.
Evaluation of Final Visualization
Goals
The goal of our project was to use data visualization as a tool to better understand and communicate the complex issue of Intimate Partner Violence (IPV). We aimed to identify patterns and trends in IPV data across different populations, periods, and geographic locations, while also exploring how factors such as gender, income level, and access to resources influence these patterns. Through the visual representation of data, our objective was to make the information more accessible, impactful, and actionable for audiences ranging from educators to policymakers. Ultimately, our team aimed to integrate visualization with thoughtful design to inform, engage, and advocate for increased awareness and change.
Target Users
This project is designed to raise awareness of IPV occurrences during times of crisis, with several key user groups in mind. First, the visualizations aim to support professionals involved in shaping public response, such as public health officials, policymakers, and social service workers. These individuals rely on accurate, accessible data to make informed decisions, educate communities, and coordinate aid and protective policies during emergencies.
Another key audience includes those managing local organizations, shelters, and advocacy centers. For them, our visualizations highlight when and where needs are most acute, helping them allocate resources more effectively and respond to heightened risk. Researchers and academics can also utilize these tools to examine long-term trends, identify risk factors, and propose evidence-based interventions.
Beyond formal systems, we also aim to reach community leaders, educators, and the general public. By presenting the data in a clear and approachable manner, we strive to break the silence surrounding IPV, particularly when stigma, societal expectations, or cultural norms discourage open discussion. Making these patterns visible can help communities better understand, anticipate, and respond to IPV during times of crisis.
Task
Our analysis focused on visualizing how intimate partner violence (IPV) rates tend to fluctuate in response to major world events. We observed that during periods of societal stability, IPV rates often plateau or decline slightly. However, during times of disruption, such as economic recessions, political transitions, or global crises, there is a noticeable spike in IPV incidents. These patterns underscore the pressing need for increased awareness, open dialogue, and access to supportive resources during times of uncertainty. As a society, we must recognize how external stressors can escalate private violence and commit to proactive measures that protect and support vulnerable individuals.
Effectiveness of overview and detail views
The visualizations we presented provide a strong foundation for merging views, allowing users to grasp trends while accessing specific data points for deeper understanding. Within the broader context of the patterns revealed through our time series line graph and grouped bar chart, which effectively illustrate overarching changes in variables such as IPV rates over the years, and how mortality rates increase. At the same time, our visuals include detailed elements that delve into further specifics, including the relevance of percentages to age groups, as well as comparisons between these types and female and male victims, and the resources utilized by these victims.
Regarding data encoding, the visualization employs horizontal and vertical positioning, placement along a standard scale, bar height, and angles. These choices enable precise analysis, as they align with an acute observation of comparing values on the chart’s axis. The presentation’s choices enhance the accessibility and clarity of our data visualization. This encompasses axis labeling, color, and the layout of the data. Components such as text legibility, consistency in marks and time intervals, and colorblind-safe palettes improve understanding and ensure the target audience can interpret the data without confusion.
The effectiveness of our IPV visualization project can be evaluated through the lens of Shneiderman’s Information Visualization Mantra, “overview first, zoom and filter, then details on demand”, and Munzner’s taxonomy of low-level actions, including select, explore, and reconfigure. Our dashboard provided a clear overview of IPV trends over time and across demographic groups, helping users quickly understand the broader patterns and how they align with major societal events such as economic recessions or shifts in political leadership. To support zooming and filtering, we implemented interactive sliders and drop-down menus that allow users to narrow their view to specific periods or demographic subgroups, enabling targeted exploration. This is beneficial for policymakers or researchers focusing on high-risk populations. For details on demand, hover and click interactions revealed exact values and timestamps, allowing users to explore the data in depth without cluttering the primary view. These features also align with Munzner’s framework: users can select data elements through interactions, explore subsets of the data by zooming or filtering, and reconfigure views to focus on meaningful patterns. Overall, the system’s interaction techniques effectively supported user goals by providing layered access to information in an intuitive and ethically responsible way.
Conclusion
To continue building stronger support systems during times of crisis, we need to ask: How might we better assist victims of intimate partner violence when incidents surge? By looking closely at the gaps in our current responses, especially within local communities. We can begin to uncover where systems are falling short and where emergency plans and actions can be strengthened. This isn’t just about improving services on paper; it’s about creating real, compassionate support that works at both the community and legal levels. It’s also vital to listen to the voices of those who’ve lived through IPV, how they talk about their experiences, whether they knew help was available, and if they felt safe enough to reach out. Their stories can reveal truths that data alone can’t, helping us build support that truly meets people where they are.
Sources
1) Bureau of Justice Statistics. “National Incident-Based Reporting System (NIBRS).” U.S. Department of Justice, Office of Justice Programs, https://bjs.ojp.gov/national-incident-based-reporting-system-nibrs. Accessed 5/15/25.
2) Bureau of Justice Statistics. (n.d.). Intimate partner violence. U.S. Department of Justice, Office of Justice Programs. Retrieved May 16, 2025, from https://bjs.ojp.gov/taxonomy/term/intimate-partner-violence?page=0
3) Nonfatal Domestic Violence, 2003–2012
Bureau of Justice Statistics. Nonfatal Domestic Violence, 2003–2012. U.S. Department of Justice, Office of Justice Programs, 2014, https://bjs.ojp.gov/content/pub/pdf/ndv0312.pdf
4) Intimate Partner Violence, 1993–2010
Bureau of Justice Statistics. Intimate Partner Violence, 1993–2010. U.S. Department of Justice, Office of Justice Programs, 2013, https://bjs.ojp.gov/content/pub/pdf/ipv9310.pdf.
5) Criminal Victimization, 2022
Bureau of Justice Statistics. Criminal Victimization, 2022. U.S. Department of Justice, Office of Justice Programs, 2023, https://bjs.ojp.gov/library/publications/criminal-victimization-2022