Helping women-identifying individuals feel & be safer when walking alone at night
Ramble is the product of a project exploring how to help women-identifying individuals in cities feel and be safer when walking alone. Research indicates that women tend to be more afraid of walking alone than men and have gender-specific fears, which may be a both a product and perpetuation of gender inequality.
A literature review and interviews with subject matter experts emphasized the following points, which influenced our ultimate solution: that it is necessary to be mindful about the differences between perceived versus actual safety, namely that it is important to target perceived safety in order to influence walking behavior; the influence that the surrounding physical environment can have on perceived and actual safety; the importance of incorporating opportunities for education around safe practices; and the value of having an outlet to discuss incidents.
Diary studies, field research, and semi-structured interviews conducted with women-identifying University of Washington students resulted in a set of design principles that chiefly informed the feature set of Ramble, including: easier communication of fear, both during and beyond the walk; ability to impact surroundings; and risk assessment and response education.
MHCID Capstone Project, University of Washington
Product Manager, Prototyper, Interaction Designer
Sara Al Mughairy — Lead Researcher
Sarah O' Connor — Lead Prototyper, Designer
Julie Sutherland — Lead Designer
April 2016 - August 2016
Secondary research, primary research, ideation, prototyping, usability testing, interaction design, interface design
InVision, Sketch, Illustrator, Photoshop, Principle
Our response to our secondary research findings and design principles is Ramble: a mobile app that makes walking safety a social focus by leveraging community activation to support women-identifying individuals navigate the fear of walking alone. Ramble targets easier communication of fear both during the walk, by virtually pairing remote walkers who can easily communicate unease through simple button presses, as well as beyond the walk through a social media feed that normalizes conversation around walking experiences and incidents.
Ramble combines location data with qualitative report data to provide organizations with influence on pedestrian safety with information regarding areas in which users feel unsafe, as well as information regarding the effectiveness of existing interventions. Ramble incorporates “Safety Tips” throughout the app to provide users with research-grounded risk assessment and response strategies, as well as ways to be more prepared and methods for mitigating the risk of being targeted in the first place.
Data collected by Gallup indicates that over the past seven years, 36% of U.S. residents reported feeling unsafe walking alone at night. This percentage increases to roughly 50% amongst women, city dwellers, and the socioeconomically disadvantaged. Data further indicates that there is a significant 18-point gap in perceived safety between the sexes; 45% of the entire female population reports feeling unsafe walking alone at night, as opposed to 27% of men (Gallup 2015).
This gap in perceived safety between men and women can likely be attributed to the unique fears that women face:
- The “fear of crime paradox," which suggests that although women are less likely to be victims of street crime than men, they tend to be more afraid of it (Pryor et al, 2013).
- Women are particularly afraid of sexual violence and sexual harassment, leading them to develop preventative strategies of distancing themselves in both space and time from potential attackers (Bianco & Lawson, 1997).
- Women tend to be more aware of environmental cues and safety risks, which may contribute to them feeling less safe (Bianco & Lawson, 1997).
- Finally, the unique vulnerability and powerlessness that women face in society may largely contribute to their fears: women, both historically and currently, experience oppression, lack of democratic control, and marginalization in their communities, contributing to feelings of helplessness (Koskela, 2000).
We realized that there was an opportunity to design an intervention addressing the specific fears that women face when walking alone. Our intention was to help women- identifying individuals in cities feel and be safer when walking alone. Residents of cities are more likely to use walking as a regular form of transportation than those in more rural environments (Saelens, 2003), and as described in the Gallup findings above, city dwellers tend to have an increased fear of walking alone, which may correspond with increased risk (Gallup 2015), making them a fitting population to target.
Over the course of three months, we conducted both primary and secondary research in an effort to understand the safety strategies women-identifying individuals currently engage in, as well as identify opportunities to improve upon or extend these strategies.
- Conducting a literature review helped us identify four broad categories of subject matter experts to contact.
- Attending the CHI Conference (Conference on Human Factors in Computing Systems) in San Jose caused a lag in our schedule.
- Synthesis and analysis of research conducted with target users began immediately and continued through the beginning of June.
- Follow-up interviews took place with diary study participants, therefore semi-structured interviews were dependent upon diary completion. Participants completed their diaries at different paces, and interviews were scheduled accordingly.
- A competitive analysis helped us identify trends in existing products. The ensuing heuristic evaluation assessed a breadth of these products.
Our literature review exposed us to a set of common themes that contribute to women’s feelings of unease while walking alone:
- There are different types of environmental factors that may signal to women that there is heightened risk of walking in a certain area. These can be what researchers refer to as social incivilities, physical incivilities, and properties of the built environment(Loukaitou-Sideris, 2006).
- Features of the built environment can contribute to fear by reducing visibility of potential threats and/or inhibiting an individual’s ability to escape (Loukaitou-Sideris, 2006).
- Regardless of whether an actual threat exists, perception is what directly impacts one’s actions and motivation. For instance, if people perceive an area as unsafe, they tend to walk less (Hong & Chen, 2014).
Subject Matter Expert Interviews
In order to understand how the construction of urban and digital spaces affect women’s perceived and actual safety, we spoke with Human Centered Design and Engineering faculty member, Daniela Rosner, as well as Seattle Neighborhood Group Senior Program Manager, Tari Nelson-Zagar.
We also met with UW staff members tasked with assessing and preventing violence on campus. These individuals include Gillian Wickwire, the UW SafeCampus Violence Prevention Response Program Threat Assessment & Management Specialist, and Natalie Dolci, the UWPD Victim Advocate.
We then reached out to Gailyn Perrin, a Taekwondo instructor and self defense expert, to learn more about what individuals can do to protect themselves.
We gained the following key insights from conducting interviews with these individuals:
- Our design solution should leverage “Community Activation," which is a Crime Prevention Through Environmental Design (CPTED) term used to encourage community members to “work together to ensure each other’s safety.”
- Crime data should expose stigmatization, rather than contribute to it.
- Our solution must be simple, and cannot require much thought or action from a user in a threatening situation; in threatening situations, executive functioning failure is common.
- The most impactful solutions are ones that lower the threshold for reporting so that more people share their experiences.
- Nonphysical aspects of self defense - awareness, body language, boundaries, and projecting confidence - increase preparedness and confidence, which are key factors in lowering your chance of becoming a statistic. This confidence is related to how much exposure and knowledge you have.
- It is vital to trust your intuition. It’s common for people to disregard feelings of unease, but human beings have developed senses to protect themselves, and we should learn to pay attention to them. Subsequently, taking action on those feelings, not ignoring them, is a significant factor in keeping oneself safe.
Popular Media Scan
We conducted an in-depth popular media scan in order to understand how the local media covers the topic of sexual assault.
We noticed a trend indicating that whenever an assault occurs, there is a flood of corresponding articles both about the specific incident, but also about preventive safety measures women can take in order to protect themselves. This trend illustrates an education cycle that is highly reactive instead of preventative. We believe that the dissemination of educational material should not be reactive, but instead should be readily available and part of a broader, normalized conversation surrounding women’s safety.
In order to better understand the landscape of existing solutions, we conducted a survey of roughly thirty existing products, both directly and tangentially related to the issue of walking alone, and categorized them into seven distinct groups based on features. The categories are as follows:
- Apps specifically designed to encourage safe walking practices (i.e., direct competitors)
- Risk and incident prevention (for example, preventing active shooter scenarios)
- Sexual harassment and assault prevention
- Apps that enable users to share their location, or monitor the location of another individual
- Apps that include crowd-sourced elements
- Physical devices
- Apps that utilize crime data
There are a number of ways in which the existing landscape of products were not sufficiently addressing the needs of our target users, but the most obvious failure across a majority of these products is that they underestimate the hesitation that women-identifying individuals may face in reaching out to others for help.
We recruited and conducted primary research with nine women-identifying University of Washington students. We chose to focus our research specifically on college students because they are likely to walk routinely, and research has shown that female college students may be extremely perceptive and sensitive of the area around them, particularly at night, resulting in increased fear (Pryor et al. 2013). From a data perspective, another benefit of studying college students is that there is robust literature on perceived fear and walking amongst this population, which allowed us to better triangulate our findings.
We began primary research by conducting a diary study in which we asked five participants to record five walks over the course of a week using a mobile fitness app, Strava. The app captured participants’ walking routes, which allowed us to observe any strange walking patterns (for example, taking inefficient routes to avoid particular areas). Participants also used the app to take and save photos during the course of a walk.
We also gave participants the option to record their experiences in a physical pocket journal that we crafted and provided them with.
We then conducted semi-structured interviews with our diary study participants. We began by discussing their recorded walks, highlighting aspects that made them feel safe or unsafe, and then moved into a few mapping activities and semi-structured questions to elicit more insight into their personal experiences and strategies.
This methodology incorporated a structured set of questions, with the flexibility to ask clarifying or follow-up questions. While the diary studies helped us identify the ways in which women negotiate space, interviews gave us the opportunity to ask specific questions regarding data collected in the diary studies. In addition to providing an opportunity to clarify diary entries, the interviews allowed us to better understand the nature and experiences that contribute to participant’s fears as well as how they communicate with others about them.
After conducting extensive research, we analyzing and synthesized our findings into three key insights, which in turn informed a set of design principles that guided subsequent ideation and prototyping.
Our participants feel safer around and seek out the presence of other people. For example, they often called or texted friends or family when walking alone at night. However, our participants expressed discomfort in reaching out to friends and family because they feared this behavior was burdensome and would cause their loved ones to worry. This hesitation to communicate also extends beyond the context of the walk.
This finding made us realize that there was an opportunity to facilitate easier communication of fear, both during and beyond the walk.
Features of the built environment impact perceived and actual safety
We decided that our design solution should enable women to impact their surroundings by making their concerns known to organizations that are capable of improving pedestrian safety.
There are difficulties assessing potential threats, as well as fear of unnecessary escalation. Research participants expressed concern regarding over-reacting to situations where they felt unsafe, not only for fear of expediting potential conflict, but also for social reasons; reacting in a defensive manner when it proves unnecessary can be an embarrassing experience, as well as potentially offensive for the person who is perceived as the threat.
Our design solution should help women feel confident in their ability to assess risks.
Our goal during the first stage of ideation was to generate as many unique ideas as possible. Over the course of several ideation sessions, we produced nearly 200 ideas. To guide us through this process we used several different ideation methods in order to approach our problem space from different angles.
Ideation methods included the following:
- Individual Ideation, in which team members referred to experience maps and stakeholder maps for inspiration.
- Concept Generating Matrix, in which we sorted our ideas into pre-walk activities and needs, activities and needs during a walk, and post-walk activities and needs.
- Principle to Opportunities Matrix, in which we based our brainstorming on our design principles, focusing on individual opportunities, system opportunities, and strategy opportunities.
- Concept Sorting, or, the begin of the concept refinement process. This consisted of grouping our ideas together into main directional themes: like “community conversation” or “aiding risk assessment.”
- Concept Synthesis, which consisted of mixing and matching the strongest components of our ideas to produce a smaller set of even stronger ideas.
We conducted three concept refinement sessions, in which we selected the most promising concepts. The first round of refinement resulted in seven concepts, while the second round of refinement resulted in three concepts.
- Virtual Reality Defense Training
A VR game that gives women practice and confidence about their ability to respond in the event of an assault. The concept aimed to educate and empower women in a fun and personalized way.
- Business-Pedestrian Partnership and Environmental Reporting
A mobile application that combines data about foot traffic, open businesses, the built environment, and potential threats in the area, to suggest a safe route. Additionally, it would provide individuals with a system to report unsafe features of the built environment to the entities capable of fixing them.
- Neighborhood Watch
In an effort to leverage “Community Activation,” this concept sought to connect walkers with local residents through a mobile application and exterior beacon system.
Ramble is a mobile application that makes walking safety a social focus. Our application bridges the real, in-the-moment need for women to feel “watched out for” with the broader scale social support needed for validation, education, and advocacy.
The design concept borrows the strongest aspects from many of our prior concepts and combines them into a solution that uses the experience of walking alone as a jumping off point to collect data, advocate for changes in the physical environment, promote safety education, and facilitate conversations regarding women's safety.
Once we converged on our final concept, we put various prototypes in front of users to both validate that we were headed in the right direction overall and to seek input on the details. Specifically, we had participants evaluate the major components of our concept, including the walking experience, the social channel, and the integration of educational safety tips.
We conducted a total of four concept prototype evaluation sessions in which we presented female UW students with a paper prototype and asked them a series of questions intended to discover who they wanted their walking partner to be, their expectations of the social media channel, and their feelings towards tips and locally-sourced incident report data.
We adopted a semi-structured interview format because although we had prepared a rigorous set of questions, this format gave me the ability to adapt to participant feedback in real-time and probe further on certain topics, or to change the order of questions as appropriate. Concept evaluation was guided by three overarching research questions:
- What is the desired level of interaction during the walking experience?
- What form and culture should the social media channel take?
- What type of education are users most receptive to?
Testing our prototype resulted in four key insights, which in turn informed a set of design implications that guided subsequent prototyping.
During primary research, we found that participants had a desire to feel the presence of another person during their walk, by for example, calling friends or family during the walk. However, they felt concerned about the burden this placed on others. Based on these findings, we created the Anonymous Walking Partner feature. Our prototype validated our hypothesis that this feature alleviates the concern of burdening or inconveniencing others.
The anonymous walking partner is an appropriate solution to this problem. Given that having a walking partner is comforting, designs should highlight that the user’s partner is a real person.
All four test participants found the presence of partner during a walk to be comforting, but wanted interactions during and after the walk to be kept to a minimum. Three out of four participants stated that they did not want the option to directly chat with their partner, and some participants went so far as to state that a 1-on-1 chat feature would deter them from using the app entirely.
Partners should only be able to interact via button presses during the course of a walk. Furthermore, the interaction should be made to be as convenient and efficient as possible.
As far as escalation is concerned, participants preferred the ability to customize their preferences; for example, choosing whether it is an emergency contact who is notified or the police, or, deciding if a certain number of contact attempts should be made before involving these other entities.
Provide two levels escalation (contact partner or dial 911) and provide users with the ability to select emergency contacts.
Participants do not necessarily want conversational “social” content, but they also expressed a desire to avoid a stream of horrendous events.
We decided to shape the content of the news feed so that it features a balance of user generated content and educational information. The social media feed should be empowering, not fear-inducing.
We iterated upon our prototype by incorporating the insights we gleaned from concept evaluation into an interactive prototype created using InVision. We tested the application’s usability and desirability by asking participants to complete including searching for safety tips, replying to a report, and engaging in a think-aloud exploration of the walking experience. Our interactive InVision prototype can be found below:
Interaction Design Prototypes
Although I was ultimately not responsible for the final visual design of our app, I was tasked with designing the flow of the final app, which involved creating interactions and animations using Principle.
I began by sketching possible interface designs and their interactions, followed by creating static high-fidelity Sketch mockups, which I then imported into Principle. My work helped guide my teammates' final visual design.
I advocated adopting a card based system because cards are flexible to content and because cards can contain blocks of content, metadata (in this case, the time content was posted and the location in which the content was posted), buttons, and so on. Cards also signal to the user that this is an entry point to more detailed information. Above all, I felt as if a card layout was appropriate for a collection of various types of data, and the social media feed portion of our application consists of four different types of posts: “Heads Up,” Safety Tips, Polls, and Questions of the Day.
I felt as if each card type should have its own distinct visual treatment so that users could easily distinguish between them. In my prototype, this distinction is conveyed through color and iconography.
Cards feature tags so that the can be easily filtered, as well as an action block. Actions include saving a card, sharing a card, or leaving a comment regarding a card’s content. Although sites like Twitter allow users to save content by pressing a “heart” button, I felt as if this action might not always be appropriate; I felt as if users needed a more nuanced way to interact with posts because it is likely that posts regarding traumatic walking experiences are not likable or worthy of “love.” For this reason, my prototype allows users to save a post by “starring” it.
Cards are rearrangeable, which is another reason why I felt as if a card based system would be most appropriate. I prototyped a searching/filtering system in which users can enter unique search terms, or, they can search by suggested tags. Cards are rearranged based on the user’s searching/filtering actions.
In an effort to integrate Material Motion principles into my prototype, the drop down menu is responsive and natural. That is, it quickly appears in response to a user’s action (tapping the search bar) and bounces as it settles to demonstrate the kinetic energy of motion.
We wanted to give users the option to either display or hide their handle and profile picture when posting“Heads Up.” Users have the option to add photos to their posts, as well as add a location for where the particular incident they are describing took place.
As demonstrated by the animation below, users can display or hide their personal information by simply pressing the “On/Off” toggle. The “on” and “off” states are represented by color (“off” in grayscale, “on” in yellow). Once a user has begun typing a message, the “Post” action becomes active, as represented by the change in color from grayscale to blue.
Our final solution is Ramble, a mobile app that makes walking safety a social focus, by leveraging community activation to support women-identifying individuals navigate the fear of walking alone.
We chose to create a mobile application for a variety of reasons:
- First, we wanted to create a product that would augment women’s existing personal safety strategies, and our research revealed that our participants tended to carry their phones in their hands, or, store them in an easily accessible location.
- Second, we wanted to create something that would be able to be utilized during all stages of the walk, which includes the walk itself, thereby necessitating a portable solution.
- Finally, we wanted our solution to be able to access a walker’s GPS coordinates.
User Flow Diagram
Our user flow diagram demonstrates the various possible paths through the Ramble application. The three main components of the application - News (i.e., the social media channel), Walking Partners, and Safety Tips - are highlighted in yellow.
Ramble connects single walkers together and tasks them with virtually monitoring each other’s progress towards their respective final destinations. In our primary research, we found that participants had a desire to feel the presence of another person during their walk. However, they were reluctant to reach out, citing reasons like wanting to avoid causing friends and family to worry, feeling embarrassed about being afraid, and feeling insecure about whether their fears were justified. By connecting strangers who are both walking alone, Ramble provides walkers with the presence of others while simultaneously eliminating the concern of burdening loved ones.
When a user is ready to walk, they can navigate to the walking screen, enter their origin and destination, and then tap “Find Walking Partner” to continue. Origin and destination information are required in order to determine the length of an individual’s walk so they can be paired with another walker whose journey home is of a similar duration. Additionally, Ramble calculates progress towards walkers’ respective destinations to provide real-time feedback to partners.
As a user waits to be paired with another walker, the UI displays a loading screen that features a safety tip. Presenting a user with a safety tip before they embark on a walk is an effort to increase a user’s awareness of their surroundings. Participants who tested the Ramble prototype preferred having a passive safety tip before they began their walk over having the partner sending a tip or not having a tip at all.
Walkers can monitor each other’s progress by viewing the progress indicator. It is important to note that although walkers can view one another’s progress, they do not know their partner’s geographic location and the progress indicator is a simple circular loading graphic (meaning it doesn’t mirror the actual walking route the user is taking).
Although walkers can view each other’s progress, we believe it is important for walkers to avoid attending to their device screen because attending to a device screen prevents individuals from being mindful of their surroundings, which may increase the susceptibility of becoming a crime victim (USA Today, 2012). We intentionally designed these screens to contain as little information as possible in an effort to dissuade users from staring at their cell phone. A walker can indicate that she has arrived at her end destination by clicking the “I Made It!’ button.
Despite participants feeling safer when they felt less alone, we found a general silence around communication of fear to others. Ramble attempts to lower the threshold for women to express their fears around walking by allowing users to signal their discomfort to their partner through discreet haptic feedback by simply pressing the volume button. Partners can provide each other with a sense of comfort by reciprocating this button press. This simple interaction allows our users to acknowledge their discomfort and express it without the concern of overreacting.
A rapid multi-press initiates a 911 voice call and shares your GPS location with the dispatcher. Walkers can also customize other escalation preferences.
If the user signaled that they felt uneasy during the walk, Ramble will prompt them to share more details about the source of said unease. Users are tasked with selecting relevant tags from a collection, giving users language about describing the environment they were uneasy in. Tags are “intelligently” populated, that is, selecting a particular tag causes other related tags to appear. For example, selecting a tag labeled “darkness” would cause a tag labeled “broken streetlights” to appear.
Social Media Channel
In addition to sharing aggregated user reports, Ramble also encourages users to share their walking experiences with other individuals in the nearby vicinity via a social media channel. This social media channel helps normalize discussions around the threats that women face while walking alone in addition to providing walkers with information that could make them aware of potentially dangerous situations in the area.
The Social Media Channel consists of both user-generated content and system generated content. and user are able to upvote, bookmark, share, and leave comments on this content:
- "Heads Up" posts are user-generated content alerting other users of nearby concerns. This format encourages users to only share information that can be seen as actionable, which in turn empowers other users to make informed walking choices.
- Safety Tips are system-generated posts incorporated into the News Feed in order to increase users’ confidence and knowledge in risk assessment and response. These tips are often related to recent local report data or recent user experiences.
- Polls are used to aggregate more information about user safety and personal experiences, in addition to surfacing common themes or situations occurring in a particular area. Polls also serve to promote further user engagement inside the app.
- “Question of the Day” posts are daily questions that appear on the News Feed in order to promote and normalize conversation about personal safety and walking experiences.
Our primary research indicated that participants were not confident in their ability to effectively assess or respond to risks. We wanted to teach our users how to better assess risk through Safety Tips, which are brief defense strategies grounded in research. These tips promote awareness of one’s surroundings, the importance of paying attention to intuition, how to carry oneself in order to reduce the chances of being targeted, and ways to respond to assailants. These tips are written so that they are memorable and simple enough to understand and execute.
Tips can be found in curated collections, which group information based on common themes. For example, the “Just For You” collection contains tips that are recommended to a user based on tips she has previously bookmarked. A list of these collections is visible on the Tip Categories page.
Selecting a category from the Tip Categories page directs the user to a tip collection, which contains a list of related tips. The page provides a brief description of the context in which these tips are useful.
The tip detail page explains how and why a particular tip works by providing context for the tip, as well as credible documentation.