Hispanics in the United States are 23% more likely than non-Hispanic whites to be obese and are at higher risk of developing hypertension and diabetes (CDC, 2015; Thomas & Allison, 2019). Additionally, according to the CDC, only 14.4% of Hispanics meet aerobic activity and muscle-strengthening guidelines (2019). Common barriers to physical activity cited by the Hispanic population include a lack of knowledge, motivation, and social support, as well as time and economic constraints. One way to help people meet the recommended aerobic and muscle-strengthening guidelines is the use of fitness applications (smartphone programs designed to help users achieve their fitness goals), which have been proven to reduce such barriers and improve general wellness.

The rapid increase in smartphone ownership around the globe, and among Hispanics in particular, makes fitness applications a great option for reducing sedentary lifestyles because these technologies can increase the delivery of health interventions on a massive scale and are more affordable. For example, while several fitness applications are free, gym memberships start at $9.99 and can cost over $330 a month (Sherman & SI Showcase + Pillar4, 2023). Furthermore, attending a gym often makes people nervous and can even be intimidating given lack of skill and fear of being judged by others who attend the gym on a more regular basis.

Despite the conceivable benefits of using fitness applications, engagement with fitness applications is often short-lived. Once the novelty wears off, users tend to stop using these applications. This means that fitness applications are not motivating users long-term. This trend is even more pronounced among Hispanics. While Hispanics are higher adopters of smartphones compared to the general United States population (98% vs 93%), they engage with apps including fitness applications at a much lower rate than the national average (36% vs 58%) (Arora et al., 2016). This represents a major barrier against ensuing health equity for Hispanic communities. In my research, my team and I are exploring why this trend is happening. In other words, why are Hispanics in the United States less likely to use fitness applications than the general population? And how can we build fitness applications to cater to the needs of Hispanic communities toward improved health outcomes?

As a first step, we conducted a content analysis of publicly available user profiles in a popular fitness application with the aid of computational methods of data collection. Here the goal was to assess trends in how the general population uses fitness applications and if the use of specific features of the technology motivates people to increase their physical activity. In this research, we found that people who engaged with other members of the community by following each other not only tracked more workouts in the application but also increased the amount of weight they lifted over time. Additionally, we found that users who followed customized programs recommended by the application based on criteria predetermined by the user (i.e., age, gender, goal) also increased the weight they lifted over time. This means that among the general population, regardless of their ethnicity, following customized workout programs and interacting with other members of the community motivates physical activity. However, it is important to note that most users of these fitness applications are not from minority communities. We wanted to know if these trends hold true for Hispanic adults as well.

So, this time we conducted a survey study to assess if the current state of fitness applications motivates Hispanic adults to engage in physical activity. We wanted to find out if the trends we found in our content analysis when looking at the general population also hold true for Hispanic adults. We found that features that allow users to interact with each other, such as sending messages, checking the activity of others, or following and being followed by other users, are a double-edged sword. While these features made Hispanic users feel a sense of relatedness and community, the same features made users feel a lack of control and competence. Additionally, unlike in our content analysis, in this survey we found that customization options centered on selecting workout activities filtered based on personal information such as gender, age, and goals, were not relevant to Hispanic users’ needs and motivations. In other words, current customized programs do not motivate Hispanic adults to engage in physical activity.

Two main conclusions can be drawn from these findings. First, current user-user interactivity features of fitness applications focus on what is known as “the quantifiable self.” The quantifiable self is the ability for users to self-track measurements and achievements, which can then be shared among members of the community. While the “quantifiable self” allows users to visualize their improvement and overall health outcomes, it is also conducive to competition and social comparative tendencies that may be detrimental to one’s health. Studies have indeed found that tracking technologies can lead to negative health outcomes such as bad eating behavior, especially among members of the population who are already predisposed to such disorders. It seems that while these interactivity features increase the sense of community of Hispanic adult users, these features also make them feel in competition with others, and thus lead to disengagement and attrition. Secondly, it was interesting to find that the customization of workout routines is highly motivational when assessing the general population, but not when assessing Hispanic users specifically. It is likely that this occurs because current customization options of fitness applications do not reflect the needs of Hispanic adults. To note, most fitness applications are built and tested with the majority population in mind. A one-sizefits-all approach may lead to the needs of specific communities not being reflected and met. Thus, customization available in current fitness applications may not align with values and needs that are important for Hispanic and other minority communities. For example, current fitness applications allow users to select workouts based on criteria such as gender and age, which may not be important criteria for certain members of the population. For example, for users who lack the skill and self-competence to start a fitness program, which is a known barrier among some Hispanic adults, a better customization option may be skill level and familiarity with equipment. Likewise, fitness applications typically recommend workout plans to be completed as an individual, but certain populations (for example, families with children) may prefer workout activities that can be completed as a family. To put it simply, what we realize through this survey is that for people to want to engage with fitness applications long term and follow recommended programs, these programs should be tailored for each individual user, considering their lifestyle, culture, needs, and motivations. Otherwise, fitness applications are only targeting some members of the community, but leaving those who may benefit the most from them out of the equation.

This survey research was very insightful. We learned that current fitness applications do not align with the needs of the Hispanic population, and thus do not help motivate them to increase their physical activity. However, what are Hispanics looking for in a fitness application? We are currently conducting interviews with Hispanic adults across the United States to find this out. While we are still in the process of conducting these interviews, a few trends have already emerged. First, to prepare for these interviews, we decided to create a video where we would walk participants through a fitness application currently on the market. Some participants were already users of fitness applications, and some were not. The goal of this video was to provide an overview of the most common features of fitness applications so that we could then ask participants what features they found useful and which ones they did not. Because we expected some participants to prefer watching the video and being interviewed in Spanish, we created two videos, one in English and one in Spanish. The process of creating the video in Spanish was insightful on its own. It turns out that some features available in the English version of the application were not available in the Spanish version! Even before conducting the interviews, and in the planning stages, we realized there were some severe equity issues in fitness applications. If your first language is not that of the majority, then you cannot benefit from all that the fitness application can offer.

Now let’s talk about the interviews themselves. It is still early in the process of data collection for us to discuss specific findings, but there are already a few emerging trends. First, our participants express liking features that allow them to track their fitness, sleeping habits, and general health. However, they prefer to keep those trends to themselves rather than sharing them with other members of the community. This is because they feel communities tend to be too competitive, which does not motivate them. One participant explained her company created a closed online group for employees to engage with each other through a fitness application, however, she decided to exit the group given its competitive nature. This finding is consistent with our survey research. The focus on competition creates a negative environment that disengages users and can lead them to quit fitness applications altogether. Participants prefer to simply track their workouts through their device or fitness application but not engage with others. When asked about closed communities (where users can create groups with friends and family of their choice), participants explained that their friends and families would not be interested in joining a community or that they already connected with friends and family about fitness through face-to-face communication, thus did not feel the need for such an online community.

Secondly, in terms of customization of workout programs and recommendations, participants each have a unique need that they would like fitness applications to tackle. One participant expressed that as a mother, she would like recommended activities to be activities that she could do with her son. Another participant stated that she would like to receive notifications regarding yoga, as this is a workout that she likes and practices regularly. Yet other participants stated that due to their disability, they would prefer fitness applications had more activities that they could actually participate in. Note that when participants were asked if they would like recommendations about their community and demographics, participants rarely mentioned their identity as Hispanic, but rather as parents, people with disabilities, and residents of geographical communities. When asked specifically if they would like recommendations based on their ethnicity, participants explained they did not necessarily see their identity as Hispanic as an important factor when considering their health and fitness. This was partially surprising, but as a Hispanic myself, it was not. Traditionally, health interventions targeted at Hispanics tend to focus on cultural aspects, often treating Hispanics as a monolith. The reality is that the Hispanic community is very diverse, as such some customization approaches often miss the mark. One of our participants stated that she often gets frustrated when systems customize her profile as a Hispanic because they often reflect stereotypical ideas of what it means to be Hispanic. Instead, she suggests customizing based on each individual’s needs. This suggestion is important because while it is imperative to build health and fitness applications that meet the needs of minority communities, such as Hispanic, it is also important to remember that a) Hispanics are not a monolith, diverse opinions and values exist among members of the community, and each individual perceives their own culture differently; and b) there is much more to the identities of Hispanics than their ethnicity. Intersectionality is an important concept to define and consider. Intersectionality refers to how social categorizations like race, ethnicity, and gender are all interconnected. We as human beings have several elements that make up our identity. Certain elements and categorizations may be salient for us in some contexts, but not in other contexts.

So, the difficulty that we are facing is the following: Current customization options of fitness applications typically focus on customizing workout activities and recommendations based on age, gender, and goals. These customization options do not meet the needs of Hispanic adults. Hispanics do not find them useful or motivational. However, customizing based on culture can be perceived as stereotypical and even rude if people’s relation and perception of their culture is not considered—barriers, needs, and values may be different among members of the population. And this is the problem that we are trying to solve with our research: How do we customize recommendations to motivate fitness among Hispanic adults while also considering each person’s relation with their cultural identity? One proposition is cultural customization at the individual level where, instead of developing a single intervention aimed at Hispanics (i.e., users who identify as Hispanic receive the same message because of their identity as Hispanic), we can customize based on each individuals’ relation to their identity. To do this type of cultural customization we need to understand 1) how each user perceives their own culture and the extent to which they identify with it; 2) specific cultural values that are and are not important to each user; and 3) other parts of their identity and lifestyle that may also be relevant to that individual (e.g., age, having children, living in a city vs. rural community, previous injuries and disabilities, type of physical activity and workouts preferred).

Cultural customization at an individual level is difficult to achieve, however, with technological advancement in machine learning, data-sensing technology, and artificial intelligence (AI), it could be feasible. In our work, we propose utilizing a chatbot that serves as a fitness coach to provide customized recommendations to users. We propose that these recommendations should align with users’ unique characteristics, as well as environmental and contextual situations. Leveraging data-sensing technology, we could utilize context-specific information such as location, weather, and calendar availability from each user to recommend activities, events, and other health-related recommendations. For example, given the location of the user, we could determine if an outside workout activity or inside activity is recommended given the weather. Additionally, the system could ask users questions during sign-up that are more in-depth than the traditional demographic questions. These questions could ask people about their lifestyle, disabilities, and preferences when it comes to physical activity. That way the system can recommend activities that are meaningful for each participant. A parent with a child could receive age-appropriate activities to engage with their children, but a single female who likes running could receive recommendations about running groups in the area. More importantly, the system could ask people about their values and motivations to further customize the user experience. Surveys asking the extent to which people identify with cultural values and worldviews already exist and can be leveraged to create a more comprehensive user profile. For example, a common misconception that exists is that because Hispanics/Latinos are a collectivist society—where social behavior is guided by the goals of the community compared to those of the individual—then community is always a primary focus among those who identify as Hispanic. Nonetheless, many variables influence the degree of collectivism of an individual, including acculturation and other personal experiences and personality traits. Furthermore, being collectivist does not mean the person never prioritizes his/ her personal goals and values. Indeed, the need for uniqueness is shared by all human beings. Furthermore, community also comes with certain drawbacks, for example, social comparison and stigma. Thus, in some contexts like physical activity people may prefer to remain private. All of this is to say that the degree to which a person identifies with a worldview such as collectivism may differ, with some people identifying more strongly with it than others. Thus, we propose a system where we could assess the extent to which a person identifies with a set of important cultural values and personalize the experience accordingly, combined with data-sensing technology. For example, in our interviews one of our participants said that nutrition advice was not relevant to her given that she does not cook frequently in her household, however, receiving recommendations for activities to remain active related to her age group and status as a mother was important to her. On the other hand, another participant said he is very interested in nutrition and healthy eating, so recommendations on this topic would be especially meaningful for him. He also stated that he would appreciate it if some food recommendations included Puerto Rican food, although not all the time. While both participants identify as Hispanic, they each have different lifestyles and their identity as Hispanic has different levels of salience in relation to their physical activity and health.

While this project is still in nascent stages, we plan on applying for funding through the National Science Foundation and the National Institutes of Health, with the goal of motivating physical activity among Hispanic adults through the use of context-specific and culturally tailored fitness applications. Ultimately, our efforts are focused on health equity and reducing the barriers to physical activity among Hispanic adults. Nonetheless, as a Hispanic researcher I have two additional important goals: 1) Through my research I aim to provide new guidelines for cultural tailoring. Oftentimes current health interventions focused on Hispanics miss the mark, and, while well-meaning, treat Hispanics as a monolith and assume that one single intervention targeted to Hispanics fits every person who identifies as Hispanic equally. Through my research I want to share the richness and diversity that exists among those of us with Hispanic and Latino heritage. The beauty of being Hispanic is that while we share many values and worldviews, we are also unique. 2) Through my research I also want to inspire others of Hispanic heritage to join the academic community. We have important perspectives to share and unique ideas to test, but there are only a handful of Hispanic researchers in the intersection of communication and human-computer interaction research. Recruiting a graduate student to assist in my project was a challenge on its own. It was important for me to recruit a Hispanic student for this project, and that this student was fluent in Spanish. I was unable to find an interested student within my college and had to expand my search. I eventually met my current graduate assistant who has been a tremendous help, but it took time to find a student with the necessary skills. As the Hispanic population keeps growing in the United States, it is important that our academic community also sees more presence of Hispanic researchers. Only then can our ideas, unique perspectives, and voices be heard, and only then can we strive to really benefit our community through our research.

---