Recruiting older adults for research studies is a challenging endeavor. We conducted an interview to understand older adults' preferences and expectations, with the goal of building a recommender system to support the selection of suitable research studies. Our findings suggest that sharing the results of the studies they participated in would motivate older adults to participate in more studies and give them a feeling of self-accomplishment and belonging. We list 15 design implications based on our user research and present a prototype system based on these design implications.
The main goal of our work is to design a recommender
system that helps older adults in evaluating and
selecting research studies to participate in. Designing an
accessible recommender system for older adults is not a
straightforward task. A number of general system
design suggestions are provided in review papers by Morris
[
Older adults constitute a large part of the world population—
a proportion projected to rise from 9% in 2019 to 16% in
2050 [
CEUR
when designing our recommender system.
However, to develop a recommender system for
participating in research studies, it is important to know what
factors influence older adults’ preferences in participating In October 2022 we conducted an IRB-approved study,
in such studies. Beyond the work by Jacelon [
Older adults are likely to participate in designing and in-depth follow-up questions whenever we desired more
developing smart technology systems if they find that it details or further explanation. The interviews lasted
will improve their lifestyles [
In designing a recommender system for older adults, it is
important to keep in mind that many of them did not have
computers in their early childhood, which impacts how
they learn to use a new system [
Note that while many studies on digital literacy
focus barriers that hinder older adults in learning a new
technology [
Upon discussing our open codes, we divided our findings into two broad categories: characteristics of research studies that attract older adults, and personal gains and expectations older adults have regarding research studies. The design implications related to our findings are displayed in Table 2. The quotes below are edited for clarity and brevity. 4.1. Characteristics of Research Studies that Attract Older Adults
[Basis for DI11 and DI14] Most participants expressed that if the topic of a research study is interesting, they will most likely participate. Of particular interest were research studies related to health and lifestyles improvements: “Once, they had a study where they put red buttons on the walls, and they tried to understand if we found that useful for emergencies or calling someone. We do not carry phones all the time. In a space like this [small patio] where most of the staf can not see us, a button is helpful to call someone quickly in case of an emergency.” —P7
Older adults also expressed a desire for challenging activities, such as pairing up with other people to create a prototype, or doing some other activity that is not beyond their limitations: “I have participated with some of the things they got over there, which were very interesting. One was for five sessions, where I paired up with two students, and the object was for the students to design something those folks could really use in their life.”—P3 “When talk to some of the students here for some of their projects, I am impressed with them because they come up with some pretty interesting challenges.”—P3 “We have a meeting once a week... It is called Greet... It’s held out in the entrance area of our my apartment. That group then gets together, and they hear each other... And so you get to see what other people, what they’re having problems with.”—P3
Beyond these arranged events, they do not get many opportunities to discuss diferent topics with other people in the facility. Research studies provide residents with good opportunities to talk about things they usually do not feel comfortable about or do not know with whom to talk about. In that case, they want the groups to be as mixed as possible because they think such groups will consist of more diversified opinions: “I get to hear all the opinions... sometimes they don’t agree with me but that’s more enjoyable to me. Uh, because I like to hear other people’s opinions on this study as to where they think we’re doing it right or wrong. Sometimes they have really good points.”—P3 4.1.3. Study Location [DI2] Participants like to go outside for studies, but most 4.1.2. Group Discussion have health issues that prohibit them from driving. They expressed that if studies provide them with transporta[DI5, DI13] All participants liked studies where they are tion and accommodations, they are happy to go to a asked to talk in a group rather than in a one-to-one dis- remote location: cussion, because they want to socialize with the people around them: “It depends on the circumstances, how you get there. Yeah, if you could go and stay, and perhaps you know... Would they give you housing and food, or would you have to arrange it yourself?”—P6 4.1.4. Computer Literacy [DI1, DI15] Our participants had moderate knowledge of how to use computers and mobile phones. Most of them use computers to communicate with others via email. Some also use social media applications to keep in contact with friends and family, and banking applications to manage finances: “Say uh the email situation. I do all my banking on the computer and all that. And then, whenever I have a question about something, I go to Google. So yeah, I do (use computers)”—P3
None of the participants ever used a computer or mobile phone to find a research study to participate in, but all said they were willing to try out an application that can suggest studies based on their preferences. Furthermore, most of them had used a computer as an instrument of a study, or to participate in studies virtually: “Sure, we’ve [used computers], but we were doing it as a six of us together on their computers...”—P4 4.1.5. Relation to Past Experiences, Professions,
and Skills [DI3, DI4, DI6] Past experiences with studies play an essential role in deciding whether to participate in future studies. If a study did not go well for them (e.g., the study was in an inconvenient location, or some study procedure made them feel sick), they are likely to avoid similar studies. One other participant recalled taking a chemistry class where students were experimenting with dogs: “They used to do experiments, so they could get involved. These mad students, they were using dogs, and that’s when I decided I didn’t want to experiment on dogs.”—P6
Participants like to use their past experiences and skills.
One participant once had a near-death experience and overcame his fear by teaching others how to do CPR. He used the same attitude when deciding to participate in studies: “I’ve had an up close and personal connection to the whole process. Either I can do this or be the test dummy, and I’ve already been the test dummy. In this instance it has a tinge of overcoming fear. Okay this happened to me so I am going to prove I am over this by teaching someone else to do it... I generally end up volunteering for the same things: I want to use the skills I have acquired over a lifetime.”—P2 4.1.6. Social Influence [DI7, DI12] While not all participants initially enjoyed participating in studies, they often get motivated by friends living in the same facility who like to participate. Some noted that if they do not participate, others might think badly of them. One participant said that he participated in studies because it made others happy: “I do not go out of my way to volunteer, especially for research studies. I do it because people around me are doing something like meals on wheels and it makes them happy.”—P2 4.1.7. Invitation from Known Faces in the Facility [DI7] In the facility where we conducted our interviews, the DCO usually contacts residents to determine whether they want to participate in the research studies: “She comes over and tells us that we’ve got a group that’s coming, and would we like to participate?”—P4 “The young lady came in earlier this week and told me students want to come over to interview you. Would you participate?
And she knew I would say sure.—P2
Usually, residents instantly accept the DCO’s invitation without knowing what the study is about. This can cause issues because participants may find out later that the study requires them to do something they dislike.
While the DCO usually tells them who is conducting the study, this information is not suficient. Instead, it would be a good practice to let participants know what the study is about, how long it takes, and whether it has any further requirements. 4.1.8. Preferences for virtual meeting platforms [DI15] Most participants had health issues restricting them from participating in studies. When they started participate in studies virtually, they began to like online meeting platforms such as Zoom. Since then, they started using them for diferent purposes, such as ”Sunday Church School.” “We zoom church and zoom ”Sunday School” now; zoom helps us to communicate without being there physically. Okay, so we do. You know, they still do that through [zoom].”—P1
Most participants did not have any privacy issues regarding their participation in research studies. They do not want their names to be on the research articles, but beyond that, they cannot think of any other ways their privacy could be violated. While linkage attacks using the information provided in the research articles could be possible [15], most studies seem to ask only general questions about the issues they face as an older adult, which most older adults seemed completely fine with: “You don’t have to use our names. Anyway, this is just a general question about senior citizens.”—P5 4.2. Gains, Expectations of Older Adults from the Research Studies 4.2.1. Self-Accomplishment [DI8, DI14] Older adults know there are not many opportunities where they can contribute through physical work, so they seek out opportunities where they can help others through their experience: “I feel inclined for example to help this kid... he wanted advice on something for his job. I like helping individuals who need help for specific things.”—P2
Similarly, our participants mentioned that they want to participate in studies that help others and contribute to society. This gives them a feeling of self-accomplishment: 4.2.3. Research follow-up [DI9, DI10] Participants mentioned they would be happy to be contacted again if researchers had remaining questions; this would make them feel their experiences are an asset to the researchers. Furthermore, they would love to know more about the outcomes of the studies they participate in. This helps the residents see that they have made a small but valuable contribution to the research community and help them achieve something: “I would like to see what some of those results mean for you. I never did hear a follow-up with results, and what I wanted to know how it all came out.”—P4
One participant asked researchers to bring in the published journal article so they could read the study’s outcomes: “We would be happy to have that, since we were a part of this. Would be nice to put in the library here: Things from the participants of [facility name].”—P2
One of the participants mentioned that he wanted more autonomy to choose what study to participate in. He mentioned that he would love to use a system that would recommend opportunities to volunteer for a study based on his experiences, skills, and preferences: “Let me make up my own mind. What I like is when I have a diversity of options, like when I am on eBay. I can go and see the details of each option too.”—P2 5. Design of the Prototype “We look at projects that are probably good to have here, that will make the place better”—P3 “I like to see they look healthy and that we are all together, but we’ll tell them tonight about our adventure with you folks today, so we discuss that.”—P4
The design implications of our research are summarized in Table 2. Our user research findings show that older adults love to contribute to society by participating in 4.2.2. Discussion with Family and Friends research studies. Importantly, there are various individ[DI12] Participants said that on the weekends they spend ual factors that can be used to match participants with a significant amount of time with family. They discuss studies: some like to participate in group studies, othwhat their week was like and share things that they find ers prefer challenging activities, and still others prefer interesting. They mentioned that they often share their studies related to health issues. Older adults also face experience participating in research studies with their a variety of limitations that prevent them from particfamily members over dinner. To them, it is an accom- ipating in certain studies. Ideally, older adults can use plishment they can share with their family: these factors to decide for themselves which studies they want to participate in. In the current process, researchers usually do not contact the participants personally, and participants have no opportunity to select studies that match their desires and limitations. Our primary focus is thus on automating recruiting participants through a rec- of a study if results are available (DI9), contact the reommender system that can help the older adults evaluate searcher, or give feedback on the study (DI10). The feedand select research studies. We note that older adults back mechanism asks 10 questions (Table 4) about the sometimes need help to express their desires and limi- study, avoids studies similar to the ones they disliked, tations. We solved this issue by introducing a feedback and promotes studies similar to the ones they liked. option, which helps our system refine future recommen- ROAFSS collects and stores personal information about dations. study participants, which may cause privacy concerns.
Furthermore, our user research findings show that Note, though, that participants in our studies did not older adults love to learn more about the results of the seem overly concerned about this. Moreover, ROAFSS studies they participated in. Currently, this generally can potentially serve as a portal for research study particdoes not happen. Our system helps “close the loop” from ipation, so that individual studies themselves do not have participating in research studies to learning about their to collect any personally identifiable information. This ifndings and contributions by making the research out- would increase the overall privacy of the study particicomes available to users. pants, since they only have to disclose their identifiable
Our Recommender System for Older Adults to Find information once, and this information is never stored
Suitable Studies (ROAFSS) consists of four main pages. alongside their research participation data.
The recommendation page (Fig. 1a) lists studies that the
user may participate in, tailored to the answers they
gave to six questions about their abilities and prefer- 6. Discussion
ences (Table 3) and the feedback they gave to past
studies. Each study is listed with a title, contribution (DI8), Older adults’ perspectives matter and they love
individual vs. group participation (DI5), remote vs. lo- to contribute to the society While existing works
cal requirements (DI15), and the names of friends who [
We conducted in-depth interviews with older adults to understand their motivations, restrictions and expectations around participating in research studies. We found that older adults participate in research studies because like to contribute to society with their work and opinions.
We also built a prototype that will recommend the users
a small number of older adults from a single residential community. Diferent communities and people from di
of the system research studies that they can choose from. Such a recommender system can provide the older adult community useful support in selecting research studies that fit their preferences and limitations.
We would like to thank Ruthie Millar at the Clemson Downs retirement community for her help coordinating our interviews, and all of the interviewed participants for their willingness to participate in our study. This research was supported in part by the NSF award IIS 2045153.