How can we motivate the elderly to exercise

Summary

Despite the fact that the elderly are going to constitute around 1/5th of the total U.S. population by the year 2040 (U.S. Census Bureau) numbers, not many technology solutions are built for them. My team decided to explore their problems.

To understand their needs and way of life, we conducted observations and interviews with the elderly residents and management team of Campbell Stone Assisted Living Community in Atlanta. We found out that staying fit is one of their crucial needs. However, there was also a reluctance to exercise and a wish for fun and enjoyment. Hence, we explored ways of motivating the elderly to exercise by making it more fun and less boring.

One solution was an exercise game concept — an arm gesture controlled kayaking game to be played on Kinect. Due to limited time, we couldn’t build a fully functional prototype nor could we test our earlier system goal of motivation. We found a middle ground by building a semi-functional prototype and evaluating sub system goals (reception level, enjoyment and effectiveness of the exercise) through Wizard-of-Oz test methodology. The exercise game concept was well received, we did notice excitement and there was indeed a potential for it to become an alternate mode of exercise but we had a long way to go.

Personally, I learned that building trust with users is crucial but it is difficult and it takes time; experienced first hand the difference between what the users say versus do; learned that there could be indirect indicators of system success or failure (all one needs to do is watch!).

Kinect Kayaking Prototype

Outline

  1. Acknowledgements
  2. Key responsibilities
  3. Choice of the problem space
  4. Understanding the problem space: Research questions and methods
  5. Insights from research
  6. Framing the problem statement and design constraints
  7. Brainstorming and narrowing down solutions
  8. Prototyping (using Unity)
  9. Concept testing: Wizard of Oz testing
  10. Insights from concept testing
  11. Reflection on challenges and learning

1. Acknowledgements

Campbell Stone Assisted Living for letting us conduct field studies at their site

Professor Dr. Bruce Walker, Georgia Tech for guidance

My amazing teammates — Liza George, Lisa Li and John Thompson


2. Key responsibilities

Contributing to the research protocol, finding users and conducting interviews & observations

Contributing to data analysis, brainstorming solutions and building a prototype in Unity

Contributing to the system evaluation and analysis.


3. Choice of the problem space

According to the U.S. Census Bureau 2012, the elderly (people above 65 years) are going to constitute around 22 % of the total U.S. population. Despite these numbers, very few technology innovations are built specifically for them. They have distinct needs due to the differences in the cognitive and physical abilities. This made the problem space important.


4. Understanding the problem space: Research questions and methods

Field Study Site: Campbell Stone Assisted Living Community

We decided to work with an assisted living community to get access to the elderly. Our intention was to develop a better understanding of the problem space through the following high level questions.

What kind of residents live in the community? What are their physical and cognitive abilities?

How do the elderly residents spend their day in the community?

What are the concerns, needs, wishes and challenges of the elderly?

What do they enjoy doing the most? What do they dislike?

What type of social interactions do they have? With whom? How?

What experience do the elderly have with technology?

What facilities are provided by the community to its residents?

What are the challenges of the management team?

Because it was a mix of attitudinal and behavioural questions, our research approach was -

Observations on different days at different times by different researchers.

Semi-structured interviews with the elderly and the management team to develop a holistic understanding.

The main challenge at this stage was building trust with the assisted living community and interviewing the elderly. I personally faced issues due to cultural gap.

I have explained how we overcame these challenges in the reflection section.


5. Insights from research

All residents were 65+ in age. Majority had limitations on either physical or cognitive ability (e.g. use of wheelchair or dementia).

The assisted living centre is equipped with computers, LCD displays, and Kinect. Elderly used a basic phone or a smart phone for communicating with their family.

The management team rated the physical, mental and social well being of utmost importance. To support those goals, the team organises social and physical activities throughout the day. The main challenge for them was to draw people towards physical activity classes.

We observed that social activities such as Bingo had more presence than the physical activity classes such as Zumba. Bingo was popular especially because of rewards.

The residents wished to have fun in their post-retirement life. Though they realised the importance of exercising there was resistance to it. Reasons being exercise was either being boring or unsafe.

“I don’t like going for outdoor walks. It is just not safe. I might fall.” — an elderly woman resident

6. Framing the problem statement and design constraints

Physical, mental and social well being were identified to be crucial needs.

We chose to not focus on the mental well being because of lack of domain expertise. There were multiple opportunities for social interactions considering the group activities and digital communication. However, there was a reluctance to exercise because of it being boring and sometimes unsafe.

How could we motivate the elderly to exercise?
How can exercise be made more enjoyable and safe? How could we accommodate the limited physical abilities? How could we better use the existing technology? How can we make it easy to learn and remember?

7. Brainstorming and narrowing down solutions

We brainstormed multiple ideas, of which the key concepts were

Concept 1 — Digital bingo with exercises instead of numbers

Concept 2 — Gestured controlled video game (e.g. kayaking, walking)

Concept 3 — A pedometer that uses emoticons

Special acknowledgement: Lisa Li for wonderful illustrations
We chose to go ahead with arm gestures controlled kayaking game for Kinect as it satisfied most of our design constraints.

8. Prototype

Sketches as low-fi prototype versus Kinect interfaced game as hi-fi prototype
The main challenge in this stage was determining the fidelity of the prototype. Just sketches (lo-fi prototype) won’t elicit enough feedback and building a Kinect game (hi-fi prototype) was time consuming. Our middle ground was to build a digital game which would be controlled using a keyboard instead of Kinect (mid-fi prototype) and later evaluate it using Wizard of Oz testing method.
Mid-fi prototype

Interaction design details

Motivation — Introducing a story of kayaking on a magical river where gold coins are just scattered everywhere and they can be collected by passing through it. Also, reaching the finish line faster gives more coins. 
Easy learning curve — Interactions based on natural movements required for kayaking left, right and straight.
Feedback — Avatar which mimicked the user action, visuals and audio after collecting a coin and reaching the finish line.

Different stages in the game

9. Concept Testing: Wizard of Oz

The main challenge in this stage was identifying measurable system evaluation metric. We realised that a longitudinal study was required to evaluate the impact of our game on user’s motivation level to exercise. Because of the limited time, we evaluated sub system goals instead — reception level, enjoyment, usability and effectiveness of the exercise.

Our key evaluation questions were

Reception — What do the elderly think about mixing exercises with game?
Enjoyment — What aspects of the game interest them and what don’t?
Enjoyment — How to make the game more engaging? 
Exercise — Is the game an effective form of exercise?
Usability — Do they face any difficulties while performing the kayaking gesture? 
Usability — Are they able to understand the interface?

Because of the mid-fidelity of our prototype we conducted a Wizard-of-Oz testing — a“wizard/moderator” observed the participant’s actions and control the game accordingly using keyboard. The participants were given instructions on how to kayak towards the finish line and collect coins. For a realistic setting, we conducted the sessions in the exercise room.

Considering the nature of our research questions, we collected both behavioural data (facial expressions, speech, movements) and attitudinal data (post evaluation questionnaire). As we had a good rapport with the elderly residents by this time, we were afraid that they won’t give honest feedback verbally. We administered a paper based questionnaire to get more candid responses.

The community manager helped advertise the study through community noticeboards. Despite this effort, very few residents initially showed up. More people started to trickle in later; through word of mouth. We observed snowball sampling effect.

Exercise room with the test environment setup

10. Insights from concept testing

We had a total of 11 participants (8f, 3m) with an average age of 76.5. We received tactical as well as strategic feedback from them.

Reception

The concept of exercise game was well received (all but one liked it)

An indirect indicator of reception was participants recommending the game to other residents. We didn’t think of this as a success metric before the test.

Enjoyment

They had varied reasons for liking the concept — fun, real life activity instead of sterile exercise routine, pleasant imagery and challenge of collecting gold coins.

We also observed the excitement while collecting coins and the smile on their face after seeing the end score and hearing the applause.

Despite being just 4 minutes, few participants complained that the game was long. We did make something that was enjoyable / attractive enough to play in the first place but we had to figure out how to create sustainable interest.

Effectiveness of exercise

We didn’t find the participants to be tired; which was a concern because we were building an exercise game after all. We had to figure out an optimum duration of the game customized according to the partcipant.

Usability

The game was rated as easy to understand in the post evaluation survey(4.18 on 5; with 5 being easy to understand).

Some participants asked how much time was left for the game to finish. This was an indicator to add a route map as a progress indicator.

Participants were delighted to hear an applause at the end of game. This was an opportunity to introduce more rewards throughout the game for example after collecting a coin.

They rated the kayaking action as easy. (4 on 5; 5 being most easy). But we observed that halfway through the game majority participants weren’t performing the exact kayaking action. My teammate later offered the participants an oar prop (stick) to help with the kayaking action. Maybe this could be a permanent prop for our game.

What participants did was different from what we observed.

11. Reflection on challenges and learning

This project is very close to my heart because of the involved challenges and learning.

Initially, we were turned down by a few assisted living communities because they were unsure of our intentions. In later attempts, I think emphasising on our affiliation with a research institute and being guided by a professor helped us demonstrate trustworthiness. We were later able to build trust and rapport through frequent visits, participation in their social activities and yes genuine interest.

Conversations with elderly went off the tangent quickly and we later started factoring in extra time for that and getting better at moderating.

Because it was my first time in the United States and working with a community far different from me, there was a cultural gap. It was personally difficult for me to think of relevant ice-breakers. I started observing how my American teammates handled these conversations and learning from them.

What users say could be different from what they do.

Sometimes there could be indirect indicators of system success. In our case, it was residents participating in the test just based on other participant’s recommendation. All one need to do is to keep an eye out.

Also, as with any other school project, we didn’t iterate on this project later. Retrospecting, there are multiple areas which we could have later worked on — how to evaluate the impact on motivation to exercise, determining the effective duration of the exercise game, understanding which games resonated with users. But most importantly deploying the solution in the Campbell Stone assisted living community so that the elderly could have some more fun; we were doing this for them after all!