An official website of the United States government
Activity tracking has the potential to promote active lifestyles among older adults. However, current activity tracking technologies may inadvertently perpetuate ageism by focusing on age-related health risks. Advocating for a personalized approach in activity tracking technology, we sought to understand what activities older adults find meaningful to track and the underlying values of those activities. We conducted a reflective interview study following a 7-day activity journaling with 13 participants. We identified various underlying values motivating participants to track activities they deemed meaningful. These values, whether competing or aligned, shape the desirability of activities. Older adults appreciate low-exertion activities, but they are difficult to track. We discuss how these activities can become central in designing activity tracking systems. Our research offers insights for creating value-driven, personalized activity trackers that resonate more fully with the meaningful activities of older adults.
more »
« less
This content will become publicly available on June 9, 2026
From Verbal Reports to Personalized Activity Trackers: Understanding the Challenges of Ground Truth Data Collection with Older Adults in the Wild
Tracking activities holds great potential to improve the well-being of older adults, yet the accuracy of activity trackers for this demographic remains in question. Evaluating this accuracy requires ground-truth data directly from older adults, which has largely been gathered in controlled laboratory settings or labeled by researchers. Moreover, considering the diversity in older adults' activity engagement and tracking preferences, personalized activity tracking appears necessary. We demonstrate that older adults can benefit from personalized activity trackers by showing that cadence thresholds for stepping intensities vary within this group. However, collecting ground-truth data from older adults in real-world settings poses unique challenges. This paper examines two sources of ground-truth labels for the smartwatch Inertial Measurement Unit (IMU) data collected with older adults. Using verbal self-reports and a thigh-worn activity tracker, we assess their viability as ground-truth sources in natural settings. Additionally, we evaluate the costs and benefits of triangulating these sources as a ground-truth proxy. Our findings reveal two main costs: data shrinkage and notable effort from both contributors and data stewards. Simultaneously, we observe improved data quality and a greater ability to identify error sources when evaluating a trained model.
more »
« less
- Award ID(s):
- 1955568
- PAR ID:
- 10639255
- Publisher / Repository:
- ACM
- Date Published:
- Journal Name:
- Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies
- Volume:
- 9
- Issue:
- 2
- ISSN:
- 2474-9567
- Page Range / eLocation ID:
- 1 to 33
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Current activity tracking technologies are largely trained on younger adults’ data, which can lead to solutions that are not well-suited for older adults. To build activity trackers for older adults, it is crucial to collect training data with them. To this end, we examine the feasibility and challenges with older adults in collecting activity labels by leveraging speech. Specifically, we built MyMove, a speech-based smartwatch app to facilitate the in-situ labeling with a low capture burden. We conducted a 7-day deployment study, where 13 older adults collected their activity labels and smartwatch sensor data, while wearing a thigh-worn activity monitor. Participants were highly engaged, capturing 1,224 verbal reports in total. We extracted 1,885 activities with corresponding effort level and timespan, and examined the usefulness of these reports as activity labels. We discuss the implications of our approach and the collected dataset in supporting older adults through personalized activity tracking technologies.more » « less
-
Self-tracking using commodity wearables such as smartwatches can help older adults reduce sedentary behaviors and engage in physical activity. However, activity recognition applications that are typically deployed in these wearables tend to be trained on datasets that best represent younger adults. We explore how our activity recognition model, a hybrid of long short-term memory and convolutional layers, pre-trained on smartwatch data from younger adults, performs on older adult data. We report results on week-long data from two older adults collected in a preliminary study in the wild with ground-truth annotations based on activPAL, a thigh-worn sensor. We find that activity recognition for older adults remains challenging even when comparing our model’s performance to state of the art deployed models such as the Google Activity Recognition API. More so, we show that models trained on younger adults tend to perform worse on older adults.more » « less
-
In this work, we explore how a real time reading tracker can be built efficiently for children’s voices. While previously proposed reading trackers focused on ASR-based cascaded approaches, we propose a fully end-to-end model making it less prone to lags in voice tracking. We employ a pointer network that directly learns to predict positions in the ground truth text conditioned on the streaming speech. To train this pointer network, we generate ground truth training signals by using forced alignment between the read speech and the text being read on the training set. Exploring different forced alignment models, we find a neural attention based model is at least as close in alignment accuracy to the Montreal Forced Aligner, but surprisingly is a better training signal for the pointer network. Our results are reported on one adult speech data (TIMIT) and two children’s speech datasets (CMU Kids and Reading Races). Our best model can accurately track adult speech with 87.8% accuracy and the much harder and disfluent children’s speech with 77.1% accuracy on CMU Kids data and a 65.3% accuracy on the Reading Races dataset.more » « less
-
Summary The accurate estimation of circadian phase in the real‐world has a variety of applications, including chronotherapeutic drug delivery, reduction of fatigue, and optimal jet lag or shift work scheduling. Recent work has developed and adapted algorithms to predict time‐consuming and costly laboratory circadian phase measurements using mathematical models with actigraphy or other wearable data. Here, we validate and extend these results in a home‐based cohort of later‐life adults, ranging in age from 58 to 86 years. Analysis of this population serves as a valuable extension to our understanding of phase prediction, since key features of circadian timekeeping (including circadian amplitude, response to light stimuli, and susceptibility to circadian misalignment) may become altered in older populations and when observed in real‐life settings. We assessed the ability of four models to predict ground truth dim light melatonin onset, and found that all the models could generate predictions with mean absolute errors of approximately 1.4 h or below using actigraph activity data. Simulations of the model with activity performed as well or better than the light‐based modelling predictions, validating previous findings in this novel cohort. Interestingly, the models performed comparably to actigraph‐derived sleep metrics, with the higher‐order and nonphotic activity‐based models in particular demonstrating superior performance. This work provides evidence that circadian rhythms can be reasonably estimated in later‐life adults living in home settings through mathematical modelling of data from wearable devices.more » « less
