ArticlePDF Available

Abstract

Background: While advancements in technology have encouraged the development of novel prompting systems to support cognitive interventions, little research has evaluated the best time to deliver prompts, which may impact the effectiveness of these interventions. Objective: This study examined whether transition-based context prompting (prompting an individual during task transitions) is more effective than traditional fixed time-based prompting. Methods: Participants were 42 healthy adults who completed 12 different everyday activities, each lasting 1-7 minutes, in an experimental smart home testbed and received prompts to record the completed activities from an electronic memory notebook. Half of the participants were delivered prompts during activity transitions, while the other half received prompts every 5 minutes. Participants also completed Likert-scale ratings regarding their perceptions of the prompting system. Results: Results revealed that participants in the transition-based context prompting condition responded to the first prompt more frequently and rated the system as more convenient, natural, and appropriate compared to participants in the time-based condition. Conclusions: Our findings suggest that prompting during activity transitions produces higher adherence to the first prompt and more positive perceptions of the prompting system. This is an important finding given the benefits of prompting technology and the possibility of improving cognitive interventions by using context-aware transition prompting.
Prompting technologies: A comparison of time-based and
context-aware transition-based prompting
Kayela Robertsona, Cody Rosascoa, Kyle Feuzb, Maureen Schmitter-Edgecombea,*, and
Diane Cookc
aDepartment of Psychology, Washington State University, Pullman, WA, USA
bDepartment of Computer Science, Weber State University, Ogden, UT, USA
cSchool of Electrical Engineering and Computer Science, Washington State University, Pullman,
WA, USA
Abstract
BACKGROUND—While advancements in technology have encouraged the development of
novel prompting systems to support cognitive interventions, little research has evaluated the best
time to deliver prompts, which may impact the effectiveness of these interventions.
OBJECTIVE—This study examined whether transition-based context prompting (prompting an
individual during task transitions) is more effective than traditional fixed time-based prompting.
METHODS—Participants were 42 healthy adults who completed 12 different everyday activities,
each lasting 1–7 minutes, in an experimental smart home testbed and received prompts to record
the completed activities from an electronic memory notebook. Half of the participants were
delivered prompts during activity transitions, while the other half received prompts every 5
minutes. Participants also completed Likert-scale ratings regarding their perceptions of the
prompting system.
RESULTS—Results revealed that participants in the transition-based context prompting
condition responded to the first prompt more frequently and rated the system as more convenient,
natural, and appropriate compared to participants in the time-based condition.
CONCLUSIONS—Our findings suggest that prompting during activity transitions produces
higher adherence to the first prompt and more positive perceptions of the prompting system. This
is an important finding given the benefits of prompting technology and the possibility of
improving cognitive interventions by using context-aware transition prompting.
Keywords
Prompting technology; cognitive intervention; assistive technology; cognitive aids
*Corresponding author: Maureen Schmitter-Edgecombe, Department of Psychology, Washington State University, Pullman, WA
99164-4820, USA. Tel.: +1 509 335 0170; Fax: +1 509 335 5043; schmitter-e@wsu.edu.
Conflict of interest
The authors report no conflicts of interest.
HHS Public Access
Author manuscript
Technol Health Care. Author manuscript; available in PMC 2016 March 22.
Published in final edited form as:
Technol Health Care. 2015 ; 23(6): 745–756. doi:10.3233/THC-151033.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
1. Introduction
Advances in technology have fueled new opportunities for health care assistance,
particularly in the field of neurorehabilitation where individuals often need reminders or
prompts to assist with activity initiation and completion [1]. Research suggests that
prompting technologies, that is, any form of verbal or non-verbal intervention delivered to
the user [2], can be beneficial for individuals with cognitive impairments [3–5]. Prompting
technologies can be as simple as reminders, notifications, and alerts, but also can be more
complex, such as machine learning algorithms that monitor an individual’s behavior and
prompt during specific tasks [3,6,7]. To further development of automated prompting
technologies, we evaluated the importance of prompt timing by investigating whether
transition-based prompting is more effective than time-based prompting.
A fixed time-based prompting system delivers prompts based solely on a pre-specified and
inflexible time, similar to how a kitchen timer or alarm clock works. Time-based prompting
has been shown to be effective in promoting task engagement, task completion, and focus
[8,9]. For example, individuals with intellectual disabilities became more engaged in tasks
after being taught to look at instructional cards when prompted at predetermined times [10].
In addition, several studies have shown an increase in prospective memory performance
after subjects with brain injuries used a paging device that delivered time-based prompts for
task completion [11,12].
One limitation of time-based prompting is that it may be delivered when the user is engaged
in another important task. Also, time-based prompting often requires the user to do extra
work (i.e., programming a schedule). For example, Ferguson, Myles, and Hagiwara [13]
found that while using a handheld computer helped a child with autism complete IADL’s
more independently, the device was very hard for the parents and teachers to use and also
required a significant amount of user input before it was effective. A user may also become
annoyed after hearing a prompt to do a task that has already been completed and problems
may arise due to the dynamic nature of some daily activities. An untimely prompt can also
increase cognitive load and reduce user attention to the intervention [14]. According to
Seelye and colleagues [15], incorporating activity awareness into time-based prompting
would address these and other similar problems. Time-based prompting with activity
recognition effectively describes activity-aware prompting, a class of context-aware
services.
Context-aware prompting systems use the environment and the status of a person (e.g., an
individual’s location) to recognize effective prompting contexts. In location-based context-
aware prompting, prompts are provided based on the location of the user, often utilizing
GPS technology (e.g., prompt when near a grocery store [16], prompt based on a to-do list
[17]). For example, Frazer and colleagues [18] devised a system to improve GPS navigation
in electric vehicles that decreased glancing at the GPS interface while driving by prompting
the driver based on relevant location information. Recently, Ramanathan and colleagues [19]
began exploratory research on the development of a smartphone-based system, and other
studies have used the Wizard-of-Oz techniques or wireless sensor networks to develop a
prompting system that improves indoor wayfinding and/or task initiation for individuals
Robertson et al. Page 2
Technol Health Care. Author manuscript; available in PMC 2016 March 22.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
with cognitive impairments [20–22]. Similar to time-based prompting, the main limitation of
location-based prompting is that the most effective time to prompt may not be dependent on
the location of the user, but on the activities the user is or has been engaged in.
Context-aware systems generally use a fixed set of parameters such as time and location to
identify a current situation. Not only is the parameter set typically fairly small, but the
parameters are considered separately. In contrast, activity-based context-aware prompts take
into account a richer state description, expressed as the user’s past and current
automatically-recognized activity. For example, Lancioni and colleagues [23] found
improvements in reading ability when a computer sent a tactile prompt upon registering
inactivity while a child with ADHD was told to read information off of a screen. Another
study used a gestural recognition system to prompt individuals to complete various
vocational activities, which also proved useful in increasing task success rates [24]. Several
studies found that activity-based context-aware prompting helped people remember to take
their medications and increased treatment adherence compared to time-based prompting
[25,26]. These systems improved upon the prompting systems described previously by
prompting only when the machine-learning algorithm predicted that the user might not take
the medication, and using multiple methods of determining context, which included motion
sensors and a wearable location sensor. Despite an improvement over classic time-based
prompting methods, context-aware methods of prompting still have limitations. Most studies
done using context-based prompting methods still require user intervention to make the
prompting work, such as requiring user feedback, or requiring a user to input a schedule
[15].
This study aims to provide prompts at times when it would be most opportune for the user to
receive them, and therefore respond to them. When delivery of a prompt is based on time or
location, the user may or may not be currently engaged in an activity where prompting
would be inappropriate or inconvenient. To address this concern, prompting during
transition periods, a period of time when the user is not engaged in an activity and may be
transitioning between activities, has been suggested as an effective prompting time
[14,27,28]. Instances of information overload can occur when prompts are delivered during
an activity [27], and individuals typically perform better on tasks if they are only doing one
activity at a time [29,30]. Additionally, interruptions during a task can increase the risk of
making errors on that task [31–33]. Therefore, if prompting a user to perform an activity
occurs while the user is engaged in another activity, time, number of errors and task
completion of each activity may be affected.
This study assessed whether activity transition-based prompting improves use and
perceptions of prompts to use a digital memory notebook compared to traditional time-based
prompting. Participants completed twelve tasks of everyday living (such as cooking,
cleaning, and watching television) in a smart home. While completing the tasks, participants
received prompts delivered by the experimenter to record information into a digital memory
notebook either every five minutes (time-based condition) or during activity transition
periods (transition-based condition). At the end of the experiment, participants were asked to
critique the timing of the prompts in order to evaluate the effectiveness of the prompt
timing. Based on dual-task management theory, which suggests that prompting individuals
Robertson et al. Page 3
Technol Health Care. Author manuscript; available in PMC 2016 March 22.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
during a transition period (rather than during a task) will lessen cognitive load [27], we
hypothesized that transition-based prompting would result in a greater number of responses
to the first prompt and greater user satisfaction compared to time-based prompts.
2. Methods
2.1. Participants
Participants were 42 undergraduate students recruited from a local university. They were
randomly assigned to either the transition-based prompting condition or the time-based
prompting condition. The majority of students received credit in an introductory psychology
course in return for their time.
2.2. Procedure
Participants completed an hour-long testing session in a campus smart apartment.
Participants were asked to complete twelve everyday tasks in a predetermined randomized
order, including putting together a puzzle, dusting the living room, gathering cooking
ingredients, and reading a magazine (see Table 1 for complete list). Each task took between
one and seven minutes to complete. To ensure smooth activity transitions, the experimenter
read the directions for the next task while the participant was still completing the previous
task via an intercom system. While completing the experiment, participants received
experimenter initiated audio prompts via an Android tablet. The prompts asked whether they
would like to record the completed activities into a daily log on the tablet. If they chose to
do so, participants entered the completed activities by either typing on the keyboard or using
the voice-to-text system.
Participants were randomly assigned to either the time-based or transition-based prompting
condition. For the time-based condition, prompts were delivered every five minutes. For the
transition-based condition, prompts were delivered during each of the transition periods
between activities and followed strict transition start and stop guidelines (see Table 1).
Activities began once participants retrieved the first item required to perform the task at
hand, and ended once the last item used to perform the current task was returned to its
original position. Transition start times were marked by the participant ending their current
task (e.g., the participant is done dusting and is putting away the duster). Transition stop
times were defined by the participant beginning a new task (e.g., the participant picks up the
magazine to start reading). Of note, there was a small variance in the number of transition-
based prompts delivered between participants in the transition-based condition due to lack of
clear transition points between certain tasks depending on the randomized order of the tasks.
If the participant chose to disregard the original prompt, another prompt was issued one
minute later. Participants received the following information about the prompts, “if it’s a
convenient time for you, record what activities you have completed when the memory
notebook prompts you. If the memory notebook prompts you at a bad time, that’s okay, it
will re-prompt you in a few moments and you can respond then”.
While participants were completing the tasks, the experimenter observed from a separate
room, watching participant performances through live feed video and communicating
Robertson et al. Page 4
Technol Health Care. Author manuscript; available in PMC 2016 March 22.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
through an intercom system. A number of motion sensors installed in the smart apartment
were used to track the participant’s movements. The experimenter used the Real-time
Annotation Tool (RAT) [34] to record what the participant was doing, when activity
transitions occurred, and when prompts were delivered. The RAT is a program developed to
replace paper-pencil coding by using a computerized annotation system that allowed
experimenters to accurately time-stamp participant’s actions sequentially. Smart home
sensor data and experimenter notes from the RAT were automatically stored in an SQL
(Structured Query Language) database for later analysis.
2.3. Outcome measures
To investigate the effectiveness of time-based versus transition-based prompting we
included multiple outcome measures. First, we recorded the number of times each
participant used the memory notebook and how many first prompts were administered to
each participant (which excluded re-prompts). Additionally, we coded whether participants
responded to the first prompt, the time at which prompts were delivered, and the time at
which participants responded to the prompts by writing in the memory notebook (i.e., during
an activity, which meant interrupting the activity, or during a transition period). We also
recorded how long it took participants to complete all twelve tasks. At the end of the
experiment, we had participants complete several questions to evaluate their perceptions of
the prompting system. Participants were asked whether the prompts were delivered at
convenient times, whether the timing seemed natural, and whether the timing seemed
appropriate. Each question was rated on a seven-point scale: strongly agree, agree, slightly
agree, neutral, slightly disagree, disagree and strongly disagree, with higher ratings
indicating a better system. Six participants (three in each condition) did not receive the
questions due to experimenter oversight.
3. Results
3.1. Analyses
To test our hypotheses, a series of independent samples t-tests were performed on the
outcome measures. For the main outcome measures, we also calculated Cohen’s d values to
examine effect sizes. General guidelines for interpreting Cohen’s d effect sizes are as
follows: 0.2 is a small effect size, 0.5 is a medium effect size, and 0.8 is a large effect size
[35]. We began by comparing participants in each group to examine whether differences
between age, education, or gender existed. To ensure that the experimental manipulation
was effective, we then examined whether the transition-based condition received more
prompts during transition periods compared to the time-based condition. We then conducted
t-tests to determine whether there were differences between conditions in the number of
prompts that yielded a response to the first prompt. T-tests also evaluated whether the
conditions differed on several other variables including (a) the number of times the
participant’s response to the prompt interrupted the ongoing task, (b) the number of times
the participant used the notebook during a transition period (a pre-determined period of time
indicated in Table 1), and (c) the time in took participants to complete all twelve tasks. In
addition, examination of differences between conditions in perception of the prompting
system was evaluated using t-tests. Finally, correlation analyses were conducted to examine
Robertson et al. Page 5
Technol Health Care. Author manuscript; available in PMC 2016 March 22.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
associations between the prompting variables and participants’ perceptions of the prompting
system. Given the number of correlations conducted, the significance level was set at p <
0.01.
3.2. Comparing participant variables
First, we compared age, education, and gender between participants in each condition. Of
note, not all participants completed the questionnaire form; therefore, we do not have data
on age, education, and gender for all of our participants. Participants in the transition-based
condition had an average of 12.42 years of education and those in the time-based condition
had 12.47 years of education, which was not statistically significant, t(32) = −1.16, p = 0.27.
Mean age of participants in the transition-based (19.88 years) and time-based (21.59 years)
conditions also did not differ, t(32) = −0.24, p = 0.81. Furthermore, there were no
differences between gender in the two conditions, X2(1, N = 32) = 0.00, p = 1.00 (6 males,
11 females, and 4 unknown in each condition).
Due to variance in participant’s speed of completing each of the tasks, participants did not
receive an equal number of prompts. Also, participants in the transition-based condition
were not always administered the same amount of prompts due to lack of clear transition
points between certain tasks depending on the randomized order of the tasks. However, there
was no significant difference between the number of first prompts administered to
participants in the transition-based condition (M = 10.00, SD = 1.95) and the time-based
condition (M = 9.52, SD = 1.40), t(40) < 1; therefore, we were able to use the raw scores in
the data analyses. The results remained the same when the data were examined as proportion
scores by adjusting for the number of prompts administered.
3.3. Comparisons of prompting responses
Participants in the transition-based condition received significantly more prompts during
transition periods (M = 9.14, SD = 2.31) than participants in the time-based condition (M =
0.70, SD = 0.80), t(40) = 15.12, p < 0.001, d = 4.89, indicating that the experimental
manipulation was effective. Overall, there were no differences between the transition-based
condition (M = 7.95, SD = 3.38) and the time-based condition (M = 8.05, SD = 2.91) in the
number of times that the memory notebook was used, t(40) < 1. As seen in Table 2,
participants in the transition-based condition responded to the first prompt an average of
7.05 times, while those in the time-based condition responded an average of 4.24 times,
which represented a significant difference between conditions, t(40) = 3.19, p < 0.005, d =
0.98. These findings support our main hypothesis that participants in the transition-based
condition would be more likely to respond to the first prompt than those in the time-based
condition.
Participants in the transition-based condition interrupted an ongoing tasks 0.67 times on
average to write in the memory notebook, while participants in the time-based condition
interrupted tasks 2.76 times on average, t(40) = −4.85, p < 0.001, d = 1.49 (see Table 2).
Participants in the transition-based condition used the memory notebook an average of 7.29
times during transition periods while participants in the time-based condition used the
memory notebook during transition periods an average of 5.29 times, which trended towards
Robertson et al. Page 6
Technol Health Care. Author manuscript; available in PMC 2016 March 22.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
statistical significance, t(40) = 2.00, p = 0.05, d = 0.62 (Table 2). Of note, the large number
of responses at the transition periods by participants in the time-based condition suggests
that even though prompts were not administered during transition periods, participants
preferred to wait until a transition period occurred to respond to the prompt by writing in the
note-book. Finally, participants in the transition-based condition completed all of the tasks
an average of 5.54 minutes quicker than participants in the time-based condition, which was
statistically significant, t(40) = −2.34, p = 0.02, d = 0.72.
3.4. Comparison of prompting system ratings
We also expected that individuals in the transition-based condition would have more
positive perceptions of the prompting system. As seen in Table 3, participants in the
transition-based condition rated prompt timing as more convenient, t(32) = 4.04, p<0.01, d =
1.39, with a mean rating of 6 (“agree” that the prompt was delivered at convenient times)
compared to a mean rating of 4 (“neutral” that the prompt was delivered at convenient
times) for the time-based condition. Participants in the transition-based condition also rated
prompt timing as more natural t(32) = 3.95, p < 0.01, d = 1.35, with a mean rating of 6
(“agree” that the prompt timing was natural) compared to a mean rating of 5 (“slightly
agree” that the prompt timing was natural) for participants in the time-based condition.
Finally, participants in the transition-based condition rated the prompt timing as more
appropriate, t(32) = 4.09, p < 0.01, d = 2.57, with a mean rating of 6 (“agree” that the
prompt timing seemed appropriate) compared to a mean rating of 5 (“slightly agree” that the
prompt timing seemed appropriate) for the time-based condition.
We then examined whether individuals’ perceptions of the prompting system were
associated with any of the prompting variables. As seen in Table 4, individuals rated the
system as more convenient when less re-prompts had to be delivered, r(34) = −0.46, p <
0.01, when they were able to respond to the first prompt more frequently, r(34) = 0.54, p <
0.01, and when they responded more frequently to prompts during transition periods, r(34) =
0.43, p < 0.01. Participants also rated the prompt timing as being more natural when less re-
prompts had to be delivered, r(34) = −0.45, p < 0.01.
4. Discussion
Advancements in technology have encouraged the development of novel assistive
technologies, which have enhanced neurorehabilitation techniques and broadened our ability
to aid individuals with cognitive deficits [3–5]. However, because technologies evolve so
quickly, it is often difficult to understand the various components of a particular technology
and what makes it more or less efficacious. For example, although numerous prompting
technologies have been employed, factors that impact the effectiveness of prompts (e.g.,
timing of prompts) are not yet fully understood. Thus, the purpose of this study was to
examine whether transition-based prompting (i.e., prompting an individual during activity
transition) is more effective than traditional time-based prompting. Participants completed
12 different IADLs in a smart home and received either time-based prompts or transition-
based prompts to record information in a digital memory notebook.
Robertson et al. Page 7
Technol Health Care. Author manuscript; available in PMC 2016 March 22.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
Consistent with our hypothesis, participants were more likely to respond to the first prompt
when in the transition-based condition, thereby requiring less re-prompting. Participants in
the transition-based condition rated the prompting system as being more convenient, natural,
and appropriate. This demonstrates that receiving transition-based prompting not only
results in higher user responses to initial prompts, it also produces more positive perceptions
of the system, which is imperative for any novel device. More positive perceptions of the
system were also associated with less re-prompting and ability to respond to prompts the
first time and during a transition period. Participants were also able to complete the tasks
more quickly when transition-based prompts were delivered, which is likely a result of
having less reprompts and less task interruptions. These findings suggest that prompting
systems should try to avoid the need to re-prompt an individual, emphasizing the advantage
of transition-based prompting. Furthermore, our analyses revealed that individuals were less
likely to interrupt a task to record information in the memory notebook when prompted
during transition periods. The dual-task management theory proposes that individuals will
be more likely to respond to prompts that do not interrupt activities because the cognitive
load is lessened [27]. This may also account for the finding that participants in the time-
based condition often waited for a transition period to occur to respond to a prompt, even
though the prompt was administered during a task. Thus, our study suggests that a
prompting system that utilizes knowledge of activity transitions will be more effective than
traditional time-based reminder systems.
This is the first study to evaluate the effectiveness of prompting during activity transitions;
however, the literature has investigated other types of context-aware prompting, such as
location and activity context-based prompting. Location based prompting systems often
utilize GPS systems to alert individual’s when they are near a specific location, which can
be helpful; however, they do not correct for potential problems of information overload.
Activity context-based prompting is similar to transition-based prompting because it
attempts to adjust for selected activity contexts such as when a person is out of the home or
is sleeping. Similar to our results, many of these activity based prompting studies have
concluded that context-based prompting is more effective than traditional methods [25,26].
As Lundell et al. [25] described, “a simple time-of-day rule to trigger an alarm, are not very
effective because the reminder is generated whether or not it is an opportune time or place”
(p. 1). In fact, we also observed that participants who were interrupted during a task were
more likely to respond during transition periods or to interrupt inactive tasks (e.g., reading a
magazine) compared to active tasks (e.g., making oatmeal). This highlights the importance
of integrated prompting systems that take advantage of “opportune” moments. This is
particularly meaningful for real-world applications because many activities will take more
time than those assessed in this study, which would likely result in more delayed (and
potentially forgotten) opportunities to use the memory notebook. Although our results are
similar to other context-based studies, our system defines this “opportune” time as a period
of activity transition, rather than inactivity or engagement in a more docile task or location.
Therefore, future studies would benefit from comparing and/or integrating these different
types of context-aware prompting systems in order to develop the most effective system.
Future studies will also be needed to demonstrate that these findings will generalize to
cognitively impaired populations.
Robertson et al. Page 8
Technol Health Care. Author manuscript; available in PMC 2016 March 22.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
To experimentally test the effects of time-based and transition-based prompting in the
laboratory, we had to set up a design that resulted in the administration of nearly 10 prompts
within a one-hour time period. This is different from the frequency of prompting that would
be expected in the real world environment and may limit generalization. In addition, some of
the activities in this study took only a brief amount of time to complete such that the time
between activity completion and the time-based prompt may have been significantly
delayed. This too may differ from time-based prompting delivered in a real-world
environment.
Although we believe that a transition-based activity-aware prompting system can be useful
in many ways, several hurdles must be overcome before such systems are a reality. In the
current study, the experimenter issued the prompts, which is not practical outside of an
experimental testbed. For this reason, we developed an activity recognition algorithm that
has the ability to detect transitions and deliver prompts automatically during activity
transitions, thus eliminating the need for human intervention [36]. In a separate study, we
evaluated detection of transition periods in scripted and unscripted environments and found
that the recognition algorithm was able to detect transition periods greater than 80% of the
time, with a false positive rate of less than 15% [37]. However, our completed prompting
system would require users to have infrared motion sensors installed in their homes. The
home-based sensor system would require very little infrastructure and set-up, but it can seem
obtrusive. Of note, most studies suggest that this is a small cost to pay if it allows
individuals to remain independent in their homes [15]. Unfortunately, this will limit the
system to in-house use only, but other back-up prompting systems can be employed if the
user is out of their house when a prompt is needed. It may also be possible to integrate this
prompting system into a smartphone, but currently sensors in existing smartphones would
not be able to detect transition periods as accurately as a home-based sensor system. It will
be important to test this system as a whole, particularly with individuals with cognitive
impairment because they are the targeted user for most cognitive interventions. We also
hope to expand the use of prompts incorporated into the digital memory notebook. For
example, instead of prompting only to use the notebook, prompts to complete particular
activities on a to-do-list or prompts to take certain medications listed in the medications
section of the notebook can be added. This type of intervention system could be a useful
compensatory tool for many individuals with cognitive impairments resulting from
neurodegenerative diseases, such as dementia, or acquired neurological conditions, such as a
traumatic brain injury. Although the prompting system has not yet been tested by those with
cognitive impairment, we would expect the results to be the same, if not amplified. We base
this on the theory that transition-based prompting is more effective because it reduces
cognitive load, which will especially be important when working with individuals with
already compromised cognition.
Overall, an effective prompting system can improve the efficacy of cognitive interventions,
minimize the necessity for human assistance, and allow users to feel more independent.
Research has shown that prompting technologies can enhance medication adherence,
improve the use of external compensatory strategies, reduce caregiver burden, and increase
functional independence in those with cognitive impairment [4,5,38]. Our research provides
empirical evidence that prompting during activity transitions produces higher user response
Robertson et al. Page 9
Technol Health Care. Author manuscript; available in PMC 2016 March 22.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
rates to the initial prompt, which is important so that the prompt is not forgotten, and more
positive perceptions of the system, which is crucial for any prompting system. Furthermore,
our proposed prompting system requires minimal user-intervention once set-up is complete
and is ideal for people experiencing memory problems that may need external cues to
successfully complete everyday activities. For people who are experiencing cognitive
difficulties due to traumatic brain injuries, stroke, and neurodegenerative processes, it is
imperative to develop technologies that will allow them to function more independently in
their own homes; thereby reducing cost of health care and burden of caregivers [39].
Therefore, it is essential that prompting systems continue to be investigated and employed in
the most effective manner possible.
Acknowledgments
This work was supported by a grant from National Science Foundation under grant No. DGE-0900781 and the
National Institute of Biomedical Imaging and Bioengineering under grant No. R01 EB009675.
References
1. Cicerone KD, Dahlberg C, Malec JF, Langenbahn DM, Felicetti T, Kneipp S, et al. Evidence-based
cognitive rehabilitation: Updated review of the literature from 1998 through 2002. Arch. Phys. Med.
Rehabil. 2005; 86:1681–1692. [PubMed: 16084827]
2. Das B, Chen C, Seelye A, Cook D. An automated prompting system for smart environments.
Toward Useful Services for Elderly and People with Disabilities. 2011; 86(8):9–16.
3. Bewernitz MW, Mann WC, Dasler P, Belchior P. Feasibility of machine-based prompting to assist
persons with dementia. Assistive Technology. 2009; 21:196–207. [PubMed: 20066886]
4. Boll S, Heuten W, Meyer HM, Meis M. Development of a multimodal reminder system for older
persons in their residential homes. Informatics for Health and Social Care. 2010; 35(3–4):104–124.
[PubMed: 21133767]
5. Boger J, Mihailidis A. The future of intelligent assistive technologies for cognition: Devices under
development to support independent living and aging-with-choice. Neurorehabilitation. 2011;
28:271–280. [PubMed: 21558632]
6. Kaushik P, Intille S, Larson K. User-adaptive reminders for home-based medical tasks. A case
study. Methods of Information in Medicine. 2008; 47:203–207. [PubMed: 18473085]
7. Lamming M, Flynn M. Forget-me-not: Intimate computing in support of human memory.
FRIEND21. 1994:2–4.
8. Lancioni GE, Brouwer JA, Bonter HP, Coninx F. Simple technology to promote independent
activity engagement in institutionalized people with mental handicaps. International Journal of
Rehabilitation Research. 1993; 16:235–238. [PubMed: 8244616]
9. Epstein JN, Willis MG, Conners CK, Johnson DE. Use of technological prompting device to aid a
student with Attention Deficit Hyperactivity Disorder to initiate and complete daily tasks: An
exploratory study. Journal of Special Education Technology. 2001; 16:19–28.
10. Lancioni GE, Coninx R, Manders N, Driessen M, Dijk JV, Visser T. Reducing breaks in
performance of multihandicapped students through automatic prompting or peer supervision.
Journal of Developmental and Physical Disabilities. 1991; 3:115–128.
11. Wilson BA, Evans JJ, Emslie H, Malinek V. Evaluation of NeuroPage: A new memory aid. Journal
of Neurology, Neurosurgery & Psychiatry. 1997; 63(1):113–115.
12. Hersh NA, Treadgold LG. NeuroPage: The rehabilitation of memory dysfunction by prosthetic
memory and cueing. Neurorehabilitation. 1994; 4:187–197. [PubMed: 24525367]
13. Ferguson H, Smith Myles B, Hagiwara T. Using a personal digital assistant to enhance the
independence of an adolescent with Asperger Syndrome. Education and Training in
Developmental Disabilities. 2005; 40:60–67.
Robertson et al. Page 10
Technol Health Care. Author manuscript; available in PMC 2016 March 22.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
14. Okoshi, T.; Nozaki, H.; Nakazawa, J.; Dey, A.; Tokuda, H. Attelia: Reducing user’s cognitive load
due to interruptive notifications on smart phones; IEEE International Conference on Pervasive
Computing and Communications; 2015.
15. Seelye A, Schmitter-Edgecombe M, Das B, Cook D. Application of cognitive rehabilitation theory
to the development of smart prompting technologies. Biomedical Engineering. 2011; 99:1–18.
16. Marmasse N, Schmandt C. Location-aware information delivery with commotion. IN HUC 2000
Proceedings. 2000:157–171.
17. Sohn T, Li K, Lee G, Smith I, Scott J, Griswold W. Place-its: A study of location-based reminders
on mobile phones in UbiComp 2005: Ubiquitous Computing. 2005:232–250.
18. McKimm F, Galli M, Cimolin V. Dynamic navigation system desgin for networked electric
vehicles. Theory, Mehtods, Tools and Practic Lecture Notes in Computer Science. 2011;
6770:156–166.
19. Ramanathan N, Swendeman D, Comulada WS, Estrin D, Rotheram-Borus MJ. Identifying
preferences for mobile health applications for self-monitoring and self-management: Focus group
findings from HIV-positive persons and young mothers. International Journal of Medical
Informatics. 2013; 82(40):38–46.
20. Chang YJ, Wang TY, Chen YR. A location-based prompting system to transition autonomously
through vocational tasks for individuals with cognitive impairments. Research in Developmental
Disabilities. 2011; 32(6):2669–2673. [PubMed: 21741205]
21. Chang YJ, Wang TY. Indoor wayfinding based on wireless sensor networks for individuals with
multiple special needs. Cybernetics and Systems. 2010; 41(4):317–333.
22. Liu, AL.; Hile, H.; Kautz, H.; Borriello, G.; Brown, PA.; Harniss, M., et al. Indoor wayfinding:
developing a functional interface for individuals with cognitive impairments; Proceedings of the
8th international ACM SIGACCESS conference on Computers and accessibility; 2006.
23. Lancioni GE, Coninx R, Manders N, Driessen M, Dijk JV, Visser T. Reducing breaks in
performance of multihandicapped students through automatic prompting or peer supervision.
Journal of Developmental and Physical Disabilities. 1991; 3:115–128.
24. Chang YJ, Chen SF, Chuang AF. A gesture recognition system to transition autonomously through
vocational tasks for individuals with cognitive impairments. Research in Developmental
Disabilities. 2011; 32(6):2064–2068. [PubMed: 21985989]
25. Lundell, J.; Hayes, T.; Vurgun, S.; Ozertem, U.; Kimel, J.; Kaye, J., et al. Continuous activity
monitoring and intelligent contextual prompting to improve medication adherence; Conf Proc
IEEE Eng Medi Biol Soc; 2007. p. 6287-6290.
26. Hayes T, Cobbinah K, Dishongh T, Kaye JA, Kimel J, Labhard M, et al. A study of medication
taking and unobtrusive, intelligent reminding. Telemedicine and Ehealth. 2009; 5(8):770–776.
27. Ho, J.; Intille, SS. Using context-aware computing to reduce the perceived burden of interruptions
from mobile devices; Proceedings of the SIGCHI conference on Human factors in computing
systems; 2005. p. 909-918.
28. Iqbal, ST.; Bailey, BP. Understanding and developing models for detecting and differentiating
breakpoints during interactive tasks; Proceedings of the SIGCHI conference on Human factors in
computing systems; 2007. p. 697-706.
29. Byrne MD, Anderson JR. Serial modules in parallel: the psychological refractory period and
perfect time-sharing. Psychological Review. 2001; 108(4):847. [PubMed: 11699122]
30. Strayer DL, Drews FA, Johnston WA. Cell phone induced attention in simulated driving. Journal
of Experimental Psychology. 2003; 9(1):23–32. [PubMed: 12710835]
31. Adamczyk, PD.; Bailey, BP. If not now when? The effects of interruptions at different moments
within task execution; ACM Conf. Hum. Factors Comput. Syst; 2004. p. 271-278.
32. Bailey BP, Konstan JA. On the need for attention award systems: Measuring effects of interruption
on task performance, error rate, and affective state. J. Comput. Hum. Behav. 2006; 22:709–732.
33. Fogarty J, Ko AJ, Aung HH, Golden E, Tang KP. Examining task engagement in sensor-based
statistical models of human interruptibility. ACM Conf. Hum. Factors Comput. Syst. 2005:331–
340.
34. Feuz K, Cook D. Real-time annotation tool (RAT). Proceedings of the AAAI Workshop on
Activity Context-Aware System Architectures. 2013
Robertson et al. Page 11
Technol Health Care. Author manuscript; available in PMC 2016 March 22.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
35. Cohen, J. Statistical power analysis for the behavioral sciences. Hillsdale, NJ: Lawrence Erlbaum
Associates; 1977.
36. Feuz K, Krishnan N, Cook D. Activity recognition on streaming sensor data. Pervasive and Mobile
Computing. 2014; 20:138–154.
37. Feuz K, Cook D, Rosasco C, Robertson K, Schmitter-Edgecombe M. Automated detection of
activity transitions for prompting. IEEE Transactions on Human-Machine Systems. 2015:1–11.
38. Greenway MC, Hanna SM, Lepore SW, Smith GE. A behavioral rehabilitation intervention for
amnestic mild cognitive impairment. American Journal of Alzheimer’s Disease & Other
Dementias. 2008; 23(5):451–461.
39. Lancioni GE, Brouwer JA, Bonter HP, Coninx F. Simple technology to promote independent
activity engagement in institutionalized people with mental handicaps. International Journal of
Rehabilitation Research. 1993; 16:235–238. [PubMed: 8244616]
Robertson et al. Page 12
Technol Health Care. Author manuscript; available in PMC 2016 March 22.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
Robertson et al. Page 13
Table 1
Tasks performed by participants and their respective transition definitions
Activity Transition start (e.g.,
ending of task) Transition end (e.g.,
beginning of task) Average time on
task (seconds)
Dust the bottom level of the apartment Puts the duster away Begins to retrieve the duster 143.68 ± 81.18
Copy a recipe in the kitchen Puts the last task item back Begins to retrieve the first task
item 248.89 ± 125.96
Pick out an interview outfit and then put it
back Puts the last item of clothing back Begins to retrieve the first item of
clothing 76.97 ± 49.80
Read a particular page in a magazine Puts the magazine back Begins to retrieve the magazine 210.80 ± 128.01
Gather ingredients to make spaghetti and then
put them back Puts the last task item back Begins to retrieve the first task
item 145.70 ± 50.39
Watch television in the living room Turns the television off Turns the television on 288.97 ± 83.49
Collect items on a list and put them in a
picnic basket Puts the last task item back Begins to retrieve the first task
item 246.00 ± 121.31
Work on a puzzle at the dining room table Puts the puzzle away Begins to retrieve the puzzle 325.04 ± 119.50
Sweep the kitchen Throwing away the dirt Begins to retrieve the broom 124.35 ± 44.71
Complete some math problems at the dining
room table Puts the pencil down and has all
questions completed Begins to retrieve the pencil and
begins problems 220.18 ± 98.02
Make a bowl of oatmeal Takes oatmeal out of microwave or
puts the last task item back
(whichever comes last)
Begins to retrieve first task item 206.15 ± 73.32
Play a handheld game at a chair by the door 5 minute time limit is up Begins to retrieve the game 293.25 ± 61.41
Technol Health Care. Author manuscript; available in PMC 2016 March 22.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
Robertson et al. Page 14
Table 2
Comparison of prompting variables in the context-aware transition condition and time-based condition
Dependent variable
Transition-based condition
(n = 21)
M (SD)
Time-based condition
(n = 21)
M (SD) t-test p−value Cohen’s d
Number of memory notebook uses 7.95 (3.38) 8.05 (2.91) −0.10 p = 0.92 0.03
Number of first prompts given 10.00 (1.95) 9.52 (1.40) 0.91 p = 0.37 0.28
Response to first prompt 7.05 (3.34) 4.24 (2.28)** 3.19 p = 0.001 0.98
Task interruptions 0.67 (0.86) 2.76 (1.79)** −4.85 p = 0.001 1.49
Number of uses during transitions 7.29 (3.43) 5.29 (3.00) 2.00 p = 0.05 0.62
Time to complete all tasks 40.13 (7.19) 45.67 (8.11)*−2.34 p = 0.02 0.72
Note:
*p < 0.05,
**p < 0.01,
M = mean, SD = standard deviation, General guidelines for interpreting Cohen’s d effect sizes are as follows: 0.2 = small effect size, 0.5 = medium effect size, and 0.8 = large effect size [35].
Technol Health Care. Author manuscript; available in PMC 2016 March 22.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
Robertson et al. Page 15
Table 3
Comparison of group ratings regarding the prompting system
Questions
Transition-based condition
(n = 17)
M (SD)
Time-based condition
(n = 17)
M (SD) t-test p–value Cohen’sd
System usability
The prompts were delivered at convenient times 6.00 (1.41) 4.06 (1.39)** 4.04 p = 0.001 1.39
The prompt timing seemed natural 6.35 (0.86) 4.71 (1.49)** 3.95 p = 0.001 1.35
The prompt timing seemed appropriate 6.24 (0.83) 4.65 (0.28)** 4.09 p = 0.001 2.57
Note:
p < 0.05,
**p < 0.01,
M = mean, SD = standard deviation, General guidelines for interpreting Cohen’s d effect sizes are as follows: 0.2 = small effect size, 0.5 = medium effect size, and 0.8 = large effect size [35].
Technol Health Care. Author manuscript; available in PMC 2016 March 22.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
Robertson et al. Page 16
Table 4
Correlations between prompting variables and perceptions of system usability
Prompting variables
System usability
Prompting was
convenient Prompt timing was
natural Prompt timing was
appropriate
Number of memory notebook uses 0.25 −0.10 0.05
Number of re-prompts given −0.46** −0.45** −0.40
Response to first prompt 0.54** 0.26 0.30
Task interruptions −0.41 −0.36 −0.34
Number of uses during transitions 0.43** 0.10 0.23
Time to complete all tasks 0.05 0.02 0.01
Note:
*p < 0.05,
**p < 0.01.
Technol Health Care. Author manuscript; available in PMC 2016 March 22.
... Two studies [34,42] (13.33%) focused on medium monitoring, based on smartphones, wearables, and sensors. The remaining two studies [38,43] (13.33%) did not use any kind of monitoring systems, and relied instead on baseline and post-intervention assessments. Light monitoring questionnaires and user self-reported data are certainly easy to implement and can gather relevant and accurate information about the users. ...
... Figure 4 shows the distribution of studies according to the type of intervention. Six of the studies [29,33,36,38,40,43] (40%) did not carry out any kind of intervention. Static interventions were conducted in five of the studies [31,34,35,41,42] (33.33%). ...
... The distribution of studies according to the technology used is summarized in Figure 5. Six of the included studies (42.86%) used mobile phones as a technological solution. They are used on their own in three studies (21.43%) [31,32,43] or in combination with wearables in two studies (14.29%) [34,42] or web applications in only one study (7.14%) [37]. For example, Wirth et al. [31] used mobile phones to report monitoring data via telephone. ...
Article
Full-text available
Brain health refers to the preservation of brain integrity and function optimized for an individual’s biological age. Several studies have demonstrated that our lifestyles habits impact our brain health and our cognitive and mental wellbeing. Monitoring such lifestyles is thus critical and mobile technologies are essential to enable such a goal. Three databases were selected to carry out the search. Then, a PRISMA and PICOTS based criteria for a more detailed review on the basis of monitoring lifestyle aspects were used to filter the publications. We identified 133 publications after removing duplicates. Fifteen were finally selected from our criteria. Many studies still use questionnaires as the only tool for monitoring and do not apply advanced analytic or AI approaches to fine-tune results. We anticipate a transformative boom in the near future developing and implementing solutions that are able to integrate, in a flexible and adaptable way, data from technologies and devices that users might already use. This will enable continuous monitoring of objective data to guide the personalized definition of lifestyle goals and data-driven coaching to offer the necessary support to ensure adherence and satisfaction.
... People who have cognitive impairment like dementia experience difficulty in everyday functional independence and therefore find it difficult to initiate daily tasks [4], [5]. Prompting technologies, that is, any form of verbal or non-verbal intervention delivered to the user, could potentially assist individuals with cognitive impairments [6], [7]. Basic and complex prompting technologies have been shown to increase adherence to instructions, decrease errors on everyday instrumental activities of daily living (IADLs), increase independence and increase activity engagement of individuals with cognitive impairment [8]. ...
... 1) Time-based Systems: these systems deliver prompts based solely on a pre-specified and inflexible time [7]. The most popular example of time-based prompting is Google calendar. ...
... On the other hand, time-based prompts may be delivered when the user is engaged in another important task. A user may also become annoyed after hearing a prompt to do a task that has already been completed [6], [7]. ...
Conference Paper
Full-text available
Intervention strategies can help individuals with cognitive impairment to increase adherence to instructions, independence, and activity engagement and reduce errors on everyday instrumental activities of daily living (IADLs) and caregiver burden. However, to be effective, intervention prompts should be given at a time that does not interrupt other important user activities and is more convenient. In this paper, we propose an intelligent personalized intervention system for smartphones. In our approach, we use context and activity awareness to time prompts when they will most likely be viewed and used. Our result based on real data collected using smartphone motion sensors demonstrate that the proposed approach can detect the time-frame of a user response with an average accuracy of 65% and reduce the inefficiency by 39% , on average, compared to different static time interventions which shows the possibilities and advantages of the proposed system to increase user satisfaction and response rate.
... Finding opportune moments at which users are receptive to mobile crowdsourcing tasks, however, remains unclear. That is, prior research attempts have sought to identify moments at which users are receptive for responding to messages [26], reading news articles [42], reading advertisements [55], responding to interventions [25,44,47], responding to questionnaires [24,41], and so on. By and large, such moments have been found to relate to users' prevailing contexts, including their personal condition (e.g., emotional state [24]), the characteristics of whatever activity they are involved in [24,28], their level of engagement with that activity [28,39], and their social context [17]. ...
... Park et al. [37] identifed four distinct types of breakpoints -long silences; a participant leaving a table; others in a group using their own smartphones; and a participant being left alone -and found that sending notifcations only during these breakpoints resulted in a (54.1%) decrease in perceptions that notifcations were interruptive. Robertson et al. [44] developed a system for delivering prompts during transitions between activities, and showed that this approach was not only more favored by users than during non-transition, but that it made initial prompts more efective. Similarly, albeit not using breakpoints, Isaacs et al. [17] proposed the concept of microwaiting to describe brief times of idleness during an activity, and suggested that people frequently requested content during such moments. ...
... Emergency systems that can help locate a lost or wandering person with dementia are often based on GPS location data [4], [5]. Some systems are even offered by the local emergency medical services (EMS). ...
... This results in either a phone call or, if the caller is unable to speak, transmits the location data to authorised entities [6], [7]. Because of the availability of different system, a basic understanding of the various components may be lacking, as well as the perceived usefulness of these technologies, which require additional usability studies with the primary user group [4], [8]. ...
... This type of technology shows distinct promise as potential ways to prompt people living with dementia to complete ADLs: touchscreen computers, for example, can be utilised by people living with dementia (Boyd et al., 2015;Jodderell & Astell, 2016;Lim, Wallace, Luszcz, & Reynolds, 2013;Smith, Mountain, & Hawkins, 2018). Moreover, studies involving participants with mild cognitive impairment confirm that using technology for prompting has assistive potential (Hedman, Nygard, & Kottorp, 2017;Lanciono et al., 2017;Robertson, Rosasco, Feuz, Schmitter-Edgecombe, & Cook, 2015;Thomas & Marsiske, 2014). What is less clear, however, is whether touchscreen devices could be used effectively by carers to guide people who are living with dementia through a sequence of actions to complete a task and whether this technology is likely to be acceptable to people who may be unfamiliar with it. ...
Article
Objectives: People who are living with dementia typically experience difficulties in completing multi-step, everyday tasks. However, digital technology such as touchscreen tablets provide a means of delivering concise personalised prompts that combine audio, text and pictures. This study was one component of a broader, mixed methods study that tested how an application (app) –based prompter running on a touchscreen tablet computer could support everyday activities in individuals with mild to moderate dementia. In this study we set out to understand the experiences of people living with dementia and their primary carer in using the prompter over a four-week period. Method: We collected qualitative data using semi-structured interviews from 26 dyads, composed of a person living with dementia and their carer. Dyads were interviewed at the start and end of this period. Transcripts were then analysed using thematic analysis. Results: The study identified three overarching themes related to: participants’ attitudes towards the technology; their judgements about how useful the prompter would be; and the emotional impact of using it. Conclusion: Consistent with the Technology Acceptance Model, carers and participants were influenced by their approaches to technology and determined the usefulness of the prompter according to whether it worked for them and fitted into their routines. In addition, participants’ decisions about using the prompter were also determined by the extent to which doing so would impact on their self-identity.
... While mobile devices have become a popular tool for notification of information news and delivery of real-time interventions, pushing notifications can also be detrimental if they interrupt ongoing activities [8] [9]. Prompting during transition periods, a period when the user is not engaged in an activity and may be transitioning between activities has been suggested as an effective prompting time [10]. In the case of adults with memory limitations, mobile delivery of activity prompts been found to support functional independence because the device acts as a type of cognitive prosthesis. ...
Article
Full-text available
Segmenting behavior-based sensor data and recognizing the activity that the data represents are vital steps in all applications of human activity learning such as health monitoring, security, and intervention. In this paper, we enhance activity recognition by identifying activity transitions. To accomplish this goal, we introduce a change point detection-based activity segmentation model which partitions behavior-driven sensor data into non-overlapping activities in real time. In addition to providing valuable activity information, activity segmentation also can be used to improve the performance of activity recognition. We evaluate our proposed segmentation-enhanced activity recognition method on data collected from 29 smart homes. Results of this analysis indicate that the method not only provides useful information about activity boundaries and transitions between activities but also increases recognition accuracy by 7.59% and f measure by 6.69% in comparison with the traditional window-based methods.
... More complex machine learning algorithm systems attempt to adjust for this by using sensors and other technologies to learn and track an individual's behavior 10,12,13 . These systems can prompt users at appropriate times (e.g., when they are not engaged in other activities) and when necessary [17][18][19] (e.g., when user leaves the stove on). For example, Pollack et al., developed a machine-learning prompting system, called Autominder 19 . ...
... More complex machine learning algorithm systems attempt to adjust for this by using sensors and other technologies to learn and track an individual's behavior 10,12,13 . These systems can prompt users at appropriate times (e.g., when they are not engaged in other activities) and when necessary [17][18][19] (e.g., when user leaves the stove on). For example, Pollack et al., developed a machine-learning prompting system, called Autominder 19 . ...
Article
Introduction: Declines in memory and executive functioning often lead to difficulties completing instrumental activities of daily living (IADLs). Prompting technologies have the potential to help promote aging in place by providing support for the initiation and accurate completion of IADLs. In this study, we evaluate preferences of older adults for different levels of prompting support based on subjective and objective measures of cognitive functioning. Method: Participants were 170 community-dwelling older adults split into two cognitive complaint groups: cognitive complaints and few cognitive complaints. After completing six IADL tasks (e.g., organize a pillbox, cook), each participant was asked to make a specific error (e.g., leave stove on) on three of the tasks. They were then prompted to correct the error with one of three different prompt modes: verbal indirect, verbal direct, multimodal verbal direct and video. Results: The cognitive complaints group reported greater preference for the multimodal prompt compared to the few cognitive complaints group. The indirect prompt was the least preferred by both groups. Furthermore, participants who recalled less on objective memory measures preferred more support in terms of prompt mode. Executive functioning did not appear to be related to prompt preference. Conclusion: Level of subjective cognitive complaints and objective memory performance may influence amount of support preferred in a prompt.
Article
Objective: We provide an overview of some biomedical technologies able to relieve everyday impairments in neurological patients. Methods: Two literature searches from 2009 to 2020 were conducted using PubMed, Embase, Cinahl and Scopus. The studies meeting the criteria for eligibility constituted 224 of the 6257 identified studies. Results: The first literature search resulted in the identification of 53 different neurological impairments. The following impairments were selected as the most general: six motor (walking/gait abnormality, paralysis/paresis, hypertonia, dystonia, tremor and ataxia), three cognitive (memory, attention/concentration and executive dysfunction), two sensory (visual and hearing impairments) and three uncategorized impairments (communication impairments, sleep abnormalities and seizures/epilepsies). The second literature search resulted in the identification of 22 biomedical technologies able to compensate or rehabilitate the neurological impairments. Conclusions: This review identified some of the common neurological impairments across diseases and showed that technology can be beneficial for neurological patients by helping them with everyday living. The review also found that different aspects such as personal aspects of the intended users (e.g., impairments) and physical and environmental context of the task play an essential role in the usefulness of the technology. • Implications for rehabilitation • Neurological diseases are globally the leading cause of disability, whereby there is a great need for rehabilitation of neurological impairments. • Assistive technology can compensate for permanent impairments or be used in rehabilitation as an alternative to usual therapy or an adjunct to increase overall therapy time. • This study provides an overview of existing assistive technology and how patients with neurological impairments can benefit from technology.
Article
Full-text available
Individuals with cognitive impairment can benefit from intervention strategies like recording important information in a memory notebook. However, training individuals to use the notebook on a regular basis requires a constant delivery of reminders. In this study, we design and evaluate machine-learning-based methods for providing automated reminders using a digital memory notebook interface. Specifically, we identify transition periods between activities as times to issue prompts. We consider the problem of detecting activity transitions using supervised and unsupervised machine-learning techniques and find that both techniques show promising results for detecting transition periods. We test the techniques in a scripted setting with 15 individuals. Motion sensors data are recorded and annotated as participants perform a fixed set of activities. We also test the techniques in an unscripted setting with eight individuals. Motion sensor data are recorded as participants go about their normal daily routine. In both the scripted and unscripted settings, a true positive rate of greater than 80% can be achieved while maintaining a false positive rate of less than 15%. On average, this leads to transitions being detected within 1 min of a true transition for the scripted data and within 2 min of a true transition on the unscripted data.
Conference Paper
Full-text available
A variety of tools have been developed to bring about the realization of pervasive context-aware systems. However, there are still many challenges faced by such applications when trying to determine the context of who, what, when, where, why and how activities are being performed. One such challenge is obtaining the necessary data to train such context-aware systems. In this paper we present a real-time annotation tool which has been developed to streamline the process of collecting and annotating data in an experimental setting. By decreasing the amount of time needed to collect and annotate data, we can increase the rate at which applications can be developed and trained to recognize the current context and react accordingly. These applications can then deliver relevant information in a timely manner supporting a broad range of tasks. Copyright © 2013, Association for the Advancement of Artificial Intelligence. All rights reserved.
Conference Paper
Full-text available
In today's ubiquitous computing environment where the number of devices, applications and web services are ever increasing, human attention is the new bottleneck in computing. To minimize user cognitive load, we propose Attelia, a novel middleware that identifies breakpoints in user interaction and delivers notifications at these moments. Attelia works in real-time and uses only the mobile devices that users naturally use and wear, without any modifications to applications, and without any dedicated psycho-physiological sensors. Our evaluation proved the effectiveness of Attelia. A controlled user study showed that notifications at detected breakpoint timing resulted in 46% lower cognitive load compared to randomly-timed notifications. Furthermore, our " in-the-wild " user study with 30 participants for 16 days further validated Attelia's value, with a 33% decrease in cognitive load compared to randomly-timed notifications.
Article
Full-text available
Many real-world applications that focus on addressing needs of a human, require information about the activities being performed by the human in real-time. While advances in pervasive computing have lead to the development of wireless and non-intrusive sensors that can capture the necessary activity information, current activity recognition approaches have so far experimented on either a scripted or pre-segmented sequence of sensor events related to activities. In this paper we propose and evaluate a sliding window based approach to perform activity recognition in an on line or streaming fashion; recognizing activities as and when new sensor events are recorded. To account for the fact that different activities can be best characterized by different window lengths of sensor events, we incorporate the time decay and mutual information based weighting of sensor events within a window. Additional contextual information in the form of the previous activity and the activity of the previous window is also appended to the feature describing a sensor window. The experiments conducted to evaluate these techniques on real-world smart home datasets suggests that combining mutual information based weighting of sensor events and adding past contextual information into the feature leads to best performance for streaming activity recognition.
Article
Full-text available
Older adults with cognitive impairments often have difficulty performing instrumental activities of daily living (IADLs). Prompting technologies have gained popularity over the last decade and have the potential to assist these individuals with IADLs in order to live independently. Although prompting techniques are routinely used by caregivers and health care providers to aid individuals with cognitive impairment in maintaining their independence with everyday activities, there is no clear consensus or gold standard regarding prompt content, method of instruction, timing of delivery, or interface of prompt delivery in the gerontology or technology literatures. In this paper, we demonstrate how cognitive rehabilitation principles can inform and advance the development of more effective assistive prompting technologies that could be employed in smart environments. We first describe cognitive rehabilitation theory (CRT) and show how it provides a useful theoretical foundation for guiding the development of assistive technologies for IADL completion. We then use the CRT framework to critically review existing smart prompting technologies to answer questions that will be integral to advancing development of effective smart prompting technologies. Finally, we raise questions for future exploration as well as challenges and suggestions for future directions in this area of research.
Article
The Personal Digital Assistant (PDA), while in use in the mainstream population, is not generally utilized by individuals with disabilities. Using a multiple-baseline design across settings we tested effectiveness of using a PDA to increase the independence of an adolescent with Asperger Syndrome (AS). Results indicated use of the PDA was effective in decreasing the student's reliance on adults to complete tasks at home and at school.
Article
Various compensatory devices typically employed in the rehabilitation of memory and attention dysfunction are reviewed. The authors then describe and evaluate a novel compensatory device, NeuroPage, which appears to operate as a prosthetic, rather than simply a compensatory device. Clinical implementation in terms of training, applications, contraindications, and cost-effectiveness is discussed.
Article
This single-subject study examined the effects of a technological prompting device for assisting a student with attention deficit hyperactivity disorder, predominantly inattentive type, to initiate and complete daily tasks. The subject, a fourth-grade male, and his family were provided with computer software and a paging device with word prompting capabilities. The family, in collaboration with the child's teacher, selected particular behaviors to target for use with the intervention. A reversal design across settings with a third return to baseline and intervention was used to evaluate the effectiveness of the prompting device in assisting the student in remembering daily tasks. Study results demonstrated that the intervention was effective in promoting the occurrence of target behaviors but did not affect overall ADHD symptomatology. In addition, the devise appeared to be more useful in the student's structured school environment than his less-structured home environment. Parent and teacher reported a high degree of consumer satisfaction with the device. Overall, the results suggest that the use of a prompting device may be an effective technological intervention for addressing some of the memory-associated symptoms of ADHD.