Transcript Marcela Rodríguez

Marcela D. Rodríguez
Autonomous University of Baja California, UABC, Mexicali, México
[email protected]

Why knowledge of experts is important to
consider when designing AHA?

How expert knowledge can be used to
abstract design principles for AHA?
Approaches for using expert knowledge
I.
a. Systematic Reviews
b. Heuristic Evaluations
II.
AIS for Elderly Medication
III.
Design Dimensions for AIS
IV.
Conclusions

Essential tool for summarising evidence
accurately and reliably.

Collate empirical evidence that fits prespecified eligibility criteria to answer a
research question.

Provides reliable findings from which
conclusions can be drawn and decisions
made.

PRISMA [Liberati et al, 2009]: 27-item checklist and a
four-phase flow diagram

The researcher may look for each study and counting
them as “votes” about the question focus.
 Ie., in a specific SR conducted in the field of medicine, the
researcher found that, among 35 valid studies, 29 showed a
positive result, while 5 showed no result, and one study
showed a negative result.

Internal comparison of studies, based on their specific
parameters:
 i.e. the negative effect must be due to a different dosage
scheme, while the five studies that showed no result were
conducted in subjects that had a different age distribution in
comparison to the 29 positive ones.

Meta-analysis [optional]:
 the original individual studies are treated as if they
were parts of one larger study
 single and final result summarizes the whole
evidence.

It is important to include descriptive evaluation
of each study
 tabular format or tables including population,
oucomes, evaluation, etc.
 This help to determine if the studies results can be
pooled to a metaanalysis








Formulation of a research question.
Define elegibility criteria
Exhaustive searching of literature
Assessing the literature for elegibility
Data extraction.
Summary and synthesis of study results (meta-analysis).
Interpretation of the results.

Set of methods that are all based on having
evaluators inspecting a user interface.

Identify Usability Problems

Inspectors evaluate the UI to check that it
conforms with a set of design characteristics

Inspector: domain and usability expert
 Non-experts: Colleague, potential users
“A small set of evaluators (3-5) examine the
interface and judge its compliance with recognized
usability principles (the heuristics)” [Nielsen, 1994]

Different evaluators
identify different UP

More evaluators, more UP
http://www.nngroup.com/articles/ten-usability-heuristics/


Prepare tasks descriptions
Individual Inspection Session
 1-2 hour session, double revision
 Each inspector prepare a list of UP

Analysis of Heuristic Inspection Data
 Integrated list of UP

Interpretation of Heuristic Inspection Data
 Set of Design Recommendations

Less expensive than users observation

Experts tend to suggest solutions

Large number of (obvious) errors are
discovered before users observation

Inspections complement users observation




Rely on the inspectors ability to predict the
kind of usability problems.
Ideally, an inspector is expert in HCI and the
domain.
Make false assumptions of what actual users
will do with the UI.
Inspectors may have their own preferences
and views toward the UI design.
Instructor:
Marcela D. Rodríguez (UABC, México)
[email protected]

Definition: ”taking the prescribed number of
pills each day, within a prescribed period”

It is reported as the % of medication taken

Patients with chronic conditions has a low
adherence

Rates of adherence drops most dramatically
after 6 months of therapy

26% to 59% of older adults have lower drugs
compliance [Orwing et al, 06]

Elders living alone, and taking more than 3
medicines are prone to medication errors

Forgetting to take their drugs, taking
incorrect doses, and to terminate drug-taking
prematurely[Lison et al, 2006].
Use a familiar language
(Fisk et al, 2003).
Easy to
comprehend
Based on their abilities
and necessities (Rogers &
Mynatt, 2003)
Aesthetic
Assistive
Technology
Easy to
use
Elders prefer aesthetic
assistive producst
(Forlizzi, 2004) (Hirsch et al, 2000).
Integrated
into the
environment
Integrated into their
ADLs
They are aesthetically pleasing applications
that provide valued information through ambient
changes and abstract representations,
that are easy to comprehend:
Ambient Displays
(Wisneski et al, 1998)
Peripherical
Displays
(Matthews et al, 2004)
AIS
Notification
Systems
(McCrikard et al, 2001)

17 older adults between 65 and 86 years old,
who take between 3 and 10 medicines

40-minute contextual interviews (MedMaIDE
index)[Orwing 06]:

What elders know about their medication
 The place where they medicate
 supporting resources they use for medicating
 involved community.
Marcela D. Rodríguez, UABC
21
Elders’ Strategies
Goal
Not forget taking their
medicines.
Have a specific
place to medicate Have the objects they
need to medicate
Sort their
medicines
Evidence
(11/17) Kitchen
(4/17) Bedroom
(2/17) Living room
Avoid taking incorrect
doses or medicines
(6/17) use containers
Write down notes Remember the health
indicating the
problem addressed by
medicine purpose their medicines
Visit periodically
to their doctors
Remember the
appointment to refill
the medicines
(3/17) write the medicine purpose
(17/17) use calendars or agendas
22
Presence of psychological problems, particularly
depression
Presence of cognitive impairment
Treatment of asymptomatic disease
Side effects of medication
Patient’s lack of belief in benefit of treatment
Motivators
Reminders
Guidance
Missed appointments
Complexity of treatment
Technological Approaches
Electronic medicationmonitoring devices
MEMS www.aardex.com
Helping Hand (Bleser et al, 2010
Reminder systems
Medtracker (Hayes et al, 2009)
Persuasion through social
games
Promote reflexion
Cadex www.e-pill.com
Movipill
(Oliveira et al, 2010)
Dispositivo para dar seguimiento
(Lee y Dey, 2011)
Representación de lo datos
CareNet Display
UbiFitGarden
Reminder
CareNet
Display
Bracelet
(Consolvo
2008)
(Mynatt,etet
2003)
(Hansson
alal,2000)
Fish’n’Steps
The Orb
BusMobile
(Lin et alet2006)
(Mankoff
al 2003

AIS for making users aware of
the status of relevant and
unpredictable events.

AIS that notify of predictable
or scheduled events.

AIS to motivate a change on
the users behavior.
Marcela D. Rodríguez, UABC
25
• Low cognitive effort, easy to monitor (Mattews et al, 2004)
 Aesthetic design
• Use of semiotics (Pousman y Stasko, 2006).
•AIS
should be perceived
for theappropriate
information provided
Modalities
for grasping
amount
• Promote
(Mankoff
et al, 2003).reflexion of users abour their behavior (Consolvo, et al,
of users attention
2009)
sources of
information
(Pousman & Stasko, 2006)
•• Use
Five few
notification
levels
for managing
user attention (Pousman
06; Matthews et al, 2004).
 Modalities
to represent
information
•Consider
the users’
life style, and
cultural aspects (Jafarinaimi, et al
•2005)
Criteria for measuring user’s interruption, reaction and
comprehension (McCrickard et al, 2001)
 How to apply or adapt these AIS design
attributes to support elders medication?
26
Remind
Guide
AIS Strategies
Motivate
To not abandon the
therapeutic treatment:
• Psicological and persuasive
techniques
• Abstract representations
• Historical information and
reflection
• Reinforcements


A stylized
representation of
medication adherence
Metaphor: Birds
growth to symbolize
repercussions on
elders health.
Current medication week
Rewards:
“Periquito Curro”
3
2
1
Growth levels
0
Medication adherence during the
3 past weeks
Remind
Guide
AIS Strategies
Motivate
Initiate the activity:
• Provide a stimulus to
demand user’s attention
• Make elder’s aware of the
importance of medicating.

Timely auditory
reminders
Medicine
Enalapril

Visual notifications
with critical
information
Health
problem
20:00
Doses
Marcela D. Rodríguez, UABC
Remind
Guide
AIS Strategies
Motivate
Complete the activity:
• Atract users attentions to the
containers medicines
• Provide critical information to
correctly take the medication.

AIS attached to medicines to make elders
aware of how to carry out the activity.
Doses
Frequency
Health-problem
Marcela D. Rodríguez, UABC
32
Periquito
curro
33
21:00
La Sra. Ana leyendo su novela
Marcela D. Rodríguez, UABC
35

Objective:
 Identify usability problems
 Design recomendations
 Validate design dimensions

17 Inspectors:
 5 Domain Experts (DE): Elderly healthcare
specialists
 7 Novice Experts (NE): Computer Engineering or
Master Degree in Computer Science
 5 Advanced Experts (AE): kwnoledge on usability
engineering; C.Sc. Doctoral students
Juan P. García
Avance de Tesis
36
Introduce
evaluation
Explanation
of the systems
Feature check list (Mankoff 03):
C1- Useful and relevant information
C2- Peripherality
• Discussion of:
C3- Match between system and
• Usability problems
real world
C4- Sufficient
design
• information
Recommendations
C5- Consistent and intuitive mapping
• Barriers elders face
C6- Easy transition
C7- Visibility of status
C8- Aesthetic and pleasing design
C9- Strategy to assist
C10- Target object of the AIS
Individual
Inspection
to adopt systems
Group
Discussion
Frecuency Problem
12/17
3
2
Dificulty for perceiving changes on the parakeets;
therefore it is not easy to associate the parakeets behavior
with the elderly medication compliance.
1
Growth levels
0
DE5(0)-C6: “Identifying
the multiple levels of
growth is difficult for the
patients”
AE1-C5-R: “Include
other parakeet
behaviour
characteristics: such
as emotions, make
the parakeet seem
seek”
Frecuency Problem
It is difficult to identify the medication week represented
5/17
by each parakeet.
AE5(0)-C6: “It will not be
easy for an older adult to
identify in which week he
correctly or incorrectly take
his medications”
AE1-C1-R: “use different
colors for each parakeet”
Frecuency Problem
5/17
The system should provide
feedback when taken or
omitted the taken of the
medication.
6/17
The system should provide
reinforcements daily.
DE2(1)-C1: “It would be
useful and more relevant to
provide explicit information
indicating the medicines
taken and not taken, and
schedules”
DE4-C4-D: “elders need to
receive an immediate
reinforcement associated
with their actions…”
DE1-C5-D: “Provide
personalized feedback:
¡Arthur, you did it; “…based
on the compliance
situation”
Frecuency
7/17
Problem
Using images of food, may cause elderly associate them
with medication instructions, instead of associating it
with the health problem.
Enalapril
20:00
DE4-D: “… the salt image may
confuse al older adult; he may think
that ir is time to consume salt”
AE3-C5-R: “I suggest to modify the
image to include a symbol to indicate
if it is a recommended or restricted
food”
Frecuency
5/17
3/17
Problem
Using high
frequency auditory
notification, may not
be perceived by
older adults.
The system does not
makes elders aware
about thei
medication
complizance.
DE1-C3-D: “some elder’s present
presbycusis, loss of ability to perceive
high frequency sounds or distinguish
between tones, therefore they may not
distinguihed this auditory
notifications”
AE3(0)-C6-D: “The system
does not allow to get
additional information sbout
my medication compliance”.
Frecuency
13/16
8/16
Problem
Asociate the geometric figures with the medication frequency
is a complex task.
The system design is not pleasant
6/16
Interpreting the meaning of the highlited sides is a complex
task
4/16
The luminosity intensity of the sides, may not be perceived
by older adults

CARE-Me should reinforce the elders’ behavior
towards the activity.
 It is important to represent more detailed and
additional information (Sufficient information,
C4=0.81; Easy transition, C6=0.52)
 Daily representations of the medication compliance
and present instructions for medicating (12/17).
 Daily rewards instead of weekly: parakeets notify
appropriate and inappropriate behavior (5/17).
44

Remind-Me is appropriate to demand elders to
timely execute an important activity.
 Audio alerts demand elders to timely execute an
activity (11/17) (Peripherality, C2=-0.17)
 Mobile devices aren’t common objects used by elders
(Harmony with the environment, C3=0.99), however
elders may get used to take them anywhere (13/17).
 Information is sufficient and relevant (C4=1.24), but
(9/17) some experts suggested to provide elders with
additional information (i.e pre-conditions)
45

Tuning of ambient information to elders’
perceptual characteristics.
 Adjust the fidelity of information representation
(metaphors) and notification modality (auditory).
 6/17 evaluators (all geriatricians) stated that the
parakeet whistle may be unpleasant and not easily
heard.
 The AIS should not be in highly harmony with the
elders environment to avoid they stop perceiving the
ambient information (2 advanced experts).
 Include the elder’s name in the rewards phrases: to
grasp her attention and open a trust linkage (all
geriatricians)
46

Basic objects used to guide the activity, should
present enough information to influence elders
to carry out a current action adequately.
GUIDE-Me
2.00
1.50
1.00
0.50
Expertos de
dominio (G)
Domain
Experts
Novatos expertos
Novice
Experts(N)
0.00
C1
Expertos Avanzados
Advanced
Experts(E)
Promedio
Average
-0.50
C2
C3
C4
C5
C6
C7
C8
C9
C1- Useful and relevant information
C2- Peripherality
C3- Match between system and
real world
C4- Sufficient information design
C5- Consistent and intuitive mapping
C6- Easy transition
C7- Visibitily of status
C8- Aesthetic and pleasing design
C9- Strategy to assist
C10- Target object of the AIS
47

Strategy to assist:
 P1- Remember to take their medicines by timely
making elderly aware of taking them
 P2- Guide medication intake by providing the
necessary information to properly perform the
medication routine
 P3- Motivating elders not to leave their medication
by encouraging them to take their medicines and
not to interrupt the recommended therapy

Activity orientation:
 P4- Demand to execute an activity through a
stimulus and critical information that indicate the
importance of performing the activity
 P5- Make elders aware of critical information to
complete properly their medication
 P6- Provide daily reinforcement to motivate elders
to medicate routinely
 P7- Provide immediate feedback to make elders
aware about the medication success or failure

Tuning ambient information
 P8- Auditory notifications should be adjusted to the
frequency range in which elders hear (presbycusis)
 P9- Use pictograms to express the rules to follow for
medicating. They are stylized figurative drawings that
refer directly to an object to communicate.
 P10- Using visual metaphors to raise awareness about
medication compliance; e.g. metaphors based on the
care of pets or plants to provide immediate feedback
and daily reinforcements on medication compliance.
CARE-Me
Enhanced Design
Recompensa
Semanal
Immedate Reinforcement
Remind-Me
Enhanced Design

Guide the design

Evaluate a design

Clasify systems

Identify design patterns
Characterize systems by using the
Design Dimensions proposed
Stratey to assist:
P1- Remember to take their medicines:
Activity Orientation:
Escalating alets
P4- Demand to medicate through a stimulus and critical information:
2 hr. windowTuning ambient information
Use of different ambient modalities: lights, sounds
P2- Guide medication P8- Auditory notifications should be adjusted (presbycusis)
Increment the intensity of reminders
Indicate which medicine
take by emphasizing
therules
cap to follow for medicating.
P9- Useto
pictograms
to express the
P5- Make elders aware of critical information to complete the medication
P3- Motivate elders notP10to leave
their
medication
Using
visual
metaphors to raise awareness about
Only indicate which medicine to take by emphasizing the cap
“Social feedback” medication
: reports cancompliance
be accessed by the elder’s
P6- Provide daily reinforcement
social care network
Do not provide daily reinforcement
* Tuning of the intensity of alerts to not disturb the patient.
P7- Provide immediate feedback
Phone calls only when he missed to take the medicine
Systematic Review:
Conduct a systematic review to select
relevant technologies that enable to
validate the design dimensions of AIS

¿What are the design principles recommended
to follow for developing technologies that
support medication adherence?

Objectives:
 Identify how these technologies address the
design dimensions we proposed for AIS.
 Identify additional design attributes

Journal, and Conference proceedings papers
(we will exclude work in progress papers:
workshops, posters, doctoral symposium)

Report an evaluation of the technology

Report technology design/functionality

Commercial available products*

PubMed Central – includes NLM Catalog,
NCBI databases, MEDLINE, and uses MeSH
(Medical Subject Headings).

ACM – enables to search conferences and
journals from IEEE, Springer, and other
relevant editorials (i.e. Elsevier)

It will enable us not retrieve several instances
of the same paper.

% of papers retrieved

Type of technology approach used (mobile
devices, electronic containers, web tech., etc)

Type of strategy supported

Group age

Type of evaluation (usability, effectivity)
714 Citations identified by
initial electronic search
139 Citations kept
13 Commercial products
sold for elderly.
Selected from:
5 Products excluded:
4 duplicated
1 was an accesory
575 Articles excluded
(do not report technology to
assist medication; work in
progress)
2 Articles duplicated were
excluded
137 Articles to analyse
(meta-analysis)
? Articles excluded
? No containing description of
technology
? No reporting evaluation
8 Commercial products
? Systems to review to
identify design principles

Suggestion 1: Carry out at the beginning of
your research a literature review based on a
systematic review.

Suggestion 2: Carry out the inspection first
with Domain Experts, then with HCI Experts

It is complex to identify “design principles”,
there are no ideal or standard methodology
62
Instructor:
Marcela D. Rodríguez (UABC, México)
[email protected]


Encouraging people to exercise has been a
major challenge over the years.
Elderly who exercise on a regular basis
obtain many benefits:
 to perform their daily activities
 maintain their independence
 improve their self-esteem and relationships
5/21/12
Wellnes Interventions and HCI
64
PLAY, MATE!
Ubifit Garden
[Berkovsky, S. et al 2010]
Flowie
Fish n Steps
[Consolvo et al 2009] [Merino I. et al, 2009]
[James J. et al 2006]
Design strategies
for persuading:
 Representations of exercise routines compliance
 Awareness of the impact of their behavior
 Reinforcements
 Support the user’s personal style
 Competition that promotes socialization
5/21/12
Wellnes Interventions and HCI
65

Challenges to implement these design
strategies for motivating elders to exercise:
- understand the needs and problems they have for
exercising,
- identify the factors that negatively and positively
may impact on their motivation.
 design appropriate awareness mechanisms
according to their perceptual levels
 independent elders prefer products that match
their aesthetic desires and functional needs.
5/21/12
Wellnes Interventions and HCI
66




Identify how the persuasion strategies for
exercising should be adapted for the elderly.
Case study: Identify barriers and
motivators to start and maintain a physical
activity.
Design: Mobile Ambient Information System
to motivate elders to exercise by requiring
minimal cognitive effort.
Evaluation: Validate the design strategies.
5/21/12
Wellnes Interventions and HCI
67



40 minutes semi-structured interviews
10 female and non frail adults
5 active, 5 passive
5/21/12
Wellnes Interventions and HCI
68

Staying healthy is the main motivation of
active elders (5/5)
 Coping with depression episodes (1/5)

Passive elders mentioned different reasons
for not being able to exercise on a regular
basis.
 It is difficult to get the habit of exercising (1/5)
 The lack of social support for exercising in late life
is a key barrier among older (1/5)
5/21/12
Wellnes Interventions and HCI
69

Active elders (a) have a more extended
social network than passive elders (b).
• Abstraction: Metaphor based
on gaining virtual coins
interchangeable for time for
accessing games.
• Triggers: daily reminders and
time-based notifications to
reach an exercising goal.
5/21/12
Wellnes Interventions and HCI
71
• Historical information and
reflection: representation
of the complied, and
planned exercising goals.
• Positive and playful
reinforcement: rewards
that do not make them
feel punished, and
promote socialization.
5/21/12
Wellnes Interventions and HCI
72

A formative evaluation to obtain the elders’
 perception regarding the effectiveness of the
system’s persuasion strategies, and
 assessing the users’ interruption, reaction and
comprehension.


One-day evaluation with 15 persons (63 to 86
years old).
We observed and interviewed the
participants.
5/21/12
Wellnes Interventions and HCI
73

Abstraction. The metaphor was well accepted by
all of them:
 enjoyed to track their goals( 6/15)
 like to obtain “gold coins or diamonds” (6/15)
 wearing the device, which “[it] is modern” (3/15)

Historical information & reflection. Participants
(11/15) emphasized the calendar was useful to
realize if they pursued or have proposed goals:
 “... Realizing that I did not exercise one day, it would
make me to effort for the next day”.
5/21/12
Wellnes Interventions and HCI
74
• Trigger. The notifications are easy to
comprehend, but the ambient noise may impede
they were perceived:
– Half of the participants (8/15) perceived the audible
notifications,
– However, we noticed that some participants (7/15)
observed it while exercising.
• Positive and playful reinforcement. Receiving
access time to play games would be fun and a
useful reward for motivating them (9/15).
5/21/12
Wellnes Interventions and HCI
75
• Playing games may be not an adequate
motivational factor for elderly (12/15)
– 5 participants suggested to use other type of rewards
regarding to manage their health
– “I would like to get permission to eat food restricted to
consume by my physician”; “…a desert”; “a free
consultation with a nutritionist”
– 7 would like to receive rewards that promote
socialization out of their homes:
– “…But I would like to win trips as rewards.”; “…, going
out or a meeting with friends”
5/21/12
Wellnes Interventions and HCI
76
• Visualizing the goals that were not pursued
were perceived (3/15), as a factor that may
negatively impact their motivation
– “[I would like the system to] eliminate copper coins,
since they are negative…”; “… copper coins”;
– “…I would like to see just gold coins”.
5/21/12
Wellnes Interventions and HCI
77



We understood that one of the main
motivators for exercising is to have the
opportunity to extend their current social
network.
They perceived that providing
representations of the compliance of their
short-term goals was useful; however.
We also identified that they perceived as
negative feedback to be informed of the
proposed goals that were not pursued.
78
Marcela D. Rodríguez
Autonomous University of Baja California, UABC, Mexicali, México
[email protected]
CARE-Me
Remind-Me
2.50
2.00
2.00
1.50
1.50
1.00
1.00
0.50
0.00
0.50
C1
C2
-0.50
0.00
C1
C2
C3
C4
C5
C6
C7
C8
C9
GUIDE-Me 1.50
1.00
0.50
Expertos de
dominio (G)
Domain
Experts
0.00
C1
Expertos Avanzados
Advanced
Experts(E)
Promedio
Average
-0.50
C2
C3
C4
C4
C5
C6
NA
C7
C8
C9
-1.00
2.00
Novatos expertos
Novice
Experts(N)
C3
C5
C6
C7
C8
C9
C1- Useful and relvant information
C2- Peripherality
C3- Match between system and
real world
C4- Sufficient information design
C5- Consistent and intuitive mapping
C6- Easy transition
C7- Visibitily of status
C8- Aesthetic and pleasing design
C9- Strategy to assist
80
C10- Target object of the AIS
Instructor:
Marcela D. Rodríguez
Universidad Autónoma de Baja California
Facultad de Ingeniería
Posgrado MyDCI, área Computación