Design Research Techniques for Elders with Cognitive Decline: Examples from Intel

1
Design Research Techniques for Elders with
Cognitive Decline Examples from Intels Digital
Health Group

• Jay Lundell, PhD
• Margaret Morris, PhD

2
Techniques Applied
Technology Development
Late
Early
Wide
Ethnography
Concept Feedback
Pilot/Probe Studies
Breadth of Functionality
Clinical Trials
Usability Testing
Narrow
3
Ethnography of Older Adults with Cognitive
Impairment

• 45 Households in US
• Range from normal aging to advanced Alzheimers
• A variety of needs -

4
Ethnography of Older Adults with Cognitive
Impairment

• 45 Households in US
• Range from normal aging to advanced Alzheimers
• A variety of needs -

Balancing foresight and optimism/denial
Denial
Perceived functioning
Actual functioning
Foresight
5
Ethnography of Older Adults with Cognitive
Impairment

• 45 Households in US
• Range from normal aging to advanced Alzheimers
• A variety of needs -

Having an impact Independence and control the
home, finances, relationships Mental
stimulation Physical activity Connection to the
outside world
6
Concept Feedback

• Focus Troupe dramatic scenarios
• Three user groups Normal aging, Mild cognitive
  impairment, Care givers/Boomers
• Focus on context of use, social implications

7
Concept Feedback
8
Context Aware Medication PromptingA pilot study
on the effectiveness of intelligent medication
tracking and reminding

• Methods
• Recruit 25 people over 65 who have difficulty
  with medication adherence (50-80 adherence)
• Six week baseline sensors in the home track
  activities, sleep patterns and when medications
  are taken
• Eight week intervention two types of reminders
  
• 1. basic alarm clock that always goes off at
  medication time,
• 2. context aware prompting that only prompts when
  user is likely to miss a dose (based on data
  collected in baseline)
• Measures effectiveness of reminders (as measured
  by adherence to a pre-determined regimen),
  subjective preference for reminders, ability of
  system to predict non-adherence

Motion Sensor Detect motion in each room in the
house. Also detects front door and refrigerator
door opening
Bed Sensor Detect movement in bed, sleep quality
iMed Tracker Detects when pills are taken
Health Spot Wrist watch that detects location of
subject
Phone Sensor Detects phone calls

• Bayesian inference engine uses data collected
  during baseline to decide when and where to
  deliver a prompt
• Detects activities such as sleep, visitors, on
  the phone, kitchen movement, taking meds

iMed Tracker LED, beep, and text display
Activity Beacon LED, beep, and voice reminder
Health Spot Beep, text display
9
Usability Testing Parkinsons disease

• 4 Patients with Parkinsons Disease and their
  spouses
• Tested for usability, learnability, and
  livability

10
Summary

• Standard design and usability approaches adopted
  for special users
• Downplay technology, emphasize environmental and
  social context
• Pilot technology in extended trials test for
  livability, use over time
• Include stakeholders and design for their needs
  as well