Tips & research
Evidence-informed principles for talking with someone experiencing memory loss — and where they come from.
Communication principles
Offer choices, not open questions
Why? Two options are far easier to answer than an open question.
Give time to respond
Why? Silence is often thinking. Wait before you fill it.
One question at a time
Why? Ask a single, short question and then stop.
Respond to the feeling, not the fact
Why? Name the emotion first. Accuracy can wait.
Long-ago memories stay reachable
Why? Childhood and early adulthood are usually the richest ground.
Tone and touch carry meaning
Why? A warm voice and a hand on theirs often says more than words.
Reflect back what you heard
Why? Say back a few of their words so they know you are following.
Adapt communication to current ability
Why? Adapting how we communicate to how the person is doing today can lower unnecessary demand.
Prefer recognition over recall
Why? Simple choices and recognition are easier than asking someone to recall a specific fact or date.
Connection over correctness
Why? Connection matters more than getting facts right — it's best not to test memory.
Relationship-centered care
Why? What works depends on both the person and who they're talking with.
Familiar and sensory anchors
Why? Familiar topics, music, photos, and the senses can spark connection without demanding precise memory.
Redirect gently
Why? Redirecting gently and validating feelings tends to work better than correcting or arguing.
Reduce demand when agitated
Why? When someone is agitated or tired, lowering the conversational demand and prioritizing comfort helps.
Share first
Why? Sharing something of your own before asking can make a question feel less like a test.
Research behind GrandChatly
GrandChatly incorporates principles informed by published research from clinicians and researchers working in dementia care, geropsychology, caregiving, and dyadic interventions. Its evidence library includes published research from Evan Plys, PhD, ABPP, and other researchers in these fields.
GrandChatly is not created by, advised by, reviewed by, or endorsed by any researcher or institution named here. Publications are referenced only for the principles they actually support.
Sources & evidence
Active Listening (Greater Good in Action)
Greater Good Science Center, University of California, Berkeley
ggia.berkeley.edu, 2024
Why used: Basis for the reflect-back and empathy phrasing used in the conversation-skill prompts.
View sourceAlzheimer's Disease: A to Z
Harvard Health Publishing
health.harvard.edu, 2024
Why used: Explains why distant memories are a safer, warmer starting point than recent events.
View sourceTips for Daily Life: Conversations (Memory and Aging Center)
UCSF Memory and Aging Center
University of California, San Francisco (memory.ucsf.edu), 2024
Why used: Source for the responding-to-feelings, repetition and upset-moment prompts.
View sourceCommunicating With Someone Who Has Alzheimer's Disease
National Institute on Aging, National Institutes of Health
nia.nih.gov, 2024
Why used: Grounds the choice-based and one-question-at-a-time prompts, and the guidance against quizzing or baby talk.
View sourceReminiscence and Reminiscence Therapy
Alzheimer's Association
alz.org, 2024
Why used: Underpins the photo, music and keepsake-based prompts.
View sourceHelping Dementia Caregivers
Centers for Disease Control and Prevention
cdc.gov, 2024
Why used: Supports the routine-based and reassurance prompts.
View sourceCommunication and Alzheimer's
Alzheimer's Association
Alzheimer's Association, 2025
Why used: Informs GrandChatly's core communication principles: adapt to ability, prefer recognition over recall, connect rather than correct, and use familiar and sensory anchors.
View sourceBehavioral Dyadic Interventions in the Context of Alzheimer's Disease and Related Dementia: A Call to Action
Vranceanu et al., including Evan Plys, PhD, ABPP
Journal of Alzheimer's Disease, 2026 · PMID 41212653
Why used: Helps inform GrandChatly's relationship-centered model, where what works depends on both the person and who they're talking with.
View sourceDeveloping a Dyadic Immersive Virtual Environment Technology Intervention for Persons Living With Dementia and Their Caregivers: Multiphasic User-Centered Design Study
Including Evan Plys, PhD, ABPP
Journal of Medical Internet Research, 2025 · PMID 40397932
Why used: Informs GrandChatly's user-centered, dyadic approach to designing technology that supports both members of a caregiving relationship.
View sourcePsychotherapy With a Caregiver of a Spouse With Parkinson's Disease: Integrating Systems and Individual Approaches
Evan Plys and Maggie Syme
Clinical Case Studies, 2025 · PMID 41031045
Why used: Used only for concepts around caregiver experience and the interaction between individual and family-level factors. Not used as evidence for any specific conversation prompt.
View sourcePilot Randomized Controlled Trial of the Feasibility of a Mobile App-Delivered Mindfulness-Based Intervention for Caregiver Stress
Evan Plys and colleagues
The Gerontologist, 2025 · PMID 39921878
Why used: Informs the caregiver-support and technology-delivered intervention context. Not used as evidence that any specific GrandChatly prompt improves dementia communication.
View sourceTips for Communicating with Someone with Dementia
NIH National Institute on Aging
National Institute on Aging, 2025
Why used: Informs GrandChatly's adaptation of question complexity and the preference for recognition and choice-based prompts over precise recall.
View sourceIf something seems like a sudden change
Markedly increased confusion, altered consciousness, new weakness, difficulty breathing, or stroke-like symptoms can be a medical issue. GrandChatly can't diagnose this — please contact a doctor or emergency services if it seems urgent.
GrandChatly provides general conversation and caregiving guidance. It does not diagnose dementia or delirium, determine disease stage, recommend medications, or replace medical care.
