ΨMCA 2026

Pre-Conference Workgroup Meeting
Workgroup 1: Validation of Classification

Dan Mungas
Jingxuan Wang

Aug 1, 2026

Agenda

01
Introductions
02
Structure and Roles
03
Conference Navigation
04
Readings and Preparation
05
Software Setup
06
Questions and Next Steps
🎥 This meeting will be recorded and shared with participants who are unable to attend.

Introductions

WELCOME TO
ΨMCA 2026
Advanced Psychometric Methods for Cognitive Aging Research
This group brings together participants at all career stages - from graduate students to senior investigators.
Every person here has something valuable to contribute.
We are glad you are here.

Let's get to know each other

Please briefly share:
01
Your name and institution
02
Your current role or career stage
03
Your research or methodological interests
04
One thing you hope to gain from ΨMCA

About Our Workgroup

WORKGROUP 1
Validation of Classification
How well do clinical and algorithmic classifications reflect underlying disease and predict future cognitive decline?
01
Why this matters
Clinical diagnosis is widely used to stage cognitive impairment and brain injury or degeneration, and to predict future decline.
02
The current gap
Algorithmic classifications are often evaluated against clinical diagnosis as the "gold standard," but both approaches are rarely compared directly against external validation criteria.
03
Our goal
Validate clinical diagnosis, algorithmic classification, and machine learning predictions against disease biomarkers and future progression of cognitive impairment.
Classification approaches we will compare
We will validate
clinical, algorithmic classification, and machine learning-based predictions
against
longitudinal MRI
other disease biomarkers
future cognitive progression
DATASET UC Davis Alzheimer's Disease Research Center

Potential Project Ideas

Multiple options to support subgroups with varying interests and skill levels

A
Example project A:
Compare the predictive validity of three dementia classification approaches:
  • Clinical diagnosis
  • Decision rule-based classification
  • Machine learning-based classification

using longitudinal MRI measures as the external validation criterion.
Methods Mixed effects longitudinal models
B
Example project B:
Compare the predictive validity of two dementia classification approaches:
  • Clinical diagnosis
  • Decision rule-based classification

for future cognitive decline.

* Machine learning-based classification used cognitive data as predictors.
Methods Mixed effects longitudinal models
→
These are example project ideas. We encourage participants to propose additional ideas, and we will finalize projects together on Day 1 in Tahoe.
We will match participants' interests, experience, and preferred methods to the available projects.

Structure and Roles

WORKGROUP CO-LEADS
How we will support the group
Dan Mungas
Dataset preparation, UC Davis ADRC expertise, and guidance on cognitive classification and measurement.
Jingxuan Wang
Project coordination, analysis planning, methodological discussion, and subgroup support.
Shared responsibilities
Facilitate discussion, clarify tasks, monitor inclusion, troubleshoot barriers, and help move each project toward a concrete product.
01
Walk through the dataset
Become familiar with the UC Davis ADRC data, available variables, biomarkers, and documentation.
02
Discuss project directions
Review the example projects, brainstorm additional ideas, and identify the most promising research questions.
03
Choose a project subgroup
Participants will join a project based on scientific interests, methodological experience, and learning goals.
04
Define individual responsibilities
Assign clear roles such as literature review, coding, analysis, visualization, interpretation, or writing.
05
Collaborate throughout the week
Work closely with teammates, share ideas, review code, and support each other's progress.
06
Daily check-ins
Set daily goals, review progress, troubleshoot barriers, and adjust tasks as needed.
◎
Our guiding principle
Everyone should have a meaningful opportunity to contribute, and, where feasible, everyone should work directly with the data.

Conference Navigation: Goals, Expectations & Culture

Our goals
What we hope to accomplish
01
Conduct meaningful, publishable research.
02
Practice collaborative team science.
03
Give everyone an opportunity to interact directly with the data.
04
Lay the groundwork for a manuscript or other concrete research product.
From the co-leads
What to expect from us
✓
Clear tasks and division of responsibilities.
✓
Daily check-ins, goal-setting, and progress reviews.
✓
Transparent and empathetic communication.
✓
Recognition of individual contributions.
From participants
What we ask of you
✓
Arrive prepared with readings and software completed.
✓
Engage actively, ask questions, and share ideas.
✓
Collaborate with people you do not already know.
✓
Communicate early when questions or concerns arise.

Welcome, First-Time Attendees!

📅
Dates and Location
August 9-14, 2026, in Tahoe. Most of the week is spent in focused workgroup sessions.
🤝
Intensive Collaboration
You will work closely with the same group throughout the week. Deep collaboration takes time - this is intentional.
💡
Questions Are Welcome
Junior and senior attendees work side by side. Questions and methodological discussions are encouraged.
🍽️
Meals and Social Time
Informal conversations are an important part of the experience. Try to connect with new people throughout the week.
Helpful resources
Orientation
Psy-MCA for Beginners
Schedule
2026 Conference Schedule
Workgroups
2026 Workgroup Descriptions
Questions before Tahoe?
Contact Dan Mungas or Jingxuan Wang at any time.

Join the Psy-MCA 2026 Community

Stay connected
Join the Discord server
Connect with other attendees before Tahoe, introduce yourself, ask questions, and share useful resources.
discord.gg/Zwn9MJAmY
Join the Workgroup 1 channel so you can connect with your teammates and receive updates before the conference.
Orientation
Psy-MCA for Beginners
psymca.org/ψmca-for-beginners
Conference planning
2026 Conference Schedule
psymca.org/2026-conference/schedule-2026
Scientific program
2026 Workgroup Descriptions
psymca.org/workgroups-2026
📹 Recording: This pre-conference meeting is being recorded and will be shared with participants who are unable to attend.

Authorship - Let’s Talk About It Early

We will revisit authorship on Day 1 in Tahoe. These are guiding principles rather than fixed mandates.

01
Start the conversation early
Discuss potential roles, products, and authorship expectations when projects and subgroups are established.
02
Encourage junior leadership
Junior members should be supported in considering first-author roles, while senior members may contribute in senior or supporting positions.
03
The first author leads
First authorship involves substantial responsibility for coordination, drafting, revisions, and moving the project toward completion.
04
Authorship requires contribution
Contributions may include analysis, coding, interpretation, writing, visualization, or substantive review. Membership alone is not sufficient.
05
Order reflects contribution
Authorship order should reflect relative substantive contributions, not career stage or seniority.
✦
Each workgroup may adapt these norms. We will aim for transparency, fairness, and ongoing discussion as contributions evolve.

Readings and Preparation

Core concepts
Machine learning-based classification
Mungas et al., 2025
Machine learning diagnosis of cognitive impairment and dementia in harmonized older adult cohorts
[link]
Analytic preparation
Classification validation
Mungas et al. (preprint), 2026
Validation of clinical diagnosis and machine learning classification of cognitive impairment
[Will be shared]
Data preparation
Dataset Documents
UC Davis ADRC Codebook
Variables, coding conventions, and study documentation
[Will be shared]
⏱
Estimated reading time: approximately 2 hours.

Software Setup

✓
Please arrive with your statistical software installed and ready to use.
Test your installation, required packages, and access before traveling to Tahoe.
Quick poll - what software do you use?
R / RStudio
Stata
Mplus
Python
SAS
Other
Resources
What we will have
▣
Documented datasets
Analysis files with codebooks, variable definitions, and supporting documentation.
☷
Analysis planning tools
A shared document for variables, task division, and analysis planning.
Need help? Contact us before the conference if you have installation, package, access, or compatibility problems.

Questions & Discussion

Open discussion
Questions?
Please share any questions, concerns, project ideas, or suggestions before we begin our work together in Tahoe.
Topics we can discuss
Scientific questions and project ideas
Dataset access and preparation
Software or technical needs
Conference logistics
Roles, collaboration, and authorship
?
Stay in touch
Contact the co-leads
Dan Mungas
Workgroup Co-Lead
Jingxuan Wang
Workgroup Co-Lead
Thank you for joining us.
We look forward to a productive, inclusive, and collaborative week at ΨMCA 2026.