Data Transparency & Access

Wearable Wellness Guide — Research Accountability & Open Science Commitment

Why Data Transparency Matters in Wearable Health Reviews

When health decisions rest on device accuracy claims, you deserve to see the evidence. Wearable Wellness Guide operates on a foundational principle: all testing data, methodology decisions, and analytical processes should be verifiable, reproducible, and accessible to independent researchers, journalists, and healthcare professionals.

This page explains what data we collect during device testing, how we protect participant privacy, how we publish our findings, and how external parties can request access to underlying datasets for validation or further research.

Our commitment: If we test it, we document it. If we publish a claim, the data supports it. If you need verification, we provide access.

Data Collection Standards

What Data We Collect

During each wearable device testing protocol, we systematically record:

Device Measurements

• Heart rate (beats per minute)
• Blood oxygen saturation (SpO₂ percentage)
• Sleep stage classifications (light, deep, REM, awake)
• Step counts and activity classifications
• Blood pressure readings (systolic/diastolic mmHg)
• Respiratory rate (breaths per minute)
• Energy expenditure estimates (calories)
• Other health metrics as relevant to device category

Reference Device Measurements

• Concurrent readings from medical-grade or clinically validated reference equipment
• Device model, firmware version, and calibration status
• Measurement timestamps synchronized to device under test

Environmental & Testing Conditions

• Testing setting (laboratory, real-world environment)
• Activity state (resting, walking, running, HIIT, sleep)
• Ambient temperature and humidity (when relevant)
• Device placement and fit verification notes
• Protocol adherence markers (e.g., movement artifacts, sensor detachment)

Participant Demographics (Anonymized)

• Age category (18-29, 30-44, 45-60, 60+)
• Biological sex
• Fitzpatrick skin tone scale classification (I-VI)
• Body mass index category (underweight, normal, overweight, obese)
• Fitness level category (sedentary, moderately active, highly active)
• Relevant health status categories when applicable to device type (e.g., diagnosed cardiovascular conditions for heart rate monitors)

Testing Metadata

• Test date and duration
• Device firmware version at time of testing
• Protocol version used
• Testing administrator initials (internal quality control)
• Any deviations from standard protocol

What Data We Do NOT Collect

To protect participant privacy and comply with research ethics standards:

❌ Participant names or identifying information
❌ Contact information (email, phone, address)
❌ Medical history beyond categories relevant to device validation
❌ Medical diagnoses or treatment information
❌ Precise location data or GPS coordinates
❌ Photographic or video recordings of participants
❌ Any individually identifiable health information that could re-identify participants when combined with external datasets

Know about our Methodology & protocol details

Data Anonymization Process

Before any data leaves our secure testing environment, we follow a six-stage anonymization protocol:

Stage 1: Participant ID Assignment

Each participant receives a unique, randomly generated alphanumeric identifier (e.g., P-4721) that replaces all personally identifying information in testing records. Assignment logs linking IDs to participants are stored separately under access-controlled encryption.

Stage 2: Direct Identifier Removal

All names, contact information, and precise dates of birth are permanently removed from analytical datasets. Dates are converted to age categories or testing sequence numbers.

Stage 3: Demographic Aggregation

Granular demographic data (e.g., age 34, BMI 27.3) is converted to categorical ranges (e.g., age 30-44, BMI 25-30) to prevent identification through unique characteristic combinations.

Stage 4: Small Cell Suppression

When subgroup analyses produce cell sizes fewer than 5 participants, data is either aggregated into broader categories or marked as “N<5, data withheld for privacy protection.”

Stage 5: Re-Identification Risk Review

Dr. Rishav Das, in consultation with data management protocols, reviews datasets for potential re-identification vectors, including:

• Unique combinations of demographic characteristics
• Outlier measurements that could identify individuals
• Temporal patterns that could link to external event data

Stage 6: Publication Approval

Only datasets passing all five prior stages receive approval for public release or external access requests.

Data Storage & Security

Storage Infrastructure

Raw testing data is stored in encrypted formats on access-controlled servers. Current security measures include:

• Encryption: AES-256 encryption for data at rest; TLS 1.3 for data in transit
• Access Control: Role-based access limited to Dr. Rishav Das and authorized research personnel
• Authentication: Multi-factor authentication required for all data system access
• Audit Logging: All data access events are logged with timestamp, user identity, and action taken

Retention Policy

Testing data is retained for a minimum of 7 years from publication date to enable:

• Long-term reproducibility verification
• Meta-analyses across device generations
• Regulatory inquiries if device safety issues emerge

After the retention period, data is either:

• Archived in anonymized form indefinitely for historical research value, or
• Securely disposed via cryptographic erasure and physical media destruction

Security Audits

Data security practices are reviewed annually. While Wearable Wellness Guide is independently operated and does not currently undergo third-party security audits, we follow industry-standard data protection frameworks appropriate for non-clinical research settings.

Breach Notification Commitment

In the event of unauthorized data access or security breach, we commit to:

• Immediate containment and forensic assessment
• Notification to affected participants if identifiable data is compromised
• Public disclosure on this website within 72 hours of confirmed breach
• Corrective actions documented and implemented

To date, no security incidents affecting participant data have occurred.

Published Datasets

What We Publish

When sufficient sample sizes are achieved and anonymization protocols are complete, Wearable Wellness Guide publishes:

Aggregated Accuracy Metrics by Device

• Mean error and standard deviation
• Mean absolute percentage error (MAPE)
• Correlation coefficients vs. reference devices
• Limits of agreement (Bland-Altman analysis where applicable)

Error Distributions & Statistical Summaries

• Histograms of measurement error across conditions
• Percentile distributions (5th, 25th, 50th, 75th, 95th)
• Outlier frequency and classification

Demographic Subgroup Analyses (when sample sizes permit)

• Accuracy breakdowns by skin tone category
• Performance differences by age group
• Effect of BMI on sensor accuracy
• Activity-specific error rates

Testing Condition Summaries

• Environmental parameters during testing
• Protocol adherence rates
• Device firmware versions tested

Anonymized Raw Data (when feasible)

• CSV or Excel format files containing:
  • Participant ID (anonymized)
  • Timestamp
  • Device reading
  • Reference device reading
  • Activity/condition category
  • Demographic category codes

Where We Publish

Datasets are made available through:

1. Open Science Repositories
   
   • Figshare, Zenodo, or Open Science Framework (OSF)
   • Each dataset receives a persistent DOI (Digital Object Identifier) for permanent citation
   • Datasets tagged with appropriate metadata for discoverability
2. Direct Website Download
   
   • Machine-readable formats (CSV, JSON)
   • Accompanied by data dictionaries explaining variable codes
   • Version-controlled with publication date stamps
3. Supplementary Materials to Published Reviews
   
   • Summary datasets embedded in device review articles
   • Full datasets linked from review pages

Current Published Datasets

As testing protocols scale and sample sizes reach publication thresholds, datasets will be listed here with:

• Dataset title
• Number of participants
• Number of devices tested
• Testing period
• DOI or download link
• Brief description

Upcoming Dataset Releases (in development as testing continues):

• Heart rate accuracy validation across Fitzpatrick skin tones (N=24+ target)
• Sleep staging algorithm performance vs. polysomnography references (N=30+ nights target)
• Consumer pulse oximeter accuracy during exercise (N=20+ participants target)

Current status: Testing protocols are operational. Initial datasets are being compiled and will be published when sample sizes meet statistical adequacy thresholds (N≥20 for most device categories).

Data Access Requests

Who Can Request Data

Wearable Wellness Guide welcomes data access requests from:

✅ Academic researchers conducting device validation or meta-analysis studies
✅ Independent validation teams seeking to replicate or extend our findings
✅ Journalists and fact-checkers verifying claims for investigative reporting
✅ Regulatory agencies or public health organizations assessing device safety
✅ Healthcare professionals advising patients on device selection
✅ Device manufacturers seeking to understand performance gaps (subject to non-commercial use terms)

Request Process

Step 1: Submit Formal Request
Email: data@wearablewellnessguide.com (or primary contact email if not yet established)

Include in your request:

• Your name and professional affiliation
• Specific dataset(s) or data categories requested
• Research question or intended use
• Proposed analysis plan (brief description)
• Timeline for use and publication plans

Step 2: Institutional Review (if applicable)
If your research involves human subjects or will be published in peer-reviewed journals, provide:

• IRB approval documentation, or
• Confirmation that your use qualifies for exemption under your institution’s policies

Step 3: Data Use Agreement Execution
Upon preliminary approval, you will receive a standard Data Use Agreement (DUA) outlining:

• Permitted uses
• Re-identification prohibition
• Attribution requirements
• Data security obligations
• Sharing restrictions

Step 4: Decision Notification
Requests are reviewed within 30 calendar days. You will receive:

• Approval with data access instructions, or
• Request for additional information, or
• Denial with explanation (rare; typically only when privacy risks exist)

Step 5: Secure Data Transfer
Approved requesters receive data via:

• Encrypted file transfer (for sensitive datasets)
• Direct repository download link (for published datasets)
• Password-protected archive with data dictionary

Data Use Terms

All data requesters agree to:

1. No Re-Identification Attempts
   Do not attempt to identify individual participants through dataset analysis, linkage to external data sources, or any other method.
   
2. Non-Commercial Use (unless explicitly approved)
   Data provided for research, journalism, or validation purposes may not be used for commercial product development, marketing, or competitive intelligence without separate written agreement.
   
3. Proper Attribution
   Cite Wearable Wellness Guide and Dr. Rishav Das in any publications, reports, or presentations using our data. Recommended citation format:
   
   
   Wearable Wellness Guide. (Year). [Dataset Title]. Retrieved from [DOI or URL]. Medical review by Rishav Das, M.B.B.S.
   
   
4. Results Sharing
   If your analysis produces findings relevant to device accuracy or safety, we request (but do not require) that you share a copy of your results with us before publication for potential inclusion in updated reviews.
   
5. Compliance with Participant Privacy Protections
   Maintain all data security measures outlined in the DUA. Do not redistribute raw data without explicit written permission.
   

Denied Requests

Requests may be denied if they:

• Pose unacceptable re-identification risks
• Lack sufficient methodological detail for us to assess appropriateness
• Request data types we do not collect
• Involve commercial uses that could create conflicts of interest

Denied requesters are provided with an explanation and, when possible, alternative data sources or modified access options.

Participant Privacy Protections

Core Privacy Principles

Principle 1: Minimal Data Collection
We collect only data necessary for device accuracy evaluation. Extraneous personal information is never requested or recorded.

Principle 2: No Individual-Level Public Release
Individual participant data is never published in any form that could enable re-identification. All public datasets are aggregated or thoroughly anonymized.

Principle 3: Consent-Based Data Sharing
Participants provide informed consent for their anonymized data to be used in research publications and shared with external researchers under controlled conditions. No data sharing occurs without this consent.

Principle 4: Purpose Limitation
Data collected for device testing is used exclusively for accuracy validation and related research purposes. It is never sold, traded, or used for unrelated commercial activities.

Principle 5: Right to Withdraw
Participants may withdraw from testing at any time and request removal of their data from future analyses (though aggregated data already published cannot be retroactively removed).

Aggregate-Only Public Datasets

All publicly available datasets contain only:

• Summary statistics (means, medians, standard deviations)
• Anonymized categorical demographic breakdowns
• Device vs. reference comparison tables
• Error distribution graphs without individual data points visible

Individual-level datasets (where each row represents one participant or one measurement) are:

• Never published publicly
• Shared only under Data Use Agreements with approved requesters
• Subject to enhanced anonymization protocols before release

Small Cell Suppression in Published Tables

When subgroup analyses produce categories with fewer than 5 participants, we:

• Suppress specific numeric values and report “N<5”
• Combine categories to achieve adequate cell sizes (e.g., “Ages 18-29 and 30-44 combined”)
• Note in methodology sections where suppression has occurred

Example:

Skin Tone Category	N	Mean MAPE (%)
Fitzpatrick I-II	8	3.2 ± 1.1
Fitzpatrick III-IV	12	3.7 ± 1.4
Fitzpatrick V-VI	<5	Data withheld

Note: Fitzpatrick V-VI data withheld to protect participant privacy due to small sample size.

No External Dataset Linkage

We prohibit:

• Linking our anonymized datasets to external databases (e.g., commercial health records, social media profiles)
• Using unique measurement combinations to reverse-identify participants
• Combining multiple published datasets in ways that could narrow identification pools

Data requesters who violate these prohibitions face immediate DUA termination and potential legal action if harm occurs.

IRB-Equivalent Review for Sensitive Data

While Wearable Wellness Guide operates as an independent consumer review entity (not a formal research institution), we apply Institutional Review Board (IRB)-equivalent ethical standards when:

• Testing involves participants with diagnosed medical conditions
• Measurements could reveal sensitive health information (e.g., arrhythmias, sleep disorders)
• Datasets include vulnerable populations (e.g., pregnant individuals, elderly participants)

In these cases, Dr. Rishav Das conducts an ethics review assessing:

• Risk-benefit balance for participants
• Adequacy of informed consent
• Privacy protection sufficiency
• Special protections for vulnerable groups

If ethical concerns cannot be adequately mitigated, testing protocols are modified or discontinued.

Data Quality Assurance

Pre-Publication Data Validation

Before any dataset is published or shared, it undergoes:

1. Completeness Check

• Verify all expected variables are present
• Confirm no critical data fields are missing
• Document any protocol deviations or missing data with explanations

2. Range and Logic Validation

• Identify physiologically implausible values (e.g., heart rate 300 bpm)
• Check for impossible value combinations (e.g., sleep stage “REM” during “running” activity)
• Flag outliers for secondary review

3. Anonymization Verification

• Confirm no personally identifiable information remains
• Re-check small cell sizes (<5) for suppression
• Validate that demographic categories are sufficiently broad

4. Reference Alignment

• Ensure device timestamps match reference device timestamps
• Verify reference equipment calibration records are documented
• Confirm measurement units are consistent (e.g., all HR in bpm, not some in Hz)

5. Statistical Summary Accuracy

• Recalculate all means, standard deviations, and error metrics
• Verify correlation coefficients and statistical tests
• Cross-check summary tables against raw data

Statistical Review of Published Summaries

Dr. Rishav Das reviews all published statistical summaries for:

• Appropriate statistical test selection (e.g., paired t-tests for repeated measures)
• Correct interpretation of p-values and confidence intervals
• Adequate sample size justification
• Transparent reporting of limitations

If statistical methods require specialized expertise beyond general medical training, external statistical consultation is sought and acknowledged.

Post-Publication Error Correction Process

If errors are discovered after publication:

Minor Errors (typos, formatting issues, non-substantive corrections):

• Corrected within 48 hours
• Correction note added to dataset page: “Updated [Date]: [Brief description of change]”
• Original version archived with “v1.0” tag; corrected version tagged “v1.1”

Major Errors (calculation mistakes, data processing errors affecting conclusions):

• Dataset immediately flagged as “Under Review”
• Correction analysis completed within 7 days
• If conclusions change: Full correction notice published explaining error, impact, and revised findings
• If conclusions unchanged: Transparent note explaining error and why results remain valid
• All users who downloaded affected datasets notified via email (if contact info available)

Retractions (rare; only if data integrity cannot be verified):

• Dataset removed from public access
• Retraction notice posted permanently with explanation
• All citing publications notified
• Affected device reviews updated or retracted as appropriate

Versioning and Correction Notices

All published datasets include:

• Version number (e.g., v1.0, v1.1, v2.0)
• Publication date and last update date
• Change log documenting all modifications
• Clear statement if dataset supersedes previous version

Example change log entry:

Version 1.1 (Updated March 15, 2026):

– Corrected mean MAPE calculation for Device C (was 4.8%, now 4.2%)

– Added 3 additional participants to Fitzpatrick V-VI category

– No change to overall conclusions regarding device ranking

---

Commitment to Open Science

Wearable Wellness Guide embraces open science principles:

✅ Reproducibility: Full testing protocols published; raw data shared upon request
✅ Transparency: Methodology limitations and funding sources openly disclosed
✅ Accessibility: Data provided free of charge for non-commercial research
✅ Collaboration: External validation efforts encouraged and supported
✅ Continuous Improvement: Protocols updated based on peer feedback and emerging standards

Our accountability standard: If we cannot justify a conclusion with verifiable data, we do not publish it. If you cannot replicate our findings with our shared methodology, we want to know why.

Frequently Asked Questions

Contact for Data Access Inquiries

For dataset access requests, privacy questions, or data security concerns:

📧 Email: data@wearablewellnessguide.com
🌐 Data Repository: (Link to be established upon first dataset publication)
📄 Sample Data Use Agreement: [Available upon request]

For general questions about testing methodology:

See our Medical Review Process and Testing Methodology sections.

About our data & transparency commitments

---

Updates to This Policy

This Data Transparency & Access policy is reviewed annually and updated as:

• Data security standards evolve
• New data sharing platforms become available
• Participant privacy regulations change
• External feedback identifies areas for improvement

Last reviewed: January 11, 2026
Next scheduled review: January 2027
Policy version: 1.0

All substantive changes will be documented in a change log below and announced via our primary communication channels.

About our data & transparency commitments

---

This page serves as the canonical authority for all data handling, privacy protection, and research transparency claims made by Wearable Wellness Guide. All testing protocols, dataset publications, and external research collaborations must comply with the standards outlined here.