ABSTRACT

Voice Up is a purpose-centered wellbeing and talent platform designed to connect measurement, lived experience, and real-world engagement pathways. This report documents the current evidence base supporting Voice Up, including psychometric validation of original instruments, large-scale qualitative analysis, behavioral engagement data, and market-level recruitment efficiency. Together, these data form a multi-tier evidence system demonstrating reliable measurement, construct convergence, behavioral relevance, and scalable demand. This report intentionally does not claim causal impact and instead defines the validated baseline from which longitudinal and experimental evaluation will proceed.

1. INTRODUCTION

1.1 Background and Rationale

Across education, workforce development, and health-adjacent systems, individuals experience increasing stress, lack of purpose, and disconnection between credentials and meaningful contribution. Traditional evaluation frameworks emphasize outputs (completion, placement) without measuring the human conditions that drive engagement and persistence.

Voice Up was designed to address this gap by:

• Measuring whole-person wellbeing and purpose
• Validating those measures empirically
• Aligning pathways (training, internships, incubator projects, mentorship) to measured needs
• Evaluating engagement behavior and market demand

1.2 Purpose of This Report

This report documents:

1. What Voice Up measures
2. How those measures were validated
3. How quantitative and qualitative evidence converge
4. How measured constructs relate to observed behavior
5. What claims are supported today and what are explicitly deferred

2. DATA SOURCES AND OVERVIEW

2.1 Quantitative Datasets

Three original instruments were developed and deployed:

2.2 Qualitative Dataset

• 6,700+ participant responses
• Coded using a hierarchical codebook
• 18 primary themes with defined inclusion/exclusion rules
• Frequency and co-occurrence tables generated

2.3 Behavioral & Engagement Datasets

• Training onboarding
• Internship and incubator interest forms
• Mentor/coach participation files
• Pathway selection indicators

2.4 Market Validation Dataset

• Dual recruitment campaigns (U.S. and India)
• Applicant volume, spend, conversion, and pipeline quality
• Used strictly as market validation, not outcome evidence

3. METHODS — QUANTITATIVE ANALYSIS

3.1 Reliability Analysis

Internal consistency was assessed using Cronbach's alpha at the scale and domain level.

• Student Wellbeing: α ≈ .93
• Wellbeing Self: α ≈ .90
• Head Coach/Mentor: α ≈ .98 (interpreted cautiously due to sample size)

3.2 Dimensionality and Factor Structure

Exploratory factor analysis (EFA) and parallel analysis were used to evaluate dimensionality. Confirmatory factor analysis (CFA) was conducted on short forms where sample size permitted.

Key findings:

• Multidimensional structures were supported
• Short forms retained strong fit and reliability
• No evidence of single-factor collapse

3.3 Measurement Invariance (Readiness & Execution)

Group variables (country, education level, role) were integrated at the response level. Invariance testing followed standard procedures:

• Configural
• Metric
• Scalar (or partial scalar where required)

This enables subgroup comparison while preserving interpretive boundaries.

4. METHODS — QUALITATIVE ANALYSIS

4.1 Coding Framework

A formal hierarchical codebook was developed with:

• Theme definitions
• Inclusion/exclusion rules
• Cross-coder audit logic

4.2 Analytical Procedures

• Theme frequency analysis
• Theme co-occurrence analysis
• Cross-dataset comparison (student, training, internship, mentor contexts)

4.3 Auditability

All coded responses are traceable from raw text to theme assignment, supporting replication and external review.

5. RESULTS — TIER 1: PSYCHOMETRIC FINDINGS

5.1 Reliability and Scale Performance

All primary instruments demonstrated strong internal consistency. Item-level diagnostics indicated appropriate discrimination with minimal redundancy.

5.2 Factor Structure Summary

• Student Wellbeing: stable multidomain structure
• Wellbeing Self: two correlated factors (balance; future orientation)
• Short forms demonstrated strong CFA fit

5.3 Subgroup Readiness

Group-level analyses indicate sufficient structure to support subgroup comparisons where Ns permit.

6. RESULTS — TIER 2: QUANTITATIVE ↔ QUALITATIVE CONVERGENCE

6.1 Theme–Domain Alignment

Strong convergence was observed between measured domains and participant narratives:

6.2 Co-Occurrence Patterns

Themes related to stress, support, and purpose frequently co-occurred, indicating systemic rather than isolated experiences.

6.3 Interpretation

These findings support construct validity and demonstrate that instruments capture lived experience rather than imposed categories.

7. RESULTS — TIER 3: MEASUREMENT ↔ BEHAVIOR

7.1 Behavioral Outcomes Defined

Behavioral indicators included:

• Application and onboarding
• Pathway selection (training, internship, incubator)
• Mentorship interest

7.2 Observed Associations

Participants with:

• Higher future orientation engaged more frequently in incubator and internship pathways
• Lower balance scores showed higher engagement with training and support pathways
• Higher connection needs aligned with mentorship interest

7.3 Interpretation

Measured constructs are behaviorally meaningful and inform pathway fit. These results demonstrate relevance, not causality.

8. RESULTS — TIER 4: MARKET VALIDATION

8.1 Recruitment Scale and Efficiency

Across two campaigns:

• 1,400 applicants
• ~$150 total spend
• ~$0.10 blended cost per applicant
• Strong conversion rates

8.2 Pipeline Quality

Applicants represented:

• Over 130 universities
• Advanced degrees
• Public service and health professions
• Cross-geographic participation

8.3 Interpretation

These results demonstrate demand, scalability, and economic viability. They do not measure wellbeing outcomes.

9. SYNTHESIS AND INTERPRETATION

Across all tiers, evidence converges on four conclusions:

1. Voice Up measures real, reliable constructs
2. Those constructs align with lived experience
3. They are associated with real engagement behavior
4. The model scales efficiently across contexts

This establishes a validated baseline for outcome evaluation.

10. LIMITATIONS

• No causal claims are made
• Longitudinal data collection is in progress
• Mentor sample size limits factor modeling
• Behavioral linkage is observational

These limitations are intentional and define the next evaluation phase.

11. NEXT-PHASE EVALUATION PLAN

11.1 Short-Term (0–6 months)

• Pre/post pilot evaluations
• Expanded subgroup analysis
• Research partnerships

11.2 Mid-Term (6–12 months)

• Longitudinal tracking
• Comparative pathway analysis
• Early causal designs where feasible

11.3 Long-Term

• Scaled deployment with embedded evaluation
• Peer-reviewed publications
• Policy-relevant reporting

12. CONCLUSION

Voice Up has established a rare foundation: validated measurement, lived-experience grounding, behavioral relevance, and market demand. The next phase is not conceptual validation but responsible impact evaluation.

APPENDICES (AVAILABLE)

• Instruments
• Reliability tables
• Invariance tables
• Codebook
• Theme frequencies and co-occurrence
• Recruitment campaign exhibit