Sources & Bibliography

Research · 36 studies

Research informing the model

Evidence labels describe research maturity, not a grade. External research informs JRX practice; it does not establish JRX outcomes.

• Review of randomized evaluations of tutoring programs

  Strong evidence

  J-PAL (Abdul Latif Jameel Poverty Action Lab) · 2020s review · Research review · 96 randomized evaluations

  Research consistently finds that structured, frequent, high-quality tutoring produces meaningful learning gains.

  Informs: Academic Specialists

  Limitations: Effects vary substantially by tutor type, dosage, grade level, subject, and program structure. The review covers structured tutoring programs, not informal help.

  What we don’t claim: Do not claim that any form of one-to-one academic assistance produces the same results, or that JRX has replicated these effects.

  Open J-PAL (Abdul Latif Jameel Poverty Action Lab)Verification pending

• Continued Progress: Promising Evidence on Personalized Learning

  Promising evidence

  RAND Corporation · 2015 · Quasi-experimental study · Approximately 62 schools and 11,000 students

  Evidence on personalized learning is promising rather than settled: implementation quality appears to matter more than the label.

  Informs: Personalized learning plans and hybrid models

  Limitations: Non-randomized comparisons, wide variation between schools, and 'personalized learning' covers very different practices.

  What we don’t claim: Do not present personalized learning as causally proven or universally effective.

  Open RAND CorporationVerification pending

• School connectedness and adolescent health and well-being

  Strong evidence

  U.S. Centers for Disease Control and Prevention · 2021 survey analysis and current guidance · Observational study

  Stronger school connectedness is consistently associated with better attendance, achievement, and well-being. Association, not proof of causation.

  Informs: Success Specialists

  Limitations: Cross-sectional and self-reported. Students with high connectedness may differ from peers in ways the analysis cannot fully separate.

  What we don’t claim: Do not claim that a connectedness program causes reductions in poor mental health, and never imply school staff provide clinical treatment.

  Open U.S. Centers for Disease Control and PreventionVerification pending

• Randomized evaluation of project-based Advanced Placement courses

  Strong evidence

  Lucas Education Research and university research partners · 2021 · Randomized evaluation

  Rigorous project-based learning can produce meaningful academic outcomes when projects are intentionally designed around academic content and high expectations.

  Informs: Seven Quests — applied, project-based learning

  Limitations: Two specific AP courses with purpose-built curricula and teacher training. Context-specific.

  What we don’t claim: Do not claim all project-based learning raises achievement by eight percentage points.

  Open Lucas Education Research and university research partnersVerification pending

• Career Academies: Long-Term Impacts on Labor Market Outcomes

  Strong evidence

  MDRC · 2008 · Randomized evaluation

  A randomized long-term study found that career-connected high school programs improved labor market outcomes years after graduation.

  Informs: School as Pathway — career-connected learning

  Limitations: Older study, specific program model and sites; effects concentrated among young men. Labor markets have changed since the study period.

  What we don’t claim: Do not present study-year dollar figures as today's dollars or promise earnings outcomes to JRX families.

  Open MDRCVerification pending

• What Works Clearinghouse reviews of dual enrollment and CTE-focused programs

  Promising evidence

  Institute of Education Sciences / What Works Clearinghouse · Current reviews · Research review

  Evidence indicates stronger postsecondary participation among students involved in well-structured dual enrollment and CTE-focused experiences.

  Informs: Dual enrollment, credentials, and trades pathways

  Limitations: Program models differ widely; some reviewed studies are quasi-experimental, and selection into dual enrollment complicates interpretation.

  What we don’t claim: Do not claim dual enrollment causes college completion for every participant.

  Open Institute of Education Sciences / What Works ClearinghouseVerification pending

• Meta-analysis of service-learning program outcomes

  Promising evidence

  Peer-reviewed education research · 2011 and subsequent syntheses · Meta-analysis

  Service learning is associated with positive effects across several domains, with the strongest results in programs that connect service to curriculum and include structured reflection.

  Informs: Leadership Quest, AEIOU, and service

  Limitations: Wide variation in program quality; stronger effects appear where programs include curriculum links, student voice, community involvement, and structured reflection.

  What we don’t claim: Do not claim that any service activity produces academic or civic gains.

  Open Peer-reviewed education researchVerification pending

• Systematic reviews of youth participatory action research outcomes

  Promising evidence

  Peer-reviewed youth development research · Recent reviews · Research review

  Research documents agency, leadership, academic, career, and civic outcomes associated with genuine youth voice, while causal evidence remains limited.

  Informs: YouthLed Community Solutions and student voice

  Limitations: Largely qualitative and small-sample studies. Causal evidence is far less developed than in tutoring or career-academy research.

  What we don’t claim: Do not present youth voice programs as causally proven achievement interventions.

  Open Peer-reviewed youth development researchVerification pending

• Physical Activity Guidelines for Americans

  Strong evidence

  U.S. Department of Health and Human Services / CDC · Current edition · Federal guidance

  Federal guidance recommends 60 or more minutes of daily physical activity, and movement is associated with benefits for health, cognition, and emotional well-being.

  Informs: Athletics Quest and daily movement

  Limitations: Guidance synthesizes health evidence; it is not a study of school athletics programs or academic outcomes.

  What we don’t claim: Do not claim athletics automatically causes higher grades.

  Open U.S. Department of Health and Human Services / CDCVerification pending

• Impact Evaluation of the U.S. Department of Education's Student Mentoring Program

  Promising evidence

  Institute of Education Sciences · 2009 · Randomized evaluation

  A major federal randomized evaluation found no significant overall impacts of school-based mentoring, which is why JRX treats mentoring quality — screening, training, matching, consistency, boundaries, duration, and monitoring — as the design problem.

  Informs: Good Souls Mentors

  Limitations: Evaluated a specific federal grant program with variable implementation, match duration, and mentor training. Later research finds effects concentrated in well-implemented, longer-duration matches.

  What we don’t claim: Do not claim mentoring automatically improves academic outcomes, and do not hide the mixed evidence.

  Open Institute of Education SciencesVerification pending

• Research syntheses on family engagement and school communication

  Promising evidence

  Institute of Education Sciences and university research centers · Current syntheses · Research review

  Two-way, accessible family–school communication is a supported practice associated with stronger attendance, engagement, and connectedness.

  Informs: Parents and families as partners

  Limitations: Engagement practices are difficult to isolate experimentally, and effects depend heavily on accessibility and cultural responsiveness.

  What we don’t claim: Do not frame limited engagement as a lack of family care; access, language, scheduling, and trust are design responsibilities of the school.

  Open Institute of Education Sciences and university research centersVerification pending

• Emerging research on AI-assisted instruction and tutoring tools

  Emerging evidence

  Multiple universities and research organizations · Ongoing · Research review

  JRX uses technology and AI as tools inside a human-led learning environment and will evaluate their effects on learning, agency, creativity, critical thinking, integrity, and digital literacy.

  Informs: Technology and AI-enhanced learning

  Limitations: Rapidly changing tools, short study horizons, few rigorous replications, and unresolved questions about academic integrity and over-reliance.

  What we don’t claim: Do not claim AI dramatically raises achievement, replaces teachers, or guarantees personalization.

  Open Multiple universities and research organizationsVerification pending

• Randomized controlled trial of the Big Brothers Big Sisters of America community-based mentoring program

  Strong evidence

  DuBois, Herrera, Rivera, Brechling & Root · Annals of the New York Academy of Sciences · 2026 · Randomized evaluation · 1,353 youth followed for four years

  A four-year randomized trial of the Big Brothers Big Sisters of America community-based mentoring program found significantly lower violence-related delinquent behavior and recurring substance use, plus favorable effects on selected additional measures. Strong evidence for this specific evaluated program.

  Informs: Good Souls Mentors — community-based mentoring design

  Limitations: Findings belong to this specific, well-implemented and well-supported program. Not all measured outcomes reached statistical significance, and effects should not be generalized to mentoring programs with different screening, training, matching, duration, or support.

  What we don’t claim: Do not imply all mentoring programs reproduce these outcomes, and never present this research as evidence that Good Souls Mentoring works. It informs the Good Souls design; JRX must generate its own evidence.

  Open Annals of the New York Academy of SciencesVerification pending

• The effects of youth mentoring programs: a meta-analysis of outcome studies

  Promising evidence

  Raposa et al. · Journal of Youth and Adolescence · 2019 · Meta-analysis · 70 studies · 25,286 young people

  Across 70 studies and 25,286 young people, mentoring shows a statistically significant positive overall effect of moderate magnitude by youth-prevention benchmarks — and the researchers ask readers to remain realistic about modest impacts as programs are currently implemented.

  Informs: Good Souls Mentors — realistic expectations

  Limitations: Pooled across very different programs, populations, and implementation quality. The authors themselves urge realism about the modest impact of mentoring programs as currently implemented.

  What we don’t claim: Do not quote the moderate effect without the researchers' own caution, and do not claim mentoring produces large or guaranteed outcomes.

  Open Journal of Youth and AdolescenceVerification pending

• Does natural mentoring matter? A multilevel meta-analysis on the association between natural mentoring and youth outcomes

  Promising evidence

  Van Dam et al. · American Journal of Community Psychology · 2018 · Meta-analysis · 30 studies

  Having a natural mentor is associated with positive youth outcomes, and the quality of that relationship is associated roughly twice as strongly. Associational evidence — not proof of causation.

  Informs: Good Souls Mentors — relationship quality over match count

  Limitations: Associational evidence from observational studies. It does not establish that a mentoring relationship causes these outcomes, and youth who form high-quality relationships may differ in other ways.

  What we don’t claim: Do not present these correlations as causal effects, and do not use them to claim any particular program improves outcomes.

  Open American Journal of Community PsychologyVerification pending

• Randomized evaluation of school-based mentoring (BBBS school-based program)

  Promising evidence

  Herrera et al. · Child Development / Public/Private Ventures · 2011 · Randomized evaluation · 1,139 students

  School-based mentoring produced modest academic benefits and stronger perceptions of scholastic competence, while several effects faded and many outcomes showed no significant difference. Mentoring is not magic; design matters.

  Informs: Good Souls Mentors — structure and program design

  Limitations: School-based matches were typically shorter and less intensive than community-based matches. Effects varied with match length and implementation.

  What we don’t claim: Do not present school-based mentoring as reliably raising achievement, and do not hide the null and fading results.

  Open Child Development / Public/Private VenturesVerification pending

• The test of time in school-based mentoring: relationship duration and rematching

  Promising evidence

  Grossman et al. · American Journal of Community Psychology · 2012 · Randomized evaluation · 1,139 youth

  Continuity appears to be part of the intervention: intact matches showed academic gains while premature endings showed none, and rematching was not a neutral fix in this analysis.

  Informs: Good Souls continuity standard

  Limitations: Duration is not randomly assigned — youth and mentors in enduring matches may differ. The rematching finding comes from one analysis and should not be treated as settled.

  What we don’t claim: Do not claim every mentoring relationship must last a fixed number of months, and do not present rematching as automatically harmful or automatically helpful.

  Open American Journal of Community PsychologyVerification pending

• The test of time: predictors and effects of duration in youth mentoring relationships

  Promising evidence

  Grossman & Rhodes · American Journal of Community Psychology · 2002 · Randomized evaluation · 1,138 adolescents

  Longer-lasting mentoring relationships were associated with the greatest number of improvements, which is why JRX recruits mentors for continuity rather than short-term inspiration.

  Informs: Good Souls continuity standard

  Limitations: Match length was not randomly assigned, so duration effects are partly confounded with mentor, youth, and family characteristics.

  What we don’t claim: Do not state that a relationship must last exactly twelve months, or that duration alone causes improvement.

  Open American Journal of Community PsychologyVerification pending

• Review of mentoring and academic outcomes

  Promising evidence

  National Mentoring Resource Center · Current review · Research review

  Mentoring's academic effects are generally characterized as small-to-moderate. Small effects can still be meaningful when mentoring functions as one part of a larger developmental system, and research suggests academic benefits may operate through self-efficacy, coping skills, help-seeking, growth mindset, school connection, and relationships with teachers and parents.

  Informs: Good Souls Mentors alongside Academic Specialists

  Limitations: A synthesis across varied programs and measures. Effects depend on program structure, relationship quality, and whether academic support is intentional.

  What we don’t claim: Do not present mentoring as an academic intervention, and do not turn mentors into unpaid tutors.

  Open National Mentoring Resource CenterVerification pending

• Practice review: pre-match mentor training

  Emerging evidence

  National Mentoring Resource Center · Current review · Research review

  Direct causal evidence that pre-match mentor training improves youth outcomes is currently rated insufficient, though studies associate training with stronger relationships. JRX trains mentors as prudent institutional practice — safeguarding, boundaries, role clarity, cultural humility — not as a proven causal lever.

  Informs: Good Souls mentor formation and training

  Limitations: Few studies isolate training from other program supports, and training content varies widely.

  What we don’t claim: Do not claim research proves mentor training produces better youth outcomes.

  Open National Mentoring Resource CenterVerification pending

• Longitudinal research on socioeconomic disadvantage and natural mentoring

  Promising evidence

  Peer-reviewed youth mentoring research · 2010s–2020s · Observational study

  Caring relationships exist across every community. Access to diverse networks, professional knowledge, and opportunity can still be unequally distributed — which is what Good Souls Mentoring is designed to widen.

  Informs: Good Souls Mentors — expanding the web of support

  Limitations: Observational and self-reported. It describes access to networks, not the presence or quality of care within families.

  What we don’t claim: Never frame low-income families as lacking caring adults, and never position a mentor as a replacement for family.

  Open Peer-reviewed youth mentoring researchVerification pending

• Natural mentoring, educational beliefs, and racial identity among Black adolescents

  Promising evidence

  Hurd et al. · Peer-reviewed developmental psychology research · 2010s longitudinal analyses · Observational study

  Natural mentoring relationships are associated with stronger educational beliefs, engagement, and positive racial identity among Black adolescents. Culturally responsive mentoring — understanding community, respecting identity, supporting identity development, listening rather than assuming — is part of good design.

  Informs: Culturally responsive Good Souls mentoring

  Limitations: Observational. These studies do not establish that race-matched mentoring alone causes these outcomes.

  What we don’t claim: Do not conclude that race-matched mentoring alone causes these outcomes, and do not create a racial-matching requirement.

  Open Peer-reviewed developmental psychology researchVerification pending

• Analysis of shared racial identity and premature termination in mentoring matches

  Promising evidence

  Peer-reviewed youth mentoring research · 2010s–2020s · Observational study · More than 82,000 matches

  Shared racial identity is associated with matches lasting longer. Race and cultural background can be relevant matching factors alongside student preference, interests, goals, personality, lived experience, mentor capacity, availability, and relationship potential.

  Informs: Good Souls matching practice

  Limitations: Observational program records. Matching is not random, and closure has many causes including mentor availability and program support.

  What we don’t claim: Do not conclude that every Black student needs a Black mentor, or convert this finding into an automatic matching rule.

  Open Peer-reviewed youth mentoring researchVerification pending

• Mentor activity and decision-making in school-based mentoring relationships

  Promising evidence

  National Mentoring Resource Center / peer-reviewed analysis · 2010s · Observational study · 47,699 youth

  Deciding activities together — rather than the adult setting the agenda alone — is associated with stronger mentoring relationships. Mentoring works with a young person, not at a young person.

  Informs: Youth voice inside mentoring

  Limitations: Observational analysis of program data; decision-making style is not randomly assigned.

  What we don’t claim: Do not claim shared decision-making causes specific youth outcomes, and do not use youth voice as a reason to remove adult responsibility for safety and boundaries.

  Open National Mentoring Resource Center / peer-reviewed analysisVerification pending

• Reviews of age-based stereotyping, prejudice, and discrimination toward young people

  Emerging evidence

  Peer-reviewed reviews of adultism and anti-youth ageism research · 2020s · Research review

  Research has documented age-based stereotypes, discrimination, and barriers to meaningful participation for young people. Age-based responsibility is not the concern; arbitrary, humiliating, dismissive, coercive, or unnecessarily controlling uses of adult power are.

  Informs: Authority Without Adultism

  Limitations: A developing field using multiple related terms with varied measures and few causal designs. It should not be presented as having the research maturity of high-impact tutoring or school connectedness.

  What we don’t claim: Do not write that adults discriminate against children, do not define adultism as adults telling children what to do, and do not use this research to argue for adult withdrawal or unlimited youth authority.

  Open Peer-reviewed reviews of adultism and anti-youth ageism researchVerification pending

• Systematic review of student participation in school decision-making and its effects

  Promising evidence

  Peer-reviewed systematic review · 2010s–2020s · Research review · 3,102 publications screened; 32 studies included

  Meaningful student participation — students holding some real influence over decisions or actions — is associated with stronger life skills, democratic skills, student–adult relationships, and school ethos. Student voice is not the same as student presence.

  Informs: Student voice, student councils, YouthLed Community Solutions

  Limitations: Included studies vary widely in design, measures, and setting; most are not randomized. 'Participation' is defined differently across studies.

  What we don’t claim: Do not claim student voice raises grades, and do not present inviting students to meetings or collecting opinions as equivalent to meaningful participation.

  Open Peer-reviewed systematic reviewVerification pending

• Student voice, teacher receptivity, and student academic outcomes

  Promising evidence

  Peer-reviewed education research · 2010s–2020s · Observational study · Approximately 1,751 students

  Student voice and adult receptivity have been associated with stronger engagement and academic outcomes in some studies. Voice is not meaningful if adults collect feedback but remain unwilling to consider it.

  Informs: Adult receptivity — Success Specialists, facilitators, staff

  Limitations: Association, not proof of causation. Students and teachers in more receptive settings may differ in ways the analysis cannot fully separate.

  What we don’t claim: Do not claim that listening to students causes higher grades.

  Open Peer-reviewed education researchVerification pending

• Girlhood Interrupted: The Erasure of Black Girls' Childhood

  Strong evidence

  Epstein, Blake & González · Georgetown Law Center on Poverty and Inequality · 2017 · Observational study

  Research has found adultification bias in which Black girls may be perceived as older, less innocent, or needing less support than White girls of the same age. JRX treats this as a professional-practice reminder: independence, maturity, confidence, leadership, or high responsibility must never be mistaken for needing less care.

  Informs: Preserving childhood; adult training and safeguarding

  Limitations: Evidence about adult perception, measured in survey conditions. It does not describe every adult, nor does every Black girl experience it identically.

  What we don’t claim: Do not write that Black girls are treated like adults, do not use this research as an accusation against teachers, parents, or schools, and do not imply every adult holds this bias.

  Open Georgetown Law Center on Poverty and InequalityVerification pending

• Meta-analysis of character-education program effects

  Promising evidence

  Peer-reviewed meta-analysis · 2010s–2020s · Meta-analysis · 214 studies; 307,512 participants

  Character-education research shows small positive average effects across many programs, with substantial variation in implementation and outcomes. Character has to be practiced, reflected on, demonstrated, and lived.

  Informs: Make a Move virtues, Daily Cypher, Conscious Crusader culture

  Limitations: Findings concern character education broadly, not any single program. Results vary by program, implementation quality, school context, adult behavior, duration, and student experience.

  What we don’t claim: Do not claim character education makes children better people, and do not attach this meta-analysis to any specific JRX or Virtues Project practice as validation.

  Open Peer-reviewed meta-analysisVerification pending

• Randomized evaluations of school-based social and character development programs

  Strong evidence

  Institute of Education Sciences / peer-reviewed randomized trials · 2010s · Randomized evaluation

  Well-designed character-education studies have found no broad positive impacts. JRX publishes this on purpose: naming virtues, displaying posters, teaching a weekly word, or delivering character lessons does not automatically change behavior or develop character.

  Informs: Why JRX does not treat virtue vocabulary as an intervention

  Limitations: Null findings reflect the specific programs, doses, and implementation conditions studied — not proof that character practice cannot matter.

  What we don’t claim: Do not use these findings to claim character education never works, and do not omit them when citing the positive meta-analytic average.

  Open Institute of Education Sciences / peer-reviewed randomized trialsVerification pending

• Pilot study of The Virtues Project in a preschool setting

  Emerging evidence

  Peer-reviewed / academic pilot study · 2000s–2010s · Observational study · Approximately 9 children

  A small preschool pilot reported encouraging changes in observed social behavior. This is exploratory evidence only.

  Informs: Virtues Project strategies with early learners

  Limitations: Very small sample and non-randomized design.

  What we don’t claim: Do not present this pilot as evidence that The Virtues Project produces behavioral outcomes.

  Open Peer-reviewed / academic pilot studyVerification pending

• Four-week Virtues Project-based intervention with fourth graders

  Promising evidence

  Academic study · 2000s–2010s · Observational study · Approximately 110 students; about 98 completed pre/post measures

  A pre/post study of a four-week Virtues Project-based intervention reported increases in self-esteem and peer-relationship measures. Promising, not causal.

  Informs: Virtues language and classroom practice

  Limitations: No comparison group. The study cannot establish that The Virtues Project caused the observed changes.

  What we don’t claim: Do not claim The Virtues Project raises self-esteem or improves peer relationships.

  Open Academic studyVerification pending

• Commitment to character programs across two middle schools

  Emerging evidence

  Academic study · 2000s–2010s · Observational study · Approximately 352 participants

  Implementation research suggests that virtues language support, discussion, and training help adults commit to character practice.

  Informs: Adult training and program commitment

  Limitations: Implementation evidence about adults and programs. It measures commitment, not student outcomes.

  What we don’t claim: Do not present this study as proof of student outcomes.

  Open Academic studyVerification pending

• Longitudinal study of Virtues Project training and leadership practice

  Emerging evidence

  Academic study · 2000s–2010s · Observational study · 9 adult leaders

  A small longitudinal study found adult leaders described changes in virtues language and leadership behavior after training. Exploratory evidence.

  Informs: Adult formation for staff, mentors, and specialists

  Limitations: Very small, self-reported, and specific to adult practice.

  What we don’t claim: Do not claim Virtues Project training improves leadership effectiveness or student outcomes.

  Open Academic studyVerification pending

• No independent evaluation exists yet

  JRX innovation — evaluation needed

  JRX Academy · In development · Observational study

  The JRX hypothesis is that virtues become meaningful when embedded consistently in adult language, student reflection, boundaries, relationships, decisions, service, leadership, conflict, learning, and opportunities to Show & Prove. JRX intends to evaluate that implementation.

  Informs: The combined JRX system — Virtues Project strategies, Daily Cypher, Make a Move virtues, HEART, Life Atlas, Success Specialists, Good Souls Mentors, Quests, service, leadership, reflection, and Show & Prove

  Limitations: JRX does not assume that teaching virtue words develops character. The combined architecture has not been independently validated as one complete intervention.

  What we don’t claim: Do not attach outside character-education, student-voice, or Virtues Project findings to the JRX architecture as if they validated it.

  How JRX will measure thisVerified

• No external evaluation exists yet

  JRX innovation — evaluation needed

  JRX Academy · In development · Observational study

  These elements are JRX innovations informed by established practice. They must be evaluated through JRX's own outcomes before any effectiveness claim is made.

  Informs: Life Atlas, HEART, Make a Move virtues, Conscious Crusader, CLVER Games, student cyphers

  Limitations: These are JRX-original components. They draw on established ideas — reflection, character practice, belonging, applied learning — but the specific architecture has not been externally evaluated.

  What we don’t claim: Do not attach outside research findings to JRX-original components as if they validated them.

  How JRX will measure thisVerified