SEL Evaluation in India: What Works, What Doesn't, and How

The single most-cited paper in the SEL evidence base — a meta-analysis of 213 school-based SEL programs covering 270,000+ students. Documents an 11-percentile-point academic gain and substantial improvements in social-emotional skills and behaviour. Required reading. Almost every Indian SEL program claims alignment with this evidence; understanding the methodological constraints (US contexts, high-fidelity implementations) is critical for honest extrapolation.

The Collaborative for Academic, Social, and Emotional Learning's framework of five competencies: self-awareness, self-management, social awareness, relationship skills, responsible decision-making. The reference point for most program design and most measurement instruments. The framework is well-thought-out; its operationalisation for non-US contexts is where serious work begins.

The WHO Life Skills framework predates CASEL by over a decade and remains influential in Indian government policy (NCERT life-skills modules; NEP 2020 references). Ten core skills: decision-making, problem-solving, creative thinking, critical thinking, effective communication, interpersonal relationships, self-awareness, empathy, coping with emotions, coping with stress. Read alongside CASEL for the framework Indian government implementations actually use.

The policy text that opened the door to large-scale SEL investment in Indian government schools. Cites "holistic development of all faculties" repeatedly; explicit on life skills, social-emotional development, mental health. Evaluators working with state systems should know exactly what NEP language enables; it is the warrant for investment.

The most coherent recent synthesis of the science of learning and development — incorporating SEL alongside cognition, motivation, environment. Useful for evaluators who want to think about SEL not as a discrete intervention but as a property of whole school environments. Implies different measurement choices than narrow program-evaluation designs.

The most rigorous synthesis of SEL evidence from low- and middle-income countries, including substantial Indian content. Major finding: effects on competencies are real but smaller than in HIC contexts, and academic-outcome effects are inconsistent. Methodological honesty about what is and isn't known. The reference point for "what does the field know about SEL in India."

A global research program with substantial Indian content — particularly on how SEL is framed and measured across diverse policy contexts. Useful for understanding how Indian SEL implementations look relative to other LMICs.

EI's Sangati program is one of the most rigorously evaluated Indian SEL implementations. Annual reports document classroom-practice change, child-level competency gains on adapted scales, and the fidelity-effect-size relationship clearly. Industry standard for what an honest NGO SEL evaluation looks like.

India's largest public-sector SEL implementation — the Delhi government's Happiness Curriculum covers ~800,000 children daily across 1,000+ government schools. The evaluations published to date have methodological limits (no counterfactual; outcome measures predominantly survey-based), but the operational learnings are unmatched. Required reading for anyone designing SEL at state scale.

Among the most carefully-evaluated low-cost SEL-adjacent interventions in India. Short, locally-recorded audio messages delivered to parents via phone. Measurable improvements in parenting behaviour and child outcomes at very low cost. Methodologically conservative — careful baseline, mixed-methods follow-up, honest about effect sizes.

The cleanest methodological treatment of why CASEL-aligned instruments need adaptation rather than translation for Indian children, and what that adaptation actually requires. Required reading before adopting any imported scale for Indian SEL evaluation.

Apni Shala has done careful self-evaluation of its school-based SEL programs across Mumbai government schools. Reports document the fidelity-effect-size relationship clearly and have evolved toward implementation-focused evaluation over impact-claims-focused evaluation. Useful as evolution of practice over six years.

Jones's broader argument — that SEL is poorly served by program-evaluation paradigms because the developmental processes it targets unfold across years and across multiple contexts — applies acutely to India. Required for evaluators tempted to do one-year impact studies of multi-year developmental processes.

The conceptual foundation for direct-assessment of self-regulation (head-toes-knees-shoulders, dimensional change card sort, etc.) — measures that bypass child self-report and adult report. Indian implementations of these direct measures have produced more reliable data than scale-based approaches; the methodological tradeoff is they capture narrower constructs.

ACER has done the most systematic India-specific work on SEL measurement. Their adapted scales are now the closest the field has to "standardised, validated, Indian-appropriate" instruments. Available on request; some are open. Foundational for serious Indian SEL evaluations.

The strongest case for direct-assessment of self-regulation over rating-scale approaches — and the practical limitations. Where SEL evaluation is moving methodologically.

How to read SEL effect sizes responsibly. Kraft's argument that the "typical" effect size for cost-effective school interventions is 0.05–0.20 SD — not the 0.30+ that program-marketing materials suggest — is required calibration for honest reporting.

Banaji's work explores why Indian children interpret SEL scale items differently than US/UK children — and what the implications are for cross-cultural validity. Item-level cultural-validation studies are tedious; this work demonstrates why skipping them produces unreliable evaluation.

The clearest brief framework for matching evaluation design to question. J-PAL's bias is toward RCTs, but the guide is honest about when RCTs aren't the right tool — and SEL is full of those situations (multi-year processes, contamination across schools, context-specific competencies).

Realist evaluation ("what works for whom in what circumstances and why") fits SEL evaluation unusually well — because the answer to "does SEL work" almost always depends on which competency, which child, in what implementation context. Required for the methodological pluralism the field needs.

For genuinely novel SEL programs in unfamiliar Indian contexts — caste-aware SEL, conflict-affected school SEL, displacement-context SEL — developmental evaluation often fits better than impact-verification. Patton's framework gives evaluators permission to prioritise learning over verdicts in years 1–3 of a program. See the Evaluator mode of the SEL Simulation Game for how this plays out in practice.

The most practical introduction to difference-in-differences, matching, and regression discontinuity for evaluators who can't randomise. The most commonly used designs for Indian SEL evaluations; this material is critical to using them well.

When (and how) to integrate qualitative and quantitative methods in an SEL evaluation. The field consensus is increasingly that mixed-methods is the default; this material treats it as a craft rather than a checkbox.

A sharp Indian-context critique of the credibility-revolution's overreach. Drèze's argument is not that RCTs are bad — it's that the field's enthusiasm for them has crowded out evaluations of questions that don't lend themselves to randomisation, and that the political economy of "RCT evidence" has its own biases. Required for evaluators who don't want to absorb J-PAL's framing uncritically.

The single most useful operational resource. Especially the "rainbow framework" for systematic evaluation design and the comparative methods pages. Used by working evaluators globally; particularly strong on non-experimental methods that the J-PAL world under-covers.

CASEL's catalogue of SEL measurement instruments — searchable by competency, age, format, and validation status. Critical resource. Note: most instruments listed are US-validated; cross-check Indian adaptation status separately.

Practical implementation challenges, in their own voice. SEL-specific content is limited; broader impact-evaluation operational content is excellent.

If you are coming to SEL evaluation from a programmatic background and need to build basic evaluation literacy, start here. Covers Theory of Change, indicator design, data collection, and analysis at practitioner level.

Interactive tool for designing your own evaluation plan — useful for the structured-thinking exercise of moving from question to design to instrument.

Filterable evidence base of teacher-effectiveness interventions in India. Useful comparison for thinking about how SEL evaluation evidence compares to other education-evaluation domains in terms of replication, effect sizes, and methodological standards.

India's most influential education thinker provides the deepest critique of SEL implementations that treat children's emotional lives as objects of intervention rather than as expressions of children's relationship to authoritarian school cultures. Kumar's argument: SEL programs that don't address why Indian schooling is emotionally constraining will produce "well-regulated" children adapted to badly-designed schools. Required.

Rampal's work argues that SEL, like all curriculum, embeds assumptions about whose emotional repertoires count as legitimate — and these assumptions disadvantage children from working-class, caste-marginalised, and tribal backgrounds. Evaluators measuring "self-regulation" or "empathy" should know whose self-regulation and empathy the measures reflect.

A growing body of work asking whether Indian SEL implementations should draw more from indigenous traditions — Tagore, Krishnamurti, Gandhian education, vipassana — than from imported CASEL frames. Some of this is intellectually serious; some is symbolic. Evaluators should know the conversation exists.

The anthropological foundation for the argument that "emotions" are culturally constructed and that imported measurement frames carry their cultures of origin into the populations they measure. Older than SEL itself; still required for evaluators thinking carefully about cross-cultural validity.

Pritchett's broader critique of how Indian education statistics overstate progress applies acutely to SEL — where the field is even more reliant on self-reported and adult-reported measures than the rest of education evaluation. A useful corrective to optimism.

What works (with replicated evidence):

— Sustained, structured SEL programs with 50–70 hours per year of dedicated time, delivered by teachers who receive ongoing coaching rather than one-time training, produce moderate effects on social-emotional competencies and small effects on academic outcomes. Effects compound over 3–5 years.

— Teacher-relationship interventions (mentoring, restorative practice, classroom-climate work) consistently outperform child-curriculum-only interventions for student wellbeing outcomes.

— Parent engagement through frontline workers (Dost-style audio nudges, integration with ASHA/Anganwadi visits) produces effects at very low cost.

— Whole-school approaches (where SEL is embedded across teaching, not isolated to an SEL period) outperform isolated curriculum approaches on every replicated measure, but are operationally hardest.

What doesn't work (or is widely overstated):

— One-time teacher training in SEL methods produces almost no lasting practice change. The single most common failure mode.

— SEL curricula delivered without protected time are routinely cut during exam preparation, festivals, and administrative crises. Designs that don't protect dosage produce null results.

— Self-reported child surveys as primary outcome measure are unreliable in Indian government school contexts, especially for younger children — social desirability bias, comprehension issues, and item-cultural-validity issues compound. Use as supplement, not as headline.

— Cross-grade comparisons within the same school year are confounded by teacher quality and class composition. Common informal design; rarely defensible.

What we don't yet know:

— Long-term (5+ year) effects of Indian SEL programs on labour-market and life-course outcomes. The Heckman-style evidence base is global; India-specific longitudinal data does not yet exist.

— Whether mid-career teacher SEL competence is more shapeable through professional development or pre-service training. Critical for policy but evidence is thin.

— How to evaluate SEL programs in conflict-affected, caste-conflict, and displaced-population schools, where standard measurement assumptions break.

Methodological priorities for the next five years:

— India-validated direct-assessment of self-regulation (replacing self-report).

— State-system integration of SEL measurement (so any program can be evaluated against a population baseline).

— Longitudinal cohort studies (which India largely lacks for SEL).

— Honest reporting of null and mixed findings — currently under-published.