A 3-hour interactive Practice Pack. Fill the forms as you go — your inputs save in your browser. By the end, click Build my brief and the capstone artefact builds itself.
4 modules~3 hoursInteractiveIndia · India-context
Your progress
0% complete
Your Capstone
One-page SEL Evaluation Design Brief
Walk in with an SEL programme. Walk out with a defensible design — research question, sample frame, selected instruments with justification, data collection plan + budget sketch, mixed-methods integration approach, reporting outline. Built automatically from your module answers.
Module 1 · ~25 min
What kind of evaluation do you actually need?
The number-one reason SEL evaluations disappoint is that the design choice happens implicitly. The team picks the method they've used before. The funder asks for "impact." Someone says "mixed-methods" without specifying what is mixed with what. Six months later, the data answers a question nobody quite asked.
Before any instrument selection, walk through three dimensions explicitly.
Dimension 1 — What is the question, really?
Outcome evaluation — "Did it work? What was the effect?" Needs a comparison group. Tells you whether to scale.
Process tracing — "Why did it happen? What was the mechanism?" Needs time-ordered evidence. Tells you how to redesign.
Theory-based evaluation — combines both. Tests the causal links inside a theory of change. Most demanding; most informative.
Dimension 2 — What evidence will be persuasive?
Quantitative — for representativeness, scale, effect sizes. For SEL: hard because most validated instruments are CASEL-aligned and US-developed.
Qualitative — for depth, mechanism, lived experience. For SEL: especially powerful with children and teachers.
Mixed-methods — most common for SEL. Consistently under-budgeted on integration time.
Dimension 3 — Snapshot or change?
Cross-sectional — one point in time. Cannot support change claims.
Pre-post — minimum credible change design. Needs baseline before the intervention.
Longitudinal — same units across multiple rounds. Strongest for causality.
Retrospective — relies on recall. Use only when no alternative.
Worked example
Sangati runs SEL in 40 Mumbai government schools, three years in. Funder wants evidence for renewal. Budget ₹15L, 9 months.
D1: Outcome primary, process secondary. D2: Mixed-methods (quant primary). D3: Quasi-experimental matched comparison. Honest about the limit (no clean baseline).
Your Design Decision Sheet
Fill these for your programme. Your answers save automatically and flow into the final capstone.
e.g., "Sangati SEL — 40 Mumbai govt schools, grades 4-6, 3 years running"
"We will use the findings to ____" — be specific. With a date.
Honest sentence about the limits.
Saved
Self-check
A client wants to know "why" their SEL programme helped some children more than others. What design type fits?
Outcome evaluation with RCT
Process tracing or theory-based evaluation
Cross-sectional survey
Retrospective design
Correct. "Why" and "for whom" are mechanism questions — process tracing or theory-based evaluation gives you the causal narrative. An outcome evaluation tells you whether it worked, not why.
Module 2 · ~30 min
Choosing your measurement approach
Three persistent SEL-instrument failure modes:
Imported instruments used without cultural adaptation — Indian children interpret items differently than US/UK populations.
Programme-developed instruments that measure exposure to the programme rather than the underlying competency.
Self-report scales used as sole measure with younger children, where social-desirability bias dominates.
Available instruments by framework
Framework
Sample instruments
India status
CASEL
SECA, DESSA, SSIS-SEL
Translated; partial India validation (Tandon 2020); item-level concerns
SEE Learning
Embedded reflection rubrics; Dalai Lama Centre toolkit
Used in Indian pilot schools; not yet rigorously validated
WHO Life Skills / NCERT AEP
NCERT AEP self-rating; SCERT well-being modules; ACER India SEL skills assessment
Most aligned with government schools. ACER India is the strongest validated option
Indian cultural validation ≠ translation. Translation handles language; adaptation handles whether "self-regulation" means the same thing in a Mumbai government school as in a Boston suburb. Budget at least one cognitive-pretesting round per instrument.
Your Instrument Selection Matrix
Pick a primary quant instrument + secondary triangulation source. These flow into your capstone.
e.g., "Grades 4-6, Mumbai government schools, Hindi/Marathi/English code-switching"
Name + framework — e.g., "ACER India SEL skills assessment (WHO Life Skills aligned)"
What does it under-measure? What bias does it carry?
Saved
Self-check
You're evaluating an SEL programme in Grade 3 government schools. Which instrument is LEAST appropriate as your primary measure?
Behavioural observation (CLASS)
Teacher rating scale
Self-report CASEL competency scale (translated, not adapted)
Direct assessment (Head-Toes-Knees-Shoulders)
Correct. Two compounding problems: Grade 3 children's self-report is dominated by social-desirability bias, and an unadapted CASEL scale has India-validity concerns. Behavioural observation or direct assessment is methodologically stronger for this age.
Module 3 · ~30 min
Designing the data collection (without breaking the budget)
Three operational realities for Indian SEL fieldwork:
School-year calendar — June–April workable; May, October (festivals), exam weeks are not. ~5-6 months of clean fieldwork per academic year.
Consent architecture — SMC approval, parent informed consent, child assent. All three. Plan a 6-week consent window.
Field team composition — women enumerators, local-language native, child-rapport trained.
Sample size, without illusions
Single-arm pre-post: 200–300 children minimum to detect ~0.3 SD effect. Below 100, uninterpretable.
Qualitative interview: ₹2,500–4,500 each (incl. transcription)
FGD: ₹6,000–10,000 each
Behavioural observation: ₹4,000–7,000 per session (2 observers, IRR)
Teacher/parent report: ₹150–300 per respondent
The integration line item
Mixed-methods studies routinely under-budget integration. Budget 10–15 days of senior researcher time post-fieldwork specifically for integration meetings, joint coding, contradictory-finding resolution. Without it: two parallel reports, not one integrated study.
Your Data Collection Plan + Budget Sketch
Fill the numbers. Totals are computed below. These flow into your capstone.
e.g., "200 children across 8 treatment schools + 200 across 8 matched control"
Map against school-year calendar. Note avoid-windows.
SMC + parent + child + IRB. By when?
Saved
Self-check
A funder offers ₹15L for an SEL evaluation. You design 400 children quant + 30 interviews + 6 FGDs + behavioural observation in 20 classrooms. The agency comes back saying it can't be done. What's the most likely budget killer?
The 400 children quantitative
The 30 interviews
The behavioural observation in 20 classrooms
The 6 FGDs
Correct. Behavioural observation needs 2 trained observers per session, IRR checks, multiple sessions per classroom. At ~₹5K per session × multiple rounds × 20 classrooms, it eats ₹4-6L by itself — typically the budget killer in ₹15L studies. Drop it or sub-sample to 5-6 classrooms.
Module 4 · ~25 min
Analysis, reporting, and the honest framing
Two patterns recur in SEL reporting:
Over-claim from underpowered data — funders who've seen this once stop reading.
Under-claim from mixed findings — programme teams stop fighting for funding.
What to expect from SEL analysis
Effect sizes will be small. Kraft (2020) calibrates "typical" school-intervention effect at 0.05–0.20 SD.
Statistical significance ≠ meaningful change. With 500 children, very small effects become "significant" without being important.
Mixed findings are the rule. One competency improves, another doesn't. One subgroup benefits, another doesn't. Report all of it.
Integration discipline (mixed-methods)
Where strands agree → strongest finding.
Where strands disagree → most interesting question.
Where one strand fills a hole the other opens → mechanism explanation.
Reporting structure that works
1-page executive summary — what we asked, what we found, what we recommend
3-5 page findings — organised by research question, not data type
2-page implications
Appendices — full results, instruments, quotes
The honesty test
Hand the draft to a colleague unfamiliar with the programme. Ask them to summarise the headline finding in one sentence. If overstated → too strong. If they can't find the signal → too weak. Iterate.
Your Reporting Plan
Fill the audience map + the honest framing.
Format + length + key questions they will ask
In what language? In what form?
How will quant + qual strands actually integrate? Joint coding? Side-by-side? Triangulated headline?
"This evaluation will tell us ___ and will NOT tell us ___."
Saved
Self-check
Your SEL evaluation finds: significant effect on self-awareness (+0.18 SD, p=0.04), no effect on regulation, positive teacher-reported climate. Which is the most honest headline?
Correct. Option 1 over-claims (one significant effect ≠ transformation). Options 3 and 4 under-claim (real findings exist). Honest framing names exactly what was found, in what direction, with what limits.
Capstone
Your SEL Evaluation Design Brief
You've completed the four modules. Click Build my brief to compile everything into a single one-page artefact. Copy as markdown, print as PDF, or share with your team.
One-Page SEL Evaluation Design Brief
Click "Build my brief" — your module answers will be pulled into the artefact. Edit/refine afterwards if needed (click in the box and type).
Your brief will appear here when you click "Build my brief".
It will draw from your answers in Modules 1-4 (which are saved in your browser). Empty fields show as placeholders — you can either go back and fill them, or edit them here directly after building.
Share the brief with one colleague before circulating widely. Their first reaction is the most accurate signal you'll get on whether the framing lands.