2026 Co-Teaching Analysis
Linking partnership-level support factors to what 1,655 students reported — and to grades in the six core co-teaching courses (IM1, IM2, English 9 & 10, Biology, Physics). Grades are for students with IEPs only; the student-survey responses come from all students in the co-taught classes — with and without IEPs.
Correlational analysis across matched partnerships. Strength shown as weighted correlation (r) with student-reported outcomes — not proof of cause.
Common planning time sits above the rankings below because co-teaching's defining practices — co-planning, co-instructing, and co-assessing — depend on it by definition. Co-teaching research and professional practice treat shared planning time as a precondition for the model to work, not as one optional support among many.
Support factors, ranked explore
Each bar is the correlation between a support being in place — rated by district staff in the co-teaching success-factors review, not reported by students — and a student outcome, for the 10 factors that varied enough — and were measured cleanly enough — to be testable. Toggle between the two outcome measures. Longer blue bars = stronger positive link; pale bars near zero = little measurable impact. This ranking is survey-based only — it reflects what students reported, not grade outcomes. In the Grade Outcomes tab, these same structural supports do not separate teams by grade performance.
How to read the numbers: each is a correlation from −1 to +1 — 0 means no link between the support and the outcome, and the further from 0, the stronger the link. As a rule of thumb, about ±0.15 is a small link, ±0.30 moderate, and ±0.50 strong — the word beside each number says which band it falls in.
“Varied co-teaching methods” flags teams that used a variety of co-teaching approaches — station teaching, team teaching, parallel teaching, or alternative teaching — beyond just the basic one-teach-one-support model.
These factors can't be evaluated here — either because they were present in nearly every partnership (no comparison group), or because the column is too noisy a proxy for the thing it's meant to capture. A near-zero or negative correlation for these is not evidence that the factor doesn't matter; the data simply can't speak to it. They are kept out of the ranking rather than shown as findings.
The dosage effect key finding
Partnerships grouped by their total number of supports in place. Student confidence growth climbs steadily as supports accumulate — the strongest pattern in the data (r ≈ 0.58).
Confidence-growth scale, −2 to +2: +2 = much more confident, +1 = a little more, 0 = about the same, −1 = a little less, −2 = much less. Higher = more growth.
Observed co-teaching quality key finding
A 0–3 score from classroom observation — one point each for using varied co-teaching models, Specialized Academic Instruction (SAI) in class, and participation in professional growth. Unlike the structural supports, this measures what the team actually does. It tracks confidence growth almost as strongly as the full 15-factor checklist (r ≈ 0.54), from a single measure.
Confidence-growth scale, −2 to +2: +2 = much more confident, +1 = a little more, 0 = about the same, −1 = a little less, −2 = much less. Higher = more growth.
What students valued
Most-helpful aspects of having two teachers, by number of students selecting each (multi-select).
Support by school
Average total success-factor score per partnership. Schools that stack more supports tend to show stronger student outcomes.
In students' own words free responses
Analysis of written comments students left about co-teaching. Sentiment runs strongly positive — and the recurring themes point to concrete things two teachers made possible.
The themes below capture the comments that named something specific, grouped by keyword — an approximate summary, not an exact count. Roughly half of all comments were short or general praise ("they're really nice," "they helped me") that don't sort into a theme, so these counts reflect the actionable feedback. Genuine dissatisfaction was rare — about 1% of comments.
Grade outcomes objective data
The Support Factors tab links district-staff assessments of each partnership to student-reported outcomes; Student Voice is what students said directly. This section uses actual quarter grades — an objective outcome — for co-taught and non-co-taught sections of the six core co-teaching courses (Integrated Math 1 & 2, English 9 & 10, Biology, and Physics). These grades are for students with IEPs only (students with disabilities) — not their non-IEP classmates; the Student Voice survey, by contrast, includes all students in the co-taught classes. Every official quarter grade counts — Q1, Q2, Q3, and Q4, the four permanent transcript grades; the interim progress marks between them are excluded.
The co-taught vs. non-co-taught gap, three ways
The same students compared three ways, all within the six core courses. Raw: co-taught students average about a quarter of a point below non-co-taught students. Same course & quarter: holding the exact class constant, about a third of a point — co-taught sections concentrate the highest-need students. Within-student: measured against each student's own grades across their whole schedule (which includes their easier elective courses), their core co-taught grades run about eight-tenths of a point lower, since core academics are harder than electives. Three different lenses on the same students, not one sliding scale.
Pass rates — co-taught vs. non-co-taught
Across the six core courses, the share of quarter grades that were passing (D or better) — co-taught sections vs. non-co-taught sections of those same courses. This holds the course constant but still doesn't account for which students enroll (co-taught sections concentrate the highest-need students). Pick a course below to see it on its own.
Do support factors explain grade performance?
Not really. Teams are grouped by how many support factors they have in place (low / mid / high). Their grade performance — how their co-taught students did vs. their own grades in other classes — is roughly flat across the groups (the support-to-grades link is about −0.10, essentially none). More structural supports did not translate into higher grades. Observed teaching quality tracks what students report (see Support Factors) — but it does not track grades either.
These supports correlate negatively with grades. This is almost certainly reverse causation: these supports are directed to teams that need extra help — coaching in particular to the newest teams — so receiving them marks a team that needed support. This is not evidence they hurt — cutting them would be the wrong conclusion.
Spread of team performance
How co-taught students did vs. their own grades in other classes, across co-teaching teams. Teams are not named — this is a guide for coaching and inquiry, not an evaluation scorecard, since it can't fully account for how student need is distributed between teams.
By school · compared handle with care
Two independent rankings, shown together so you can see where they agree and disagree. Grades only ranks each team by how its co-taught students did vs. their own grades in other classes, across all four quarter grades (Q1–Q4) — so 0 means right at their usual level, a minus means lower and a plus means higher (it is not a raw pass rate). Survey only ranks them by their students' ratings — how much two teachers helped, and confidence growth. Where a team tops one column but not the other, the objective and experiential signals disagree — itself worth a closer look. Survey ratings cluster high, so that column separates teams by smaller margins and is the noisier of the two. Adelante and Independence are not shown.
By school · combined handle with care
This tab blends two independent signals into one ranking, weighted 50/50: the team's grade performance vs. students’ own other classes, and its students' survey ratings (how much two teachers helped, and confidence growth). Each signal is standardized across all qualifying teams before blending, so neither dominates by scale. Only teams with both ≥12 graded students and ≥8 survey responses are included. Where this ranking differs from the grades-only tab, it means grades and student experience disagree — itself worth a closer look.
What the evidence points to
- ▲Observed quality is the most consistent signal. Teams that actually use intentional teaching, SAI, and pursue professional growth score higher on student confidence and on adjusted grades — the only factor that holds across both outcome measures.
- ▲On perceptions, stacking supports helps. The number of supports in place tracks student confidence growth (r ≈ 0.58), even though no single structural factor stands out alone.
- ▬On grades, structural supports don't separate teams. Low- and high-performing teams carry almost the same total support score. Supports may set conditions, but they don't by themselves predict grade outcomes.
- ●Coaching & PD look negative — but aren't. They are directed to struggling teams, so they mark a team that needed help. This is reverse causation, not evidence against them.
- ●Co-taught grades are lower by design. Co-teaching serves the highest-need students; the raw grade gap nearly vanishes once the same courses are compared. It is not a measure of co-teaching quality.
- ●Several factors can't be judged. Training prior to start, admin understanding, need-based assignment, and common prep were near-universal or measured too noisily. This is not evidence against them.
- ▲Students endorse it in their own words. Of 1,407 written comments, half were positive in tone and only ~1% negative — most often praising clearer explanations and faster, more available help.
