The Strength-
A four-audience survey that helps a school see where it's strong in student-centered learning and where it's growing. Built on the seven principles from Education Evolving, with MLCN's equity lens layered across.
This page describes what the survey measures, how scores are computed, what schools see in their report, and the privacy commitments that go with it.
A four-level hierarchy.
Every question on the survey lives in a four-level taxonomy: a cluster contains principles, each principle contains indicators, and each indicator is worded in one item per audience.
Social Needs · Education Experience · Student Agency
Top-level groupings for the highest-level rollups and the report's slider casing.
The Education Evolving framework + cross-cutting Equity
Positive Relationships, Foundational Needs Met, Positive Identity, Student Agency & Ownership, Real-World Relevant, Competency-Based, Anytime, Anywhere Learning.
The constructs being measured
An indicator is the thing we're actually asking about. Reports roll up to this level — strengths, growth areas, and perception gaps are all indicator-level.
One statement per indicator × audience
The actual text shown to a respondent — 26 indicators rendered in four audience-appropriate variants. Items differ in phrasing per audience but measure the same construct, which is what makes a student answer and an educator answer comparable on the same thing.
Four parallel instruments.
The same set of indicators is administered to four different audiences, each with audience-appropriate wording. Schools choose which to run; most administer at least one student instrument.
MS Student
Simplified phrasing of every indicator; one N/A option (“I'm not sure”).
HS Student
More nuanced phrasing of the same indicators; two N/A variants (“I don't know” and “I don't understand this question”) so the instrument can learn from confusion as well as ignorance.
Direct Educator
Any adult who works in the school — broader than “teacher.” Includes its own N/A (“Doesn't apply to my role”) for items that fall outside a specialist's daily work.
Caregiver
Parents and guardians, when a school wants their lens too. Optional — most pilot schools haven't used this instrument yet.
A school can administer one, two, three, or all four. Many start with student-only and add the adult instruments in year two — the report architecture handles partial sets gracefully (cells with no data are blanked, not zeroed).
Five points, asymmetric on purpose.
Each item is answered on a 5-point asymmetric Likert scale, plus an N/A. Three positive levels and two negative — no neutral middle.
| Value | Label | Counts? |
|---|---|---|
| 4 | Strongly Agree | Yes |
| 3 | Agree | Yes |
| 2 | Somewhat Agree | Yes |
| 1 | Disagree | Yes |
| 0 | Strongly Disagree | Yes |
| N/A | Per-audience N/A — “I'm not sure” (MS), “I don't know” / “I don't understand this question” (HS), “Doesn't apply to my role” (DE), “I don't have enough information” (caregiver) | Excluded |
Why asymmetric?
Three positive levels, two negative, and no neutral. Adding a third positive level (“Somewhat Agree”) creates an intentional pull toward the strength side — which is the whole point of a strength-indicator.
Why N/A doesn't hurt the score
N/A responses are excluded from both the numerator and the denominator. A school with high N/A on an indicator still gets a clean Strength score among the people who could answer — but the report flags “interpret with caution” when ≥20% mark N/A and withholds the score entirely at 40% or more.
“I don't understand”
A distinct N/A variant on the student instrument. Selecting it opens an optional text field; what students write goes back to MLCN as instrument-quality signal so the wording can improve over time.
A Strength score from 0 to 100.
Not a percentage of respondents who agreed. The Strength score is the average position respondents took on the 0–4 agreement scale, rescaled to 100. A school where every respondent simply “Agreed” scores 75; reaching 100 requires unanimous “Strongly agree.”
# item (one indicator × one audience) score = Σ (count × value) # 0..4 across scoring options total_possible = 4 × valid_n Strength score = score / total_possible × 100 # principle (within an audience or union pool) principle score = mean of its indicator scores # cluster cluster score = mean of its principle scores
Mean of scores, every level
Each principle is the mean of its indicators; each cluster is the mean of its principles. A 1-indicator principle counts equally with a 4-indicator one — the framework decides the weight, not the indicator-count accident.
Suppression + N/A
Indicators with fewer than 5 valid responses are hidden for anonymity. Indicators where ≥40% of respondents marked N/A are withheld too — the effective base is too thin. Between 20% and 40% N/A, the score still renders with a caution flag. A principle drops out of its cluster mean if fewer than 2 of its indicators survive.
Audiences stay separate
Reports are organized per audience — educators, students (MS / HS), caregivers. The gap between audiences is the signal the instrument is designed to surface, so we don't collapse them into a single school number. The comparison tool lets schools compare any two pools.
No imputation
Missing data is missing. We never fill in a value for someone who didn't answer.
Six sections, one story.
Once the survey closes, school administrators get a story-forward report that opens with the headline answer and drills down only as far as you want to go. No respondent names, and every indicator with fewer than five valid responses is suppressed to protect anonymity.
Summary
A short narrative paragraph above an orientation stat strip, then Top 3 strengths and Top 3 growth areas. The narrative pulls in two strengths, two growth areas, the headline implementation gap (if any), and three reflection questions. There's a slot for a human-authored follow-up so a school can frame the numbers in their own words next to ours.
Clusters
Cluster averages as a per-audience bar chart — one bar per audience for each of the three clusters. A filters card above the chart lets you hide audiences or layer a comparison view (e.g., one teacher's class) as an extra bar alongside the audience bars, so the slice reads as a comparison rather than a subset.
Insights
Editorial highlights. A perception-gap headline naming the biggest implementation gap or experience strength, and a trend card showing the principles that moved most since the prior administration — when one exists.
Principles
A heptagonal radar — adult view vs student view across all seven Education Evolving principles. Equity recurs inside every cluster, so it sits alongside the radar as a cross-cutting callout rather than a single spoke. Each principle's one-line definition is available inline.
Gaps
The perception-gap dumbbell. Each row is a principle, with two dots showing where Pool A and Pool B land. You pick the pools — students vs adults, MS vs HS, whatever's useful. Indicator-level gaps of 10 or more points are listed below the dumbbell, each tagged with a quadrant (implementation gap, experience strength, aligned strength, aligned growth) so the actionable patterns surface first.
Indicators
The dense view, one cluster tab at a time: every indicator × audience scored against optional network averages and prior-administration trends. Click any column header to sort. A “Show survey wording” expander on each row reveals the exact statement each audience saw.
Plus a free-text Comments section when respondents leave cluster-end reflections (admin-only), a Participation table at the bottom, and a revocable share link that can hide the network average or prior trend line if you don't want to expose them to the recipient.
School-only by default.
Three structural commitments that hold regardless of MLCN's policies — they're how the system is built, not how we promise to behave.
Network-level analysis we publish leans into rollups. Individual respondent rows are not shipped upstream.
Cluster-end reflections are visible to that school's administrators only. They never appear in MLCN aggregate views and are not visualized to anyone in the report.
Default retention is 90 days after the administration close date. Names, emails, class lists, and roles are deleted; the answer rows (which never carried personal identifiers in the report scope) remain. The purge is auditable inside.
A “network average” is only shown when two or more schools contribute, so no single school's results can be read off it. Your own school is always excluded from the network number on your report.
The full picture — how FERPA, PPRA, COPPA, and Minnesota data law apply, what we collect, and our subprocessors — lives in the privacy policy. Districts that need a signed agreement before using the platform can review our Data Privacy Agreement (or send us their own to sign).
Built on the framework — refined for MLCN.
The seven principles of student-centered learning come from Education Evolving, a long-time MLCN partner.
MLCN's additions
Equity is treated as a cross-cutting principle that appears inside every cluster, not as an add-on principle among the seven. The instrument wording has been refined across several pilot years with feedback from the five schools that ran the previous Google Forms version of the survey.
Versioning
The platform stores indicators, items, and scales as data, not code. Future instrument revisions are migrations — the report architecture stays the same. Schools running last year's instrument can still re-open last year's report against the historical version even after the active instrument moves on.
Bring the Strength-
The tool is included with MLCN membership. Join the network and you can launch an administration the same day.