26 KPI Components, One Framework

Computed by the base framework, with Risk and Promise tiers.

What it needs

• Daily attendance per student per course
• Course enrollment roster
• School calendar (instructional days)

How it’s calculated

Present days ÷ Scheduled days, calculated per student per course; aggregated upward.

What it tells you

The single strongest predictor of academic outcomes in K-12 research. A student below 90% in any course is statistically at-risk for that course; below 80% predicts course failure with high reliability.

Act on it

• Student/parent contact at 90% threshold
• Counselor referral at 85%
• Formal intervention plan at 80%
• Teacher dashboard highlights students approaching thresholds

Can you act on it alone?

STRONG CORRELATION, LIKELY CAUSAL One of the most robust findings in education research. Direct mechanism (missed instruction → missed learning). Safe to act on directly.

The Promise reading

A student consistently at or above 95% attendance is one of the school's most reliable indicators of healthy engagement. When sustained across multiple courses the student is a candidate for peer-mentoring roles, leadership opportunities, or simply public recognition of a year of showing up — which research consistently shows is itself an intervention against future drift.

Recognise it

• Teacher mention at quarterly progress reports — students rarely hear "you've been here for 95% of the year"
• Counselor invitation to peer-mentoring or activity leadership
• Principal recognition at term-level (with student consent)
• Cohort-level reporting at District tab — "schools with 75%+ Steady-or-above" is a board-level health metric

Lens: Symmetric — same KPI carries both directions

Computed by the base framework, with Risk and Promise tiers.

What it needs

• Current course grade per student per course
• Passing threshold per course (typically 50% or 60%)
• Grade calculation method (weighted categories)

How it’s calculated

Current grade − passing threshold; flag when delta < 5 percentage points or grade is below threshold.

What it tells you

Direct indicator of academic standing. A student within 5 points of failing has limited room to absorb a poor assessment. Surfaces students who can still be helped vs. those already failing.

Act on it

• Tutoring referral at "within 5 points" flag
• Assignment recovery offered before final grade locks
• Teacher checks for missing assignments first (often the simplest fix)
• Parent notification at failing status

Can you act on it alone?

DIRECT MEASURE Not an inference; this IS the academic outcome. No causal interpretation needed.

The Promise reading

A student consistently 10+ points above the passing threshold has academic margin — room to take on harder material, mentor peers, or recover from one bad assessment without falling into the risk tiers. Students Climbing into this zone after a bad term are the strongest pedagogical signal in the dataset; the recovery is more durable when met with explicit recognition before the next assessment.

Recognise it

• Teacher feedback specifically naming the improvement, not just the grade ("your last three assessments have been 65 → 72 → 78")
• Department-level recognition at the Principal tab
• Distinguish Star (consistent high) from Climbing (recent improvement) — the praise content differs
• Pair with attendance KPI for compound recognition

Lens: Symmetric — points above passing is the direct Promise signal

The base uses each student’s unexcused-absence rate to raise their attendance tier. The full component adds the rate as a KPI in its own right.

What it needs

• Daily attendance with reason codes
• Reason taxonomy (excused / unexcused / disciplinary / school activity)

How it’s calculated

Unexcused absences ÷ Scheduled days, per student per course.

What it tells you

The strongest behavioural risk signal in attendance data. Unexcused absences carry roughly 2–3× the predictive weight of excused absences for course failure and dropout.

Act on it

• Truancy protocol activation
• Counselor outreach (not punitive — often signals a problem at home or with the class)
• Differentiated from medical absences which route to wellness support instead

Can you act on it alone?

STRONG SIGNAL, MULTIPLE CAUSES Unexcused absence is itself a behaviour, but the underlying cause varies (disengagement, family issues, mental health, transportation). Treat the KPI as a flag for inquiry, not a diagnosis.

The Promise reading

Asymmetric KPI. "Low unexcused rate" at the student level isn't actionable as celebration — the absence of misconduct isn't itself a Promise signal. Aggregated at the school or class level, a sustained low rate (or a falling trend) is a useful school-health indicator and feeds the District Promise lens.

Recognise it

Not at the student level. School-level: report alongside chronic absenteeism trend for board context. Class-level: name teachers whose sections show consistently low unexcused rates in coaching conversations (with the same governance caution as KPI #19).

Lens: Asymmetric — "low unexcused" isn't actionable celebration on its own

What it needs

• Per-assignment submission status
• Per-student gradebook entries (including zeros for missing work)

How it’s calculated

Assignments submitted ÷ Assignments due, per student per course.

What it tells you

Often the proximate cause of failing grades — students fail because of zeros, not because of poor performance on completed work. Completion rate is the leading indicator; grade is the lagging one.

Act on it

• Missing work recovery before grade is finalized
• Late work policy reviews at the school level if patterns emerge
• Particularly powerful when paired with attendance — high attendance + low completion suggests a different problem than low attendance

Can you act on it alone?

DIRECT MEASURE OF BEHAVIOUR Strong predictive link to grades, but doesn't explain WHY students don't submit (unable, unwilling, unaware). Use it to surface the question, not answer it.

The Promise reading

A student consistently at 95%+ completion is demonstrating the foundational habit that predicts academic success in K-12 research. Completion rate is also the cleanest Climbing signal at the individual assignment level — a student who was at 50% completion last term and is now at 90% deserves immediate acknowledgement before the new pattern lapses.

Recognise it

• Teacher reinforcement in the same week of submission — recency is more important than scale
• "Habit-naming" feedback — focus on the consistent behaviour, not the grade outcome
• Explicit pairing with attendance — students with both high attendance AND high completion are the Steady tier's foundation
• Class-level recognition when section completion rate rises (avoid single-student leaderboards)

Lens: Symmetric — completion ≥ 95% is a strong Promise signal

The base student risk summary already combines attendance and grade. The full component adds absence reasons.

What it needs

• All of the above KPIs combined
• Configurable thresholds and weights

How it’s calculated

Tiered flag (None / Watch / At-Risk / Critical) based on combination — e.g., "At-Risk" if attendance < 90% OR grade within 5 points of failing OR unexcused rate > 10%.

What it tells you

Single actionable signal for frontline staff. Avoids the trap of monitoring 12 metrics in isolation. Tiered structure matches tiered intervention systems (MTSS).

Act on it

• Drives the daily "who do I check on today" list for teachers and counselors
• Feeds escalation workflows to administrators
• Triggers reason-appropriate routing (medical → nurse, behavioural → admin, academic → tutoring)

Can you act on it alone?

AGGREGATE INDICATOR The flag is reliable; the WHY behind any individual flag still requires human inquiry. Don't let the dashboard replace the conversation.

The Promise reading

Inverts to a "Composite Flourishing Flag" — same tiered structure, opposite end. A student at the Star tier on this composite has at least two of: ≥95% attendance, ≥10 points above passing in every course, low unexcused rate. Single actionable signal for staff who want to start the day with the good news — the Promise complement of "who do I check on today" is "who do I commend today".

Recognise it

• Drives the daily / weekly recognition list parallel to the at-risk list
• Feeds positive-outreach workflows (mentor invitations, leadership opportunities)
• Pairs with reason-coding for stronger inference (Star + low unexcused = especially robust)
• Triggers reason-appropriate recognition routing (service hours → ministry recognition, academic → tutoring leadership)

Lens: Promise inversion: "Composite Flourishing Flag", same logic, opposite filter

The base tracks rolling attendance windows for the Newly / Persistent / Chronic trend. The full component adds the slope itself.

What it needs

• Daily attendance over time
• Rolling 10, 20, 30-day windows

How it’s calculated

Linear regression slope of attendance rate over the rolling window; flag negative slopes exceeding a threshold.

What it tells you

Catches deteriorating students BEFORE they cross absolute thresholds. A student moving from 98% to 91% over 30 days deserves attention even though they're still "above 90%".

Act on it

• Early outreach before formal at-risk status
• Teacher conversation to identify emerging issues
• Detects recovery trajectories too — useful for affirming students post-intervention

Can you act on it alone?

TRAJECTORY SIGNAL Identifies change, not cause. A declining slope can mean disengagement, illness, family crisis, or schedule change. Useful precisely because it prompts inquiry early.

The Promise reading

PRIMARY PROMISE INDICATOR. Positive trend slope IS the Climbing tier source. A student moving from 85% to 93% over 30 days is the most pedagogically valuable acknowledgement opportunity in the catalogue — the recovery is in progress, and the praise lands while the new pattern is still forming. Promoted from Tier 2 (Risk lens) to Tier 1 (Promise lens) because this KPI surfaces Climbing students directly.

Recognise it

• Teacher conversation within the same week to name the trend explicitly
• Pair with the Climbing tier badge on the Teacher tab — the dashboard does the surfacing, the teacher does the relational acknowledgement
• Cohort-level reporting at Principal / District tab — "schools with 20+ Climbing students this term"
• The single highest-ROI Promise action in the framework — praise IS the intervention

Lens: Higher in Promise — positive slopes are the Climbing signal, more useful than current-level for celebration

What it needs

• Gradebook entries with dates
• Assessment type/weight

How it’s calculated

Slope of grades over time, calculated per student per course; flag declining trajectories.

What it tells you

Identifies students whose current grade looks acceptable but is falling fast. Critical for catching mid-semester disengagement before final grades crystallize.

Act on it

• Conference with student to identify cause
• Check assignment completion concurrently
• Useful for end-of-term predictive flagging

Can you act on it alone?

TRAJECTORY SIGNAL Same caveat as attendance slope. Don't infer cause from slope alone; pair with qualitative inquiry.

The Promise reading

PRIMARY PROMISE INDICATOR. Upward grade slope is the Climbing tier source on the academic side. A student whose grade moves from 55 → 68 → 75 across three assessments is recovering — explicit recognition tied to the trend (not just the current grade) makes the trend more durable. Tier 1 in Promise lens.

Recognise it

• Teacher feedback explicitly referencing the trend ("65 → 72 → 78 last three") — the trajectory itself is the praise object
• Avoid leaving the recovery uncommented — silence is the default when a struggling student improves, and the silence is itself a missed teaching moment
• Pair with attendance trend (KPI #6) for compound recognition — students Climbing on both deserve the most explicit acknowledgement
• Useful for term-end recognition decisions independent of absolute grade level

Lens: Same — upward slope is the strongest Climbing indicator on the academic side

What it needs

• Current grade per course
• Student's full course load

How it’s calculated

Count of courses where grade is below passing threshold, per student.

What it tells you

Distinguishes "struggling in one subject" from "systemic disengagement". A student failing 1 course needs subject support; a student failing 4 needs holistic intervention.

Act on it

• 1 course failing → subject-specific tutoring
• 2–3 courses failing → counselor case management
• 4+ courses → wellness/family intervention, possible schedule review

Can you act on it alone?

RELIABLE INDICATOR High counts reliably indicate systemic issues, though specific cause still requires inquiry.

The Promise reading

Asymmetric KPI. The natural inversion — "number of courses currently passing" — is uninteresting at the student level because most students pass most courses; the count rarely discriminates. Promise reading lives at the OnTrack composite level (KPI #5) rather than per-course count. Not in Promise ranking.

Recognise it

Not at the student level. See KPI #5 for the composite Promise signal.

Lens: Risk-only — no symmetric inversion ("number of courses currently passing" is uninteresting at the per-student level)

What it needs

• Gradebook entries flagged as zero/missing.

How it’s calculated

Count of zero/missing entries per student per course.

What it tells you

Often the SINGLE biggest driver of failing grades. A capable student with 3 zeros can fail a course they would otherwise pass with a B.

Act on it

• Missing work recovery — frequently the highest-ROI intervention possible
• Pattern detection: same student missing assignments across multiple classes vs. one class only
• Teacher-level review when zeros cluster in one section

Can you act on it alone?

DIRECT BEHAVIOURAL MEASURE Doesn't tell you why (forgot, didn't understand, didn't care, didn't have time, didn't have resources at home), but tells you what to fix.

The Promise reading

Inverts to "zero zero-grades" but the signal is weak as standalone celebration — most students don't have many zeros, so the absence of zeros doesn't discriminate. More useful as a component of the Steady tier classification than as an independent recognition signal.

Recognise it

Don't acknowledge "no zeros" directly — it sets the wrong baseline. Use as part of the Steady / Star composite recognition via KPIs #1, #4, and #5.

Lens: Inverts as "zero zero-grades" but the signal is weak as celebration on its own

What it needs

• Attendance data aggregated to student-year level
• Standard definition (typically <90% attendance)

How it’s calculated

% of students missing 10% or more of instructional days.

What it tells you

The standard system-level indicator used in most jurisdictions. Comparable across schools and against external benchmarks. Predicts graduation rates more reliably than test scores.

Act on it

• School-level resource allocation
• District comparison and priority targeting
• Public reporting and accountability

Can you act on it alone?

AGGREGATE RELATIONSHIP Strong group-level predictor of system outcomes. Be careful applying group statistics to individual students (ecological fallacy).

The Promise reading

Inverts to "Steady Attendance Rate" at school / district aggregate — the percentage of students at 95%+ attendance across the year. Board-reportable. A school that has reduced chronic absenteeism from 22% to 14% over three years has produced 8% of its student body as additional Steady-tier candidates — the Promise lens makes that gain legible alongside the Risk-side metric.

Recognise it

• School-level public reporting paired with chronic absenteeism trend
• District benchmarking by Promise growth, not just Risk reduction
• In board reports, both numbers together tell a fuller story than either alone — "down 8 percentage points in chronic absenteeism AND up 11 percentage points in Steady"
• Multi-year cohort recognition (graduating class that maintained 75%+ Steady across all four years)

Lens: Inverts as "Steady Attendance Rate" at school/district — useful aggregate Promise KPI

What it needs

• Attendance with reason codes
• Reason taxonomy

How it’s calculated

Distribution of absences across reason categories at student, course, school, or district level.

What it tells you

A school with 60% unexcused absences has a very different problem than one with 60% medical absences. Shapes the response strategy entirely.

Act on it

• Targets the right intervention type at the right level
• Flu outbreak vs. truancy spike are visible from the data quickly
• Disciplinary-driven absences identify potential behaviour-support resource needs

Can you act on it alone?

DIAGNOSTIC TOOL Reliably differentiates intervention types. Reasons themselves are self-reported / school-coded; data quality of the coding directly affects KPI reliability.

The Promise reading

Weak Promise signal — the Promise reading would be "low unexcused, high school-activity" which isn't well-formed as celebration. Primarily useful in the Risk lens for diagnostic shaping of response strategies. The reason-coded data feeds KPIs #3 and #10, where the Promise inversion lives.

Recognise it

Not directly. Reason-coded data quality (KPI #26) is itself a back-office acknowledgement target.

Lens: Diagnostic both directions but more useful for Risk

What it needs

• Per-student attendance and grades across ALL their courses
• Course/section/teacher identifiers

How it’s calculated

For each student, compute attendance and grade in each course; compare each course to the student's own multi-course average. Aggregate the deltas across each teacher's roster.

What it tells you

If a student is at 95% attendance everywhere except one course, that course is the signal — not the student. Aggregated across many students, persistent negative deltas in a single section suggest a section-level issue.

Act on it

• Coaching and support conversations with teachers (NOT evaluation)
• Course design or scheduling reviews
• Identifying high-performing teachers whose practices could be shared

Can you act on it alone?

CORRELATION, NOT CAUSATION Section differences may reflect teacher effectiveness OR course difficulty, period of day, demographic composition, class size, curriculum, or random variation in small samples. Requires multi-year data and statistical controls before drawing conclusions. NEVER use for personnel decisions without rigorous methodology and union/HR consultation.

The Promise reading

Symmetric — low cross-course variance is itself a Promise signal: the student performs consistently across all courses, suggesting they bring engaged learning behaviour to every classroom rather than performing in some and disengaging in others. Aggregated across many students, persistent positive deltas in one section identify Star teachers whose practice could be shared.

Recognise it

• Identify low-variance students at the counselor level — they're the Steady tier's foundation
• Identify high-performing sections for peer observation and practice-sharing (parallel to Risk-side underperformance review)
• Same governance constraints as Risk side — never use for evaluation or compensation without methodology and union consultation

Lens: Symmetric — high variance means inconsistency, low variance is a Promise signal (the student performs steadily across all courses)

What it needs

• Extracurricular enrollment with start/end dates
• Attendance and grade history before and after enrollment
• Comparison cohort of similar non-participating students

How it’s calculated

Pre/post comparison of attendance and grades around enrollment date; compared to matched non-participants (similar baseline attendance, grade level, demographics).

What it tells you

Strong research base showing extracurricular engagement correlates with academic recovery for previously isolated or at-risk students. Particularly powerful for students with no other connection point to the school.

Act on it

• Targeted invitation of at-risk, low-engagement students to activities
• Identification of which activities show strongest recovery effect for which student profiles
• Resource allocation toward "anchor" activities

Can you act on it alone?

SELECTION BIAS IS CENTRAL Students who join activities differ from those who don't in unmeasured ways (parental encouragement, baseline disposition). Naive pre/post comparisons WILL overstate the effect. Matched comparison or own-trajectory baseline required for credible inference. Useful as hypothesis-generator and program-evaluation tool, not as proof of causation.

The Promise reading

PROMOTED IN PROMISE LENS to Tier 2. Extracurricular engagement is itself a Promise signal — not just a moderator of risk. Students engaged in two or more activities show stronger Steady-tier candidacy than those engaged in only academics. Use as a direct Promise indicator at the student level.

Recognise it

• Activity-coach recognition at term end — naming the student inside the activity context where the engagement happens
• Pair with attendance and grade KPIs for the Star tier — the multi-domain student
• Use for cohort acknowledgement at school assemblies (with the same selection-bias caveat from the Risk side — don't tell students "joining a club caused your improvement")

Lens: Higher in Promise — extracurricular engagement is itself a Promise signal, not just a moderator of risk

What it needs

• Religion course grades and attendance
• School Mass attendance
• Retreat participation
• Service hours (count and depth)
• Ministry/leadership roles

How it’s calculated

Composite participation score across formation programming components. Distinguishes Catholic-specific (sacramental) from universal (Religion class, service) components.

What it tells you

Provides a mission-aligned view of student engagement that purely academic KPIs miss. A student disengaged academically but deeply involved in service or ministry presents a very different picture than one disengaged across the board.

Act on it

• Identifying mission-aligned strengths in academically struggling students
• Chaplaincy outreach to students with no faith-formation participation AND academic risk flags
• Program evaluation for retreats and formation activities

Can you act on it alone?

PARTICIPATION IS NOT FORMATION This measures opportunity and attendance, not interior spiritual life or moral development. A devout student who doesn't sign up for the retreat isn't "less formed". Frame indicators as participation only. Confidential chaplaincy interactions must NEVER be tracked. Diocesan/chaplaincy input on appropriate measures is essential before rollout.

The Promise reading

CRITICAL FRAMING NOTE. A student showing high participation across formation programming (Religion class, service hours, retreat attendance, ministry leadership) is a Promise signal at the participation level only. The Promise lens MUST NOT infer spiritual depth, devotional state, or moral formation from participation rate. A devout student who doesn't sign up for the retreat isn't "less Promise". Frame celebration as opportunity-engagement, not interior life. Chaplaincy and religious-ed leadership consultation required before any Promise-lens deployment of this KPI.

Recognise it

• Recognition tied to specific service acts or leadership roles, not aggregate index rank
• Consider whether any acknowledgement should happen at the dashboard level at all — chaplaincy may prefer all recognition be relational
• Never publish faith-formation Promise rankings, internally or externally
• Same guard rails as the Risk side, applied symmetrically — the prohibition against tracking spiritual depth applies to celebrating it, too

Lens: Mission-aligned. Promise framing must explicitly avoid inferring spiritual depth from participation. See KPI #14 detail card.

What it needs

• Academic indicators (attendance, grades)
• Co-curricular participation
• Faith formation participation
• Service hours
• Anchor-adult indicator (if available)

How it’s calculated

Weighted composite across intellectual, social, spiritual, and service domains. Aligned with Catholic education frameworks (NSBECS, Harvard Human Flourishing Program, ICLE).

What it tells you

Whole-person view consistent with the Catholic educational mission. Identifies students engaged in some domains but not others — often the most actionable profile (the student has at least one foothold to build on).

Act on it

• Strength-based intervention planning
• Annual flourishing review at the school level
• Diocesan benchmarking against peer schools

Can you act on it alone?

COMPOSITE INDEX Inherits all the cautions of its components. Especially sensitive to weighting choices; different weights produce different rankings. Use as a conversation-starter framework rather than a decisive metric.

The Promise reading

PROMOTED to Tier 2 — this IS the Promise concept. The Flourishing Index directly ranks Star-tier candidates by definition: students engaged across multiple domains (academic, social, spiritual, service). Its strength as a Promise indicator is that it surfaces students who would not appear at the top of any single-KPI Promise sort but whose multi-domain engagement makes them the school's most fully-formed students.

Recognise it

• Annual flourishing recognition program at the school level
• Diocesan benchmarking against peer schools (Promise growth, not Risk reduction)
• Principal use at term-end for naming Star-tier students
• Pair with KPI #14 with the same NEVER-infer-depth guard rail

Lens: Higher in Promise — flourishing is itself the Promise concept; the index ranks Star candidates directly

What it needs

• Attendance with day-of-week and period stamps.

How it’s calculated

Per-student absence frequency broken down by day of week and class period; flag patterns deviating significantly from baseline (e.g., 80% of absences fall on Mondays or in period 6).

What it tells you

Patterns reveal causes: Monday/Friday absences often signal weekend issues; period-specific absences signal class avoidance; consistent late arrival signals transportation or morning routine issues.

Act on it

• Schedule review for transportation issues
• Conversation with the specific teacher for class-avoidance signals
• Family check-in for weekend pattern absences

Can you act on it alone?

PATTERNS PROMPT INQUIRY A Friday pattern doesn't prove the cause; it suggests where to ask. Period-avoidance patterns are more diagnostic than day-of-week patterns.

The Promise reading

Risk-only. Pattern detection identifies anomalous patterns by definition; there isn't a useful "positive pattern" reading — a student who attends consistently is just attending, not exhibiting a pattern. Not in Promise ranking.

Recognise it

Not applicable.

Lens: Risk-only — pattern detection has no symmetric Promise reading

What it needs

• Daily attendance
• Shared course/schedule data (implicit peer network)

How it’s calculated

Identify groups of students whose absence days co-occur at rates significantly higher than chance, weighted by schedule overlap. Cluster detection on the resulting graph.

What it tells you

Detects social contagion in disengagement — groups of students whose attendance patterns move together. Useful for understanding that an individual's decline may be socially driven, not personal.

Act on it

• Group-level intervention rather than individual-only
• Counselor awareness of peer dynamics
• Activities that intentionally remix peer groups

Can you act on it alone?

CORRELATION ≠ INFLUENCE Co-absence may reflect shared friend group, shared external circumstance (same neighbourhood, same bus, same after-school job), or coincidence. Cannot conclude one student is "influencing" another. GOVERNANCE CONCERN: peer-network analysis risks labelling students by association. Use only for counselor-level intervention awareness, never in dashboards visible to students/parents or in records.

The Promise reading

Risk-only. The Promise inversion — "peer co-attendance" — is just attendance and doesn't add information beyond KPI #1. Network-effect analysis remains Risk-focused. Not in Promise ranking.

Recognise it

Not applicable.

Lens: Risk-only

What it needs

• Peer co-occurrence score (isolation indicator)
• Extracurricular enrollment timing
• Pre/post attendance and grade trajectories

How it’s calculated

Identify low-peer-correlation students (isolation proxy); track outcomes for those who subsequently join an activity vs. matched non-joiners.

What it tells you

Tests the hypothesis that activity engagement supports isolated students. Operationally identifies which students might benefit most from targeted outreach, and which activities have strongest recovery signal.

Act on it

• Targeted activity invitation program
• Activity-coach awareness when an isolated student joins
• Program evaluation: which activities work best for whom

Can you act on it alone?

MULTI-LAYER CAUTION Inherits selection-bias issues from extracurricular KPI AND inference issues from co-occurrence KPI. Strong as a program-design tool, weak as an individual-level predictor. Effects measurable only with matched comparison or randomized invitation studies.

The Promise reading

Symmetric. Students who joined an activity AND showed measurable improvement post-onboarding are the Climbing-via-engagement profile — pedagogically the most powerful Promise signal when the data supports it. The student found a foothold. Same selection-bias caution as the Risk lens applies.

Recognise it

• Activity-coach acknowledgement during the year, not just at exit
• Program-evaluation lens — which activities reliably produce Climbing students
• Never use as marketing for the activity without rigorous matched-comparison evidence
• The student's own reflection on what changed is more valuable than the dashboard's classification

Lens: Symmetric — "activity onboarding success" is a Promise signal, but causally weak in both directions

What it needs

• Within-student cross-course variance (KPI #12)
• Multi-year section-level outcomes
• Course/period/demographic controls

How it’s calculated

Multi-year aggregation of section-level deviations, controlling for course, period-of-day, class size, and student demographic mix. Variance-decomposition approach.

What it tells you

When done rigorously, can identify sections persistently underperforming or outperforming expected outcomes. Most defensibly used to identify high-performing teachers for practice-sharing.

Act on it

• Coaching support for teachers in underperforming sections
• Practice-sharing from high-performing sections
• Curriculum or scheduling review

Can you act on it alone?

HEAVY GOVERNANCE BURDEN Causation cannot be established from observational data alone. Confounds (course difficulty, demographic loading, period effects, sample size) routinely produce false signals. Requires statistical methodology beyond simple averages. In most unionized districts, teacher-effect analytics require explicit union consultation and clear use-of-data agreements BEFORE rollout. Never use for personnel decisions without rigorous methodology.

The Promise reading

Symmetric. Persistent positive deviation in a section identifies the high-performing teacher. With multi-year data and demographic controls, this can credibly support practice-sharing — observable classroom techniques worth documenting and spreading. Same heavy methodology + governance burden as the Risk lens.

Recognise it

• Coaching support FROM the high-performing section to others (not just from administration TO the underperforming)
• Practice-documentation projects rather than awards — the documented practice outlives the recognition cycle
• Never use for performance bonuses or differential compensation without explicit union consultation and methodology
• Treat positive deviation with the same statistical skepticism as negative deviation — both can be sample-size artifacts

Lens: Symmetric — section deviation cuts both ways, but governance burden remains

What it needs

• Intervention records (date, type, target student)
• Pre/post attendance and grade trajectories
• Comparison to similar non-intervened students

How it’s calculated

For each documented intervention, compare trajectory before and after; aggregate by intervention type and student-risk-profile to identify which interventions work for which students.

What it tells you

Closes the loop. Without this KPI, the entire system generates risk flags but never learns whether responses work. Critical for resource allocation and continuous improvement.

Act on it

• Scale up effective interventions
• Discontinue or redesign ineffective ones
• Match intervention type to student-risk profile

Can you act on it alone?

REGRESSION-TO-MEAN IS A MAJOR CONFOUND At-risk students often improve somewhat regardless of intervention, simply because they were flagged at a low point. Effective causal inference requires matched comparisons. Useful for program improvement even at correlation level, but be cautious about declaring specific interventions "proven".

The Promise reading

Symmetric. Documents which interventions reliably produce Climbing-tier students. The Promise lens here is the same as the Risk lens — does the intervention work? — but framed as "did the student improve" rather than "did the risk subside". Regression-to-mean caution applies in both directions.

Recognise it

• Program evaluation and scale-up decisions
• Recognition for intervention designers (counselors, social workers, coaches) whose programs show consistent Climbing outcomes
• Pair with KPI #18 for the activity-driven recovery subset
• The Promise framing makes funding conversations easier — "this program produced 30 Climbing students last year" is more grant-friendly than "this program reduced risk"

Lens: Symmetric — effective interventions are both Risk reduction and Promise creation

What it needs

• Family/household identifier linking siblings
• Attendance and grade data for sibling sets

How it’s calculated

Within-family correlation of attendance patterns and grade trajectories.

What it tells you

When siblings show simultaneous absence patterns or declines, the cause is likely at the family/household level rather than school level. Reframes the intervention.

Act on it

• Social work or family-services referral
• Coordinated outreach across siblings rather than independent per-student interventions
• Recognition that the school-level levers may not be the right ones

Can you act on it alone?

FAMILY-LEVEL SIGNAL Generally reliable for identifying family-level causes when present, but absence of pattern doesn't rule out family causes (siblings may respond differently to same circumstance). DATA PRIVACY: linking sibling records requires careful governance.

The Promise reading

Symmetric. Family-level positive patterns (sibling sets all in Steady tier, or all Climbing simultaneously) suggest family-level protective factors worth understanding. As with the Risk side, the family is the unit of intervention — and the unit of acknowledgement when the news is good.

Recognise it

• Family-level recognition when appropriate (parent-teacher conferences, family awards programs)
• Careful about implying family causality from sibling patterns — same data-privacy and governance caution as the Risk lens
• Catholic-district context: family-as-unit recognition aligns naturally with parish involvement and sacramental life, but the same NEVER-infer-depth caution applies (KPI #14)

Lens: Symmetric

What it needs

• Tardy records per student per course.

How it’s calculated

Count and rate of tardies per student per course.

What it tells you

Often signals transportation issues, morning routine problems, or class-specific avoidance (especially first-period). Less predictive than absences but useful diagnostically.

Act on it

• Transportation outreach
• Family routine support
• Distinguishing first-period tardies from mid-day tardies

Can you act on it alone?

WEAK INDIVIDUAL PREDICTOR But useful in combination with other indicators.

The Promise reading

Symmetric, weak. Low tardy rate is part of the Steady habit pattern but isn't strong enough as a standalone Promise signal. Useful as a component of the Steady composite, not as an independent recognition trigger.

Recognise it

Don't acknowledge "on time" directly — it sets the wrong baseline (punctuality as a low bar rather than a learned habit). Use as part of the Steady composite recognition (KPIs #1, #4, #5).

Lens: Symmetric

What it needs

• Gradebook entries with assessment-type tags (formative/summative/participation).

How it’s calculated

Per-student variance between formative and summative grades, or participation and assessment grades.

What it tells you

Large gap between participation grades and assessment grades may indicate learning gaps masked by effort. Reverse pattern (strong assessments, weak participation) suggests engagement issues with the class structure itself.

Act on it

• Differentiated support based on the pattern
• Assessment design review

Can you act on it alone?

DIAGNOSTIC Identifies patterns worth exploring; doesn't explain them.

The Promise reading

Symmetric. Low variance between formative and summative grades means consistent demonstration of learning, not just task completion. A Promise signal that distinguishes "engaged learner" from "task completer". Use as a Steady-tier component.

Recognise it

• Teacher-level recognition for the consistent-demonstration profile
• Useful diagnostic for assessment design when variance is low across many students (the assessments may be measuring the same thing twice)
• Don't acknowledge directly at the student level — the variance metric is too granular for student-facing language

Lens: Symmetric

What it needs

• School-wide attendance by period.

How it’s calculated

School-wide attendance rate per period; identify periods with significantly lower attendance.

What it tells you

Last-period and first-period attendance dips are widespread. The size of the dip is the diagnostic — large dips suggest schedule or culture issues.

Act on it

• Schedule design adjustments
• First-period or last-period programming review

Can you act on it alone?

STRUCTURAL EFFECT Well-documented in education research. Safe to act on at aggregate level.

The Promise reading

Symmetric. Sections with high attendance regardless of period-of-day (especially first and last period) are anomalously engaging — worth understanding what teachers or course designs make them attendance-resilient. Promise reading at the section / teacher level, not the student level.

Recognise it

• Identify and document practices from period-resilient sections
• Share at department or PD level
• Pair with KPI #19 for the teacher-effect signal — and with KPI #19's same governance constraints

Lens: Symmetric

What it needs

• Multi-year aggregated KPIs.

How it’s calculated

Same KPIs compared across reporting periods year-over-year, controlling for population changes.

What it tells you

Trend information is more informative than snapshots. A school "below district average" may be improving fast; one "at district average" may be declining.

Act on it

• Board reporting
• Strategic planning
• Resource allocation review

Can you act on it alone?

TREND CONTEXT Trends reflect many factors (policy, demographics, world events); attributing year-over-year change to any single cause requires careful analysis.

The Promise reading

Symmetric — trend information is as important on the Promise side as the Risk side. A school whose Steady-tier fraction is growing year-over-year is producing improvement at the cohort level; a school flat on Risk metrics but growing on Promise metrics is shifting the distribution in the right direction even if averages haven't moved.

Recognise it

• Board reporting alongside Risk reduction — both numbers tell a fuller story than either alone
• Multi-year cohort celebration (graduating class that maintained 60%+ Steady all four years)
• Diocesan or district benchmarking on Promise growth, not just Risk reduction
• Useful framing for grant and budget conversations — "we produced N additional Steady students" is more grant-friendly than "we reduced our at-risk count"

Lens: Symmetric — applies to both Risk reduction and Promise growth

What it needs

• Attendance records with and without reason codes.

How it’s calculated

% of absences with a coded reason vs. uncoded/unknown.

What it tells you

Every reason-based KPI above depends on this. If 40% of absences are uncoded, reason-mix KPIs are unreliable. This is the foundation that holds up everything else.

Act on it

• Front-office training
• Process review
• System defaults review

Can you act on it alone?

META-INDICATOR Doesn't measure student outcomes but determines reliability of indicators that do.

The Promise reading

Risk / operational only. Data quality enables every other KPI in both lenses but isn't itself a Promise indicator at the student level. School-level data-completeness rate could feed a "data-health" badge on the Pipeline (IT) surface, but that's an operational acknowledgement story, not a student one.

Recognise it

Not at the student level. See the Pipeline (IT) surface for operational data-quality acknowledgement (front-office training, coding-discipline recognition).

Lens: Risk / operational only