---
name: summative-assessment
description: Use when asked to design an end-of-unit or end-of-course assessment (test, final project, exam) that evaluates whether mastery was reached, as distinct from [[formative-assessment]], which is an ongoing, low-stakes check during instruction meant to inform teaching rather than assign a grade.
---

# Summative Assessment

A summative assessment evaluates, at the end of a unit or course,
whether students reached the mastery the instruction was aiming for —
a test, final project, or exam that produces a grade. It's the
counterpart to [[formative-assessment]], which checks understanding
*during* instruction specifically to inform what happens next in
teaching; a summative assessment happens after the teaching is done
and measures the result rather than shaping it.

## Key components

- **Objective coverage** — every learning objective the unit or course
  was actually built around should be assessed somewhere, so the grade
  reflects what was taught rather than a subset of it or content that
  was never actually the point.
- **Format** — whether the assessment is a written test, a
  demonstration, a project, or a mix, chosen for fit with what's being
  measured rather than defaulting to whatever format is easiest to
  grade.
- **Scoring instrument** — a rubric or answer key prepared before the
  assessment is given, not improvised while grading, so the standard
  doesn't shift between the first and last paper graded — see
  [[rubric-design]].
- **Weighting relative to formative work** — how much the summative
  result counts toward the final grade relative to the formative
  checks along the way, decided as policy rather than case by case.
- **Retake or resubmission policy** — whether and how a student who
  falls short gets a second opportunity to demonstrate mastery, decided
  before the assessment rather than negotiated after a low score.

## Assessing the objective, not the unit's content

The most common gap between a summative assessment and the unit it's
meant to evaluate is that the assessment ends up testing what was
easiest to write questions about, rather than what the unit's actual
learning objectives were. A unit built around a specific skill — say,
constructing an argument with evidence — can end up assessed almost
entirely with recall questions about the topic's factual content,
because factual recall is easier to write test items for than a
direct measure of the skill itself. The result is a test that's
straightforward to grade but doesn't actually tell anyone whether the
unit's real objective was met; a student can pass it without ever
having demonstrated the skill the unit was designed to build.

Designing backward from the objective — the same discipline
[[curriculum-design]] uses at the unit level — means writing assessment
items that require the actual skill or understanding the objective
names, even when that's harder to write and score than a set of
recall questions on the surrounding content. If the objective is a
skill, the assessment needs a task that requires performing it, not
just a set of questions about the topic it was practiced on.

## Common pitfalls

- **Tests content, not the objective** — items that are easy to write
  and grade but measure recall of surrounding facts rather than the
  actual skill or understanding the unit targeted.
- **Rubric or key written after grading starts** — improvising the
  scoring standard while grading lets the first few papers set a
  different bar than the last few.
- **No connection to formative data** — a summative result that
  contradicts what formative checks showed throughout the unit, with
  no process to reconcile the discrepancy, leaves it unclear which
  signal to trust.
- **All stakes concentrated in one assessment** — a single high-stakes
  summative event with no retake path turns one bad day into a
  disproportionate grade outcome unrelated to actual mastery.
- **Coverage gaps** — objectives that were taught but never actually
  assessed, so the final grade is silent on whether they were learned.

## Learn more

- [[formative-assessment]] for the ongoing checks during instruction
  a summative assessment is not a substitute for.
- [[rubric-design]] for building the scoring instrument a summative
  assessment is graded against.
- [[unit-plan]] for the instructional sequence a summative assessment
  evaluates at its end.