Formative checks tell you where learning is going; summative assessments tell you where it landed. When it’s time for the unit test, the midterm, or the benchmark that actually counts, you need a build process that’s a little more deliberate than throwing ten questions into a quiz. This guide shows you how to create a summative assessment in pear assessment — the platform that grew out of Edulastic — from planning the blueprint to choosing scoring settings that hold up when grades go in the book.
The workflow below assumes you’ve made quick quizzes before and now want the full, careful version: standards-aligned items, sensible points, sections for a longer test, and settings that match the stakes. If you’re brand new to the platform, skim our overview of how to use Pear Assessment first, then come back here for the summative-specific decisions.
What makes an assessment “summative” in practice
A summative assessment evaluates learning at the end of an instructional stretch — a unit, a term, a course. Three practical consequences follow for how you build it. First, coverage matters more than speed: the test should sample the whole unit’s standards, not just the last week’s lessons. Second, the results feed grades and reports that other people read — students, parents, department chairs — so scoring rules and answer keys need to be airtight before anyone sits down to take it. Third, security and consistency settings (shuffling, navigation, time limits) matter more than they do on a Tuesday exit ticket, because the results carry weight.
None of that requires a different tool. It requires using the same builder with more care at each step, which is exactly what the following sections cover.

Step 1: Start with a blueprint, not the item bank
Resist opening the item bank first. On a sticky note or a doc, list the standards or topics the test must cover and roughly how many points each deserves based on how much class time it got. A unit that spent two weeks on fractions and two days on decimals should not have a fifty-fifty test. This blueprint becomes your checklist while you build, and it’s the difference between a test that measures the unit and a test that measures whatever the item bank happened to surface first.
Most teachers land between 20 and 40 items for a full-period summative test, mixing formats: auto-graded multiple choice and technology-enhanced items for breadth, plus a small number of constructed responses for depth. The item bank makes the breadth easy — it’s large and standards-tagged — but the blueprint keeps it honest.
Step 2: Build the test from the item bank
With your blueprint in hand, create a new assessment and start adding items. Search and filter the Pear Assessment item bank by grade, subject, and standard so every blueprint line gets its items. A few habits separate a professional build from a rushed one:
- Favor certified and vetted items for high-stakes questions; community items are great for practice but vary in polish.
- Preview every item as a student. Items that read fine in the editor sometimes behave differently in the test player — passages split oddly, drag targets shrink, an equation renders small.
- Watch difficulty balance. A test of all medium items tells you little. Include a few accessible openers, a solid core, and one or two genuine stretch questions.
- Tag standards as you go. Retrofitting tags later is nobody’s idea of fun, and untagged items are invisible to your standards reports afterward. Our rundown of the Pear Assessment question types helps you pick formats that measure what you actually taught.
As you add items, keep a running tally against the blueprint. When every line has its questions and the total points look right, you’re ready for the settings that make it summative.

Step 3: Set scoring like you mean it
Scoring decisions made casually here become grade disputes later. Walk through them deliberately:
Points per item. Weight items by importance and difficulty, not by habit. If everything is worth one point, your hardest standard counts the same as a vocabulary recall item. Partial credit matters too: for multi-part and technology-enhanced items, decide whether a partially correct response earns partial points or none, and be consistent across the whole test.
Answer keys. Open every item and verify the key yourself — don’t trust that a bank item’s key matches your curriculum’s conventions, especially for “select all that apply” and ordering questions. This ten-minute review is the highest-value step in the entire build.
Rubrics for constructed responses. For essays and short answers, attach a rubric at build time rather than inventing criteria mid-grading. You’ll grade faster, more consistently, and with a defensible record if a grade is questioned.
Step 4: Apply summative-appropriate settings
Settings are where a quiz becomes a test. For most summative builds, teachers enable question and answer shuffling so neighbors get different versions, restrict backward navigation if the exam format demands it, and set a time limit that matches the class period with a sensible buffer. Decide whether calculators, reference sheets, or read-aloud supports are allowed and configure them now — not in response to a hand raised five minutes into the exam.
Also decide the release plan before you assign: will students see scores immediately on submission, after you review, or not until you release them manually? For summative work, manual release after a quick key-check is the safest pattern, because it protects you from the one scenario every teacher dreads — a bad key grading thirty tests before anyone notices.
can i build one long pear assessment test with multiple sections
Yes — and for midterms, finals, and benchmarks, you probably should. Pear Assessment lets you organize a single test into sections, so one long assessment can hold a multiple-choice section, a reading section built around a passage, and a constructed-response section, all under one assignment with one score. Students experience it as one test; you get the structure of several.
Build sections deliberately. Group items by format or by topic so instructions can be written once per section instead of repeated on every item (“Questions 1–10: choose the best answer”). Sections also let you pace a long test: a reading section with its own passage, a math section where a calculator is appropriate, a writing section at the end. If your test has parts students should tackle in order, arrange the sections in that order and use navigation settings to keep the flow.
One caution: longer is not automatically better. A test that outruns the class period measures endurance, not mastery. If your blueprint genuinely needs more items than one sitting allows, split it into two assessments on two days rather than building a marathon with sections — your data will be cleaner and your students will thank you.

Step 5: Assign, monitor, and grade
Assign the finished test to your classes with open and due dates that reflect your real testing window, including a make-up plan. During the session, the live view lets you watch progress and spot a student who’s stuck or disconnected early enough to help. Afterward, auto-graded items score themselves; your job is the constructed responses and a sanity pass over the results. Open the item analysis before releasing grades — if one question shows a bizarre pattern (everyone picked the same wrong answer), check the key one more time before those scores become official.
Common summative build mistakes (and how to avoid them)
Teachers who build summative tests under time pressure tend to make the same handful of mistakes. Watch for these:
Letting the item bank set the curriculum. If you pick items by browsing rather than by blueprint, the test quietly drifts toward whatever has the most bank coverage. Ten extra items on a minor topic don’t make the test better; they make it a test of the minor topic. Build from the blueprint, and let gaps in the bank become items you write yourself.
Skipping the student preview. Editor view lies by omission. Passages, drag-and-drop items, graphing tools, and equation editors can all behave differently at student screen sizes. Previewing the whole test as a student takes five minutes and catches the problems that otherwise surface as thirty raised hands.
One-point-everything scoring. Uniform points are easy, but they flatten your data. When a foundational skill and an enrichment stretch item count identically, the final grade stops meaning anything specific. Weight by importance; your future report-reading self will notice the difference.
Releasing scores before reviewing the item analysis. Auto-grading is fast, which is precisely the danger. A wrong key on a bank item — rare, but real — can mark a whole class wrong in seconds. A two-minute scan of the results for impossible patterns (a question literally everyone missed, a top student scoring 40%) before release is cheap insurance.
Building the entire exam the night before. Summative builds reward a second session with fresh eyes, especially for keys and rubrics. If the calendar allows it, build on Monday, review on Wednesday, assign on Friday. The test your students get is noticeably better for the gap.

Frequently asked questions
How is a summative build different from a regular quiz in Pear Assessment?
The builder is the same. The difference is care: full standards coverage from a blueprint, verified keys, weighted scoring, tighter settings, and manual score release.
Can I reuse a formative quiz as a summative test?
You can clone and expand it, but audit it first — formative quizzes often include hint-heavy or collaborative items that don’t belong on a graded exam.
How long should a summative test be?
As long as your blueprint requires and no longer. For most classes, a full period supports roughly 25–40 well-chosen items plus a constructed response or two. If students routinely can’t finish, the test is measuring speed; if everyone finishes in fifteen minutes, it’s probably not measuring much at all.
Should a summative test include questions students have never seen?
The items can be new, but the skills shouldn’t be. Summative assessment is fairest when formats and expectations mirror what students practiced formatively — new surface, familiar demands. Surprise belongs in engagement hooks, not in graded exams.
Where do I learn more about scoring and reporting features?
The official Pear Assessment documentation covers the current settings and reporting tools in depth.
The bottom line
Creating a summative assessment in Pear Assessment is the regular builder used with professional discipline: blueprint first, item bank second, keys verified, scoring weighted, sections for long tests, and scores released only after a human has glanced at the results. Do those things and the grade book numbers you produce will be numbers you can defend — to students, to parents, and to yourself.

