Types of Assessment Tools: When to Use Each One

Assessment types fall into two questions a teacher answers separately: when in the unit you assess, and what you assess with.

The "when" group covers diagnostic, formative, benchmark, summative, and ipsative assessment, plus the norm-referenced and criterion-referenced choice.

The "what with" group covers the tools that collect the evidence, including quizzes and tests, rubrics, and portfolios.

Below are all ten types, when each one fits, and a quick way to choose between them for your next unit.

What assessment tools are

An assessment tool is anything a teacher uses to find out what a student knows or can do. The assessment is the event itself, such as a quiz on a Friday or an end-of-term paper. The assessment tool, on the other hand, is what you use to run the assessment and score it, like the quiz itself or the rubric you grade the essay against.

Assessment serves three different purposes, and most teachers learn them in training:

• Assessment of learning measures what students have learned after the teaching is completed.
• Assessment for learning takes place during the teaching process and guides the teacher on what to teach next.
• Assessment as learning lets students judge their own academic progress.

The same tool serves any of the three purposes. What changes is what you do with the result.

The ten types at a glance

Ten assessment types cover almost everything a K-12 teacher uses, split by when you assess and what you assess with.

Seven of the ten are about timing and what a score gets measured against. The other three are the instruments that collect the evidence. You will usually match one from each group, such as a summative assessment scored with a rubric.

The table gives a quick overview of all ten before the details.

When you assess: seven types

Seven assessment types are defined by where they sit in the teaching sequence.

The last two on this list are about what a score gets measured against, but they belong here because you decide on them at the same planning stage.

1. Diagnostic assessment

A diagnostic assessment tests the skills students have before any teaching takes place.

Give one before a unit or school year starts, when you need to know where to start teaching and which students need extra help.

For example, an English teacher starts a persuasive writing unit by asking students to write a short argument for a later start to the school day.

Reading those paragraphs shows the teacher which students can already make a claim and back it with evidence. The limitation of this method is effort. Students know it doesn't count, so some won't try, and a weak result can mean "didn't bother" as easily as "doesn't know."

2. Formative assessment

Formative assessment tells you whether your teaching is landing while the unit is still running.

Use it often, at any point in a lesson where the result can change your plan for the next day.

A three-question exit ticket after a lesson on thesis statements is a typical example. Students write a thesis statement on an assigned topic, and you sort the slips into three piles before they leave.

A limitation of this method is follow-through. Formative checks only pay off if you act on what they show, and with five classes a day, reading every slip is a real time cost.

3. Benchmark and interim assessment

Benchmark or interim assessments measure student progress against a standard at set points during the year.

Districts usually run them two or three times a year across all their schools, to see whether students are on track for the year-end exam.

One common approach gives the same writing prompt in fall, winter, and spring and scores all three with one rubric, so the three scores can be compared.

The drawback here is timing and specificity. Results take days or even weeks to come back, and they're usually too broad to plan next Tuesday's lesson from.

4. Summative assessment

Summative assessment measures what students have learned at the end of the teaching process.

Use it at the end of a unit, term, or course, when you need a result for the gradebook or report card.

A five-paragraph argumentative essay at the end of a persuasive writing unit, scored against the unit rubric, is a typical example.

Its limitation is timing: it comes too late. By the time you know half the class struggled with counterarguments, you've already moved on.

5. Ipsative assessment

Ipsative assessment compares a student's work with their own previous performance, not with other students' work.

Use it across drafts or when students repeat a task. Every writing teacher does this when comparing a second draft with the first and seeing which rubric cells have changed, for example when a student raises their evidence score from 2 to 3.

The disadvantage is that progress does not equal meeting the standard. A student can improve a lot and still be below grade level, so it needs pairing with a criterion-referenced check.

6. Norm-referenced assessment

Norm-referenced assessment shows how a student performs compared with a larger group of students. Standardized tests with national norms in reading and mathematics use it most.

A reading score at the 72nd percentile is a norm-referenced result: the student scored higher than 72 percent of the comparison group.

The weakness of a percentile is that it shows where a student ranks, not what the student can do.

7. Criterion-referenced assessment

Criterion-referenced assessment measures whether a student meets a standard, regardless of what others in the class achieve.

Teachers use it wherever there is a fixed standard, from a classroom quiz to an end-of-year state test.

A classroom example is a quiz on commas after introductory clauses, where a student passes by answering 8 out of 10 questions correctly.

The limitation is that the result is only as good as the standard and the cut score. If the bar for "meets the standard" is too low or vague, a pass doesn't mean much.

What you assess with: three tools

Three tools do most of the work, and each suits a different kind of evidence.

Quizzes and tests

A quiz or test is a set of questions students answer within a set time, in formats such as multiple choice, short answer, or essay questions.

You can use them anywhere in a unit, from a quick formative check to a full summative exam.

A typical example is a 15-item multiple-choice quiz on cell organelles followed by two short-answer questions.

The main disadvantage of tests and quizzes is the lack of depth. A multiple-choice item can't show how a student builds an argument or writes an explanation, and some right answers are lucky guesses.

Rubrics

A rubric is a scoring guide that defines the criteria of a task and describes each level of achievement. Use one when you are judging the quality of a piece of work, such as an essay, a report, a project, or a presentation, rather than whether an answer is right.

For instance, a four-row argumentative essay rubric, covering claim, evidence, reasoning, and conventions, each with four levels, is a standard example.

A rubric can serve six of the seven "when" types. The exception is norm-referenced assessment, because a rubric scores work against criteria, not against other students.

Wording is the main limitation. Vague criteria like "uses evidence well" lead different teachers to different ratings, and writing a good rubric takes time. If you'd rather start from one that's already been tested, browse the rubric library first.

Portfolios

A portfolio is a collection of a student's work over time, with the student's own reflections on each piece.

Portfolios work best across a term or a full year, when you need to show growth.

A writing portfolio typically holds a personal narrative from September, an argument essay from January, and a research paper from May, each with a short note on what the student would change.

A portfolio shows what no single test can: how the work changed over time.

The drawback is time. Collecting, organizing, and scoring a portfolio is slow, and without a clear rubric, the scoring can drift from one student to the next.

Where peer and self-assessment fit

Peer and self-assessment are not separate types, they are ways of running one. Both are assessment as learning, because the student evaluates and learns at the same time.

Self-assessment helps students see the gap between their work and the target, for example by checking their essay against every rubric criterion before they submit it.

Peer assessment gives a student an extra reader, and reviewing someone else's writing helps them spot mistakes they miss in their own.

Both come with two practical catches. First, students need a rubric to do either well. Without one, self-assessment becomes "I think it's good," and peer assessment becomes "nice job."

Second, students need to be taught how to assess. Model the marking of one essay with the whole class and show how to write a useful comment before you ask students to assess each other's work.

How to choose

The right assessment depends on what you need to find out and what you will do with the answer. Ask yourself these five questions, and the type you need becomes clear quickly.

• What will you do with the result? If nothing in your lessons will change based on the result, you need a summative assessment. If the result can change tomorrow's lesson, you need a formative one.
• Will you compare students or measure against a standard? That is the choice between norm-referenced and criterion-referenced assessment. Criterion-referenced is usually more useful for classroom decisions, since knowing a student ranks as average tells you nothing about what to teach next.
• How often can you realistically mark it? Formative assessment only works when it is cheap to run. A daily check that takes an hour to grade will not last past week two, so keep it quick enough to sort.
• Does it need to show progress? Portfolios and ipsative comparisons show progress. A single summative test cannot, since it captures one day of a student's performance.
• Is it going on a report? Then it needs criteria that families can read. A result of 14 out of 20 is hard to explain to a parent. "Meets the standard for using evidence" is not.

How that result gets reported depends on your school's grading approach, and the guide to types of grading systems covers the options.

Where AI fits, and where it does not

AI can score against criteria you define, and it cannot decide what is worth assessing. That split is what makes it useful in an assessment plan.

AI grading handles the repetitive part of marking. AI grading applies the same rubric to every paper, so the thirtieth essay gets scored against the same criteria as the first. That consistency is the main benefit.

AI grading writes specific feedback for each rubric row, faster than a teacher can, which frees time for planning and conferencing with students.

What AI grading cannot do is decide what counts. The teacher chooses the criteria an assessment uses.

AI grading doesn't judge the originality of a student's ideas either, and it doesn't assess a class discussion, a science practical, or an oral presentation. Those all need a teacher in the room.

EssayGrader, a rubric-based AI essay grading platform for K-12 and higher-education teachers, covers the first half of this split for writing assignments. It grades each essay against the rubric criteria you define and reports mastery against writing standards separately from the assignment grade, so one pass gives you both results.

You still define the criteria yourself, and if you don't have the rubric you need, you can build a custom rubric first.