Feedback is among the most powerful influences on student achievement.John Hattie
You can use the list below as your checklist for the module. It's best to finish each item before you move into Module 4.
The Videos tab has walkthroughs for all four chapters (one playlist covers Marzano 1 and 2, and Seifert 11 and 12 each have their own video), a talk on writing proficiency scales, and two outside talks (Wiliam on formative assessment, Wormeli on redos and retakes).
You can use these buttons to track where you are in Module 3. Mark it In progress when you open the readings and Complete after you submit the assignments and the exit ticket. Your choice is saved in this browser only.
Before you grade the work you have to know what the work was supposed to show. In Module 3 you'll name the target, build an instrument to detect it, and read the data the instrument produces without misreading it. You wrote the lesson in Module 2, and in Module 3 you'll find out whether the lesson worked.
In Marzano Chapter 1 you'll study proficiency scales: leveled descriptions of one learning target written so a student can read them, restate them, and locate the work against them. In Chapter 2, Marzano describes four assessment types (informal, formal, common, student-generated) and asks you to pick one by the question you're asking rather than by the test bank packaged with the textbook. Seifert & Sutton Chapter 11 covers the technical floor: reliability, validity, fairness, item formats, the table of specifications, and rubric design. In Chapter 12, Seifert & Sutton build on the same floor and show you how to read a standardized score report without making claims the report doesn't support.
In the Assessment Plan you cover a whole unit: formative, summative, and diagnostic measures, one rubric, an alignment table, and the grading philosophy behind them. The Questioning Strategy Blueprint covers the day-to-day half: you'll write a sequenced set of Bloom's-labeled questions for one lesson and name your equity-of-participation moves. In the Summative Assessment Design you build one finished instrument with a rubric, a table of specifications, accommodations, and a reflective analysis of the design choices. By the end of Module 3 you should be able to point at a piece of student work and explain what claim it supports, what claim it doesn't support, and what the next instructional move is.
If you can't describe what a 2, 3, and 4 look like in plain student-facing language, you don't yet have a goal you can grade. In Chapter 1, Marzano shows you how to turn vague targets into proficiency scales a student can self-locate against.
Informal, formal, common, and student-generated assessments each give you a different kind of evidence. The right choice is the assessment that answers your question, whatever happens to be in the test bank.
These three qualities tell you whether a score deserves the weight you give it. Reliability means consistency, validity means the use of the score is justified, and fairness means the instrument doesn't disadvantage some students by its design. You can check an assessment for all three qualities before you give it to students.
A percentile tells you how a student ranked against a norm group, and a grade equivalent doesn't mean the child can do work at the grade level. The standard error of measurement tells you how far the true score could be from the reported one, so you can't treat it as a rounding detail. You have to understand what each number on a score report means before you can use the report to make decisions about a student.
You'll find everything for Module 3 on this page, organized into the eight tabs at the top. You can work through them in order, which follows the logic of the chapters, or jump to whichever tab you need. When you click a card below, you'll go straight to the tab.
On this tab you'll find the at-a-glance checklist, the four core ideas, the guiding questions, the seven learning objectives, the four required readings, the theorist row, the twelve key terms, and Part 7 with the three projects, the exit ticket, and the two discussions.
The tab on Providing and Communicating Clear Learning Goals covers proficiency scales, rubrics, status vs. growth, the three general types of assessment in Marzano's table 1.2, and prioritized standards.
The tab on Using Assessments covers informal voting and response-board techniques, formal assessment formats, common assessments built across a department, student-generated assessment, and the assessment-as-conversation frame.
The tab on Teacher-Made Assessment Strategies covers reliability, validity, fairness, item formats and where each one breaks, table of specifications, performance assessment, and rubric design.
The tab on Standardized and Other Formal Assessments covers norm-referenced and criterion-referenced reporting, score types and their misreadings (percentile, grade equivalent, scaled, performance level), and how to translate a score report for a parent conference.
The tab has three chapter walkthroughs (one playlist for Marzano 1 and 2, one video each for Seifert 11 and 12), a talk on writing proficiency scales, and two outside talks on formative assessment (Wiliam) and grading practices (Wormeli). You can watch each walkthrough alongside its chapter reading.
You can work through three branching scenarios on assessment and grading decisions, a classify quiz on assessment purpose, a sort by reliability/validity/fairness, and a vocabulary cloze. You'll see each choice play out, and you can replay any scenario to try a different path.
This tab has the fill-in-the-blank self-check, a theory-match classify quiz, four consolidation reveals, the Final Pass flash deck, and the Bonus Content "Going Deeper" resources. Use it for a last review before you mark Module 3 complete on the hub.
You can carry the questions below with you as you work through the chapters and activities. You'll see them again in the assignments and in the two discussions.
Each objective is tagged with the SEC 507 Course Learning Objective (CLO) it addresses.
You'll want to read each chapter before you work through its tab on this page. Under each reading, you'll find what to focus on and which assignment the chapter connects to.
The row below previews the researchers whose work informs Module 3. You'll study Marzano in the two Marzano chapter tabs, watch Wiliam and Wormeli in the Videos tab, and find books by Wiliam, Brookhart, and Stiggins in the Review tab's bonus links.
Plan on at least two, and three is plenty: one after direct instruction, one during practice, and a quick one at closure. Beyond three, the checks start to take time you need for instruction.
You should almost never curve. A curve turns your grades into norm-referenced grades, ranking students against each other. If your scores cluster low, the problem is in the assessment, the instruction, or the standard, so fix the cause instead of adjusting the scores.
Under standards-based grading, the most recent demonstration of mastery is the most valid current measure. When you average, you punish a student for early failure even after later success has replaced it. Use the most recent score as your default unless your school requires something else.
Return it within a week of submission, because feedback helps most while students still remember the work and can use the comments on the next task. If you can't grade the work in a week, the assignment was too large for the time you have, so make the next one smaller.
You'll complete four graded assignments in this module, the three projects below and the exit ticket, plus two discussions. You'll find a rubric for each project below. Each rubric lists every criterion you need to know before you start, and you'll be graded on the rubric as written. When you're finished, you can submit your assignment in Canvas.
You'll choose one unit you teach (or plan to teach) that covers two to three weeks of instruction, and then you'll design the assessments to measure it. Your plan describes what you'll measure, when, with what instrument, how each assessment maps to an objective, and how the results turn into a grade a family can read.
1. At least three assessment types. You'll include formative, summative, and diagnostic assessments, and each one has a specific instructional purpose: checking prior knowledge before you start, adjusting your teaching mid-unit, or certifying what students learned at the end.
2. One complete rubric or proficiency scale. You'll write it for at least one of those assessments, with criteria, levels, and student-facing descriptors matched to your unit objectives. A 2.0/3.0/4.0 scale works here if it describes each level in plain language.
3. An alignment table. Your table maps every assessment in the plan to the objective it measures, and a reader can scan it in one pass.
4. A grading philosophy statement. You'll name your approach (criterion-referenced, norm-referenced, or standards-based), give your reason for choosing it, and describe what students and families learn about a student's work from the grades. Discussion 5 (The Grading Debate) is your rehearsal space, and you can bring the feedback you get there into this section.
Format: You'll submit a single document of 1,000 to 1,500 words, plus the rubric and the alignment table.
Open assignment brief → Submit on Canvas →| Criterion | Pts | Exemplary | Proficient | Developing | Needs Improvement |
|---|---|---|---|---|---|
| Assessment Variety | 25 | You include formative, summative, and diagnostic assessments and justify each one with a specific instructional purpose, such as checking prior knowledge, measuring mastery, or guiding reteaching. 25 | You include at least two assessment types with clear purposes, and one of them needs a stronger justification. 19 | You lean on a single assessment type, or you include variety without saying what each type is for. 12 | You describe one assessment, or you do not distinguish between assessment purposes. 6 |
| Rubric or Scoring Guide | 25 | You include a complete rubric for at least one assessment, with criteria, levels, and descriptors that line up with your objectives. 25 | You include a rubric that connects to your objectives, and one or two levels need sharper descriptors. 19 | You attempt a rubric that is missing criteria, levels, or descriptors, and its alignment to your objectives is unclear. 12 | You include no rubric or scoring guide. 6 |
| Alignment to Objectives | 25 | You map every assessment to a specific learning objective and show the alignment in a table or matrix. 25 | You connect most assessments to objectives and state the alignment without showing it in a table. 19 | Your assessments relate to the general topic but do not map to specific objectives. 12 | You draw no connection between your assessments and your objectives. 6 |
| Grading Philosophy | 25 | You name your grading approach, give a reason for choosing it, and describe what students and families learn about a student's work from the grades. 25 | You name your grading approach, give it a sentence of justification, and mention how you communicate it. 19 | Your grading approach is implied but you neither name nor justify it. 12 | You articulate no grading philosophy. 6 |
You'll build a questioning blueprint for one lesson in your content area. Your blueprint is a Bloom's-leveled question set crossed with equity-of-participation moves: cold-call protocols, wait-time targets, response-board moves, and structured turn-taking.
1. Eight to ten questions. You'll write them for the single lesson, label each one with its Bloom's level, and tie each one to the standard the lesson targets. You should start at recall and reach analysis or creation by the end of the lesson.
2. A sequencing rationale. You'll explain your order and give the reason you placed each question where you did.
3. Wait-time targets. Your blueprint specifies the wait-time strategy for each question type (written response before verbal for recall, think-pair-share for analysis) and how you'll hold yourself to it.
4. Equity-of-participation moves. You'll name at least three moves for surfacing answers from students who won't raise a hand (random calling, equity sticks, small-group response) and say when each one fits. Your blueprint also includes a plan for reaching English learners and reluctant participants.
You may submit the blueprint as a written document, or as a visual infographic or flowchart with a five-minute audio recording walking through your sequencing rationale. The rubric is the same either way, so you can pick whichever format lets you do your best thinking.
Format: You'll submit a single document or table of 600 to 900 words, or the infographic plus the five-minute recording.
Open assignment brief →| Criterion | Pts | Exemplary | Proficient | Developing | Needs Improvement |
|---|---|---|---|---|---|
| Question Quality & Variety | 25 | You write questions at several Bloom's levels, label each one, and move from recall to analysis or creation inside a single lesson. 25 | Your questions span at least three Bloom's levels, and the progression could be sharper. 19 | Your questions sit at one or two Bloom's levels, and the cognitive demand stays flat. 12 | Your questions are all recall, or they show no awareness of cognitive levels. 6 |
| Wait Time & Sequencing | 25 | You specify a wait-time strategy for each question type and sequence the questions so each one builds on the last. 25 | You include wait-time strategies and sequence the questions in a general low-to-high pattern. 19 | You mention wait time without naming a strategy, and your sequence is not deliberate. 12 | You give no attention to wait time or to sequencing. 6 |
| Equity & Participation | 25 | You name specific techniques for reaching every student, such as random calling, equity sticks, or small-group response, and you plan for English learners and for students who hold back. 25 | You name at least two equity techniques and acknowledge that students do not all participate the same way. 19 | You mention one technique and note equity concerns without a strategy for them. 12 | You include no strategy for equitable participation. 6 |
You'll design one full summative assessment for a unit you teach, with the items, the rubric, the table of specifications, the accommodations, and a reflective analysis of the choices behind them. Your directions and keys should be complete enough for another teacher to give the assessment without asking you any questions.
1. Table of specifications. Your table is a grid that maps each item to a content topic and a Bloom's level, with item counts per cell, and you'll justify the weighting.
2. Items. You'll include the full assessment, with answer keys for selected-response items and exemplars for constructed-response items.
3. Rubric. Your student-facing rubric has concrete level descriptors and point allocations for each criterion. Two graders using it should score the same student work within one level of each other.
4. Accommodations. You'll write specific accommodations for at least three learner profiles (English learners, students with IEPs, gifted students), and each one preserves cognitive demand.
5. Reflective analysis. In one to two pages, you'll answer three questions: why you chose this format over the alternatives, how the assessment fits the grading approach you named in your Assessment Plan, and what you'd change once you saw what students did with it.
Format: You'll submit a single packet (the table of specifications, items, rubric, accommodations) plus the reflective analysis.
Open assignment brief →| Criterion | Pts | Exemplary | Proficient | Developing | Needs Improvement |
|---|---|---|---|---|---|
| Assessment Design | 25 | Your assessment measures the objectives you stated at the Bloom's level you taught, and your item types match the work students practiced. 25 | Your assessment targets the objectives, and most items match the cognitive demand of your instruction. 19 | Your assessment relates to the topic, and the items work at a different Bloom's level than your instruction did. 12 | Your assessment is generic, off topic, or unconnected to the objectives you stated. 6 |
| Differentiation & Access | 25 | You build in accommodations for English learners, students with IEPs, and other profiles, and the accommodations keep the thinking demand intact. 25 | You address at least two learner profiles, and most of your accommodations keep the thinking demand intact. 19 | You mention accommodations that lower the demand, such as supplying answers instead of scaffolding toward them. 12 | You address no accommodations. 6 |
| Scoring Criteria | 25 | You include a rubric or answer key with point allocations specific enough that another teacher would score a paper the way you would. 25 | You include a scoring guide that another teacher could use after one clarification. 19 | Your scoring guide is vague or incomplete, and two scorers would disagree. 12 | You provide no scoring criteria. 6 |
| Reflective Analysis | 25 | You explain why you chose this format over the alternatives, how it fits your grading system, and what you would change once you saw what students did with it. 25 | You address your choice of format and its connection to grading, and your revision plan stays general. 19 | Your reflection stays on the surface and does not take up alternatives or revision. 12 | You provide no reflection. 6 |
| Criterion | Pts | Exemplary | Proficient | Developing | Needs Improvement |
|---|---|---|---|---|---|
| Content & Depth | 18 | You show you understand the module's core concepts, name specific readings, theories, or Marzano elements, and connect them to what happens in your classroom. 18 | You engage the core concepts and reference the readings, and your connections to practice stay general. 14 | You stay on topic at the surface and mention the readings in passing without an example. 9 | You do not engage the module content, or you go off topic. 4 |
| Peer Engagement | 16 | You reply to at least two peers and add something: a follow-up question, a counterpoint, or a new example that extends their idea. 16 | You reply to two peers with relevant, constructive comments, and one of them could go deeper. 12 | You reply, and the replies are brief affirmations without substance. 8 | You reply to fewer than two peers, or your replies do not engage what they said. 4 |
| Presentation Quality | 16 | Your video is clear and organized, runs within the time guidelines, and you talk to the cohort rather than read to them. 16 | Your video is clear and holds its shape, and you read from notes while keeping the audience with you. 12 | A viewer can follow your video, and it is disorganized or read start to finish from a script. 8 | Your video is unclear, too short, or missing. 4 |
You'll bring an assessment (yours or one you've observed) and walk through it on camera in two to three minutes. What's the assessment measuring, what's it claiming to measure, and where do the two diverge? Your reading should use the chapter language: validity, reliability, fairness, item formats, table of specifications.
Initial post. Your post names a specific assessment, the claims it makes, and the gaps between what it claims and what it measures, in the chapter terminology.
Two peer replies. Each reply identifies a specific design improvement and explains the chapter or element that drove the suggestion.
Open Discussion 4 prompt →In a two- to three-minute video response, you'll take a position on one contested grading scenario. The scenarios include averaging vs. most-recent-mastery, the role of standardized scores in classroom grades, redos and retakes, and zero-as-an-F. Your defense uses chapter evidence, and your replies engage peers who took a different position.
Initial post. You take one clear position on the scenario you pick and defend it with evidence from the chapters.
Two peer replies. At least one reply engages a peer who took the other side and answers their strongest point.
Open Discussion 5 prompt →Of your four graded assignments, the three projects fit together as one design. In the Assessment Plan you map formative, summative, and diagnostic measures onto the objectives of one unit and name the grading philosophy behind them. The Questioning Strategy Blueprint covers the day-to-day half: what you ask, how you ask it, and who answers. In the Summative Assessment Design you build one finished instrument with a rubric, a table of specifications, accommodations, and a reflective analysis. You'll work through the two discussions alongside all of it, taking up a sample assessment in Discussion 4 and the policy questions in Discussion 5.
Your plan holds together when your objectives, your formative checks, and your summative items measure the same behavior. Where they don't, you have a gap, and your job in this module is to find it before the assessment goes to a student.
Once you've finished the readings, activities, and assignments for Module 3, post your response to the prompt below in the Exit Ticket: Module 3 assignment in Canvas.
Chapter 1 is Providing and Communicating Clear Learning Goals. Before you can assess what students learned, you have to name what they were supposed to learn. Chapter 1 of The New Art and Science of Teaching comes first in Module 3 because you base every assessment decision this term on a goal a student can read, restate, and track.
In Chapter 1, Marzano describes what students should understand about a clear learning goal. The student's job is to know what the target is, where they stand relative to it, and how to close the gap. Your job is to make the target easy to read. When the goal is vague, the assessment built on it is vague too, and students and families can't trust the grade it produces.
You'll build both the Assessment Plan and the Summative Assessment Design on this chapter. Before you draft either one, write what a 2, 3, and 4 look like for your goal in plain student-facing language.
A proficiency scale is a leveled description of performance for a single learning target, often 0 through 4, with 3.0 stating the target itself and 4.0 describing application beyond what was taught.
A rubric is a scoring guide tied to a specific task or product. It's useful when you need to judge a performance against multiple criteria at once.
Status is where a student stands today on the scale, and growth is how far the student has moved from a prior point. If a grade hides one of them, the family gets only half of the information.
In table 1.2, Marzano names three general types of assessment and shows when each one happens in the flow of instruction. You can read the table and name the three types before you start the Assessment Plan. In Chapter 2, Marzano uses a separate scheme with four types (informal, formal, common, and student-generated), so keep the two lists apart in your notes.
Marzano prioritizes standards because a course has more content than the school year has time to teach in depth. You name the standards worth deep coverage and give the rest lighter treatment.
Assessment Plan (100): Write a proficiency scale for one priority standard before you draft a single item. From the scale you learn what items you need. Questioning Strategy Blueprint (75): Sort your questions by their level on the scale: recall questions probe 2.0, application questions probe 3.0 and 4.0. Summative Assessment Design (100): Decide up front whether you'll report status, growth, or both, and design backward from the choice. Discussion 4 (Assessment Reflection): When you examine the assessment you bring, ask whether a student could have read the goal and predicted the items. Discussion 5 (The Grading Debate): You can use Marzano's distinction between status and growth as evidence for the side you take in the debate.
Write a one-sentence answer of your own before you open the reveal, so you can compare your thinking with the explanation.
Teachers often write goals that never reach the students. The goals are written in teacher language, posted on a slide for thirty seconds, and replaced by the activity. You can direct instruction with a goal only when the student can read it, restate it, and self-locate against it. If you can't point at student work and name which level of the scale it matches, the goal didn't guide the instruction. Marzano's fix is the proficiency scale, because a scale makes you write the 2.0, 3.0, and 4.0 in language a fourteen-year-old can understand.
If you can't turn a goal into a proficiency scale, what you have is still a topic. From a topic, students learn what the lesson is about, and from a scale, they learn what they're supposed to be able to do.
Type the missing term in each blank, and press Tab to move from one blank to the next. Synonyms count where they fit. When you're done, click Check to see how you did.
Marzano's tool that breaks one learning goal into score-point levels, written so a student can read them and locate the work, is the proficiency . The level that names the target itself is the . The distinction Marzano draws between where a student stands now and how far the student has come is .
A scoring guide for one task with criteria and levels (different from a proficiency scale, which scores a target) is a . The grading approach that reports a separate mark for each standard instead of averaging points across mixed work is called grading.
In Chapter 2, Using Assessments, Marzano continues from the goals chapter. Once you have a target, you need a way to know who's hitting it, who's close, and who's far from it. Marzano treats assessment as information flowing in two directions: students learn where they stand, and you learn what to teach next.
In Chapter 2, Marzano sorts classroom assessment into four buckets: informal moves you invent on the fly (thumbs vote, fist-to-five, quick-write at the door); formal moves you plan in advance; common assessments two or more teachers give and score together; and student-generated assessments where the student proposes how to demonstrate mastery. He wants you to pick the bucket by the question you're trying to answer about student learning. The items in a textbook test bank seldom match the question you have about your own students.
When an assessment fails, the usual cause is a mismatch between the technique and the question you're asking. A multiple-choice test can't answer "did the student learn to analyze a primary source." A clipboard rubric can't answer "do thirty students share the same misconception about photosynthesis." You can read this chapter as the source of the matching decisions for your Assessment Plan and your Summative Assessment Design.
Informal assessments are quick checks, often whole-class, such as voting techniques (thumbs, fist-to-five) and response boards. They cost little, you can run them often, and they give you a read in the moment.
Formal assessments are planned in advance. They include selected-response items, short constructed response, essays, oral reports, demonstrations, and probing discussions, and each format gives you a different kind of evidence.
A team of teachers builds a common assessment and scores it against a shared scale. The team's first job is to agree on what the scale levels mean, because if teachers read the levels in different ways, their scores can't be compared.
The student proposes how to show 3.0 or 4.0 mastery. The format gives the student the most flexibility, and it tests how well the student understands the scale.
Both teacher and student walk away with new information. If the assessment gave new information to just one of you, add a step so the other person learns something too.
Assessment Plan (100): You can build the plan as a mix of daily informal checks, a weekly formal probe, and a unit-end measure. You need all three, because each one answers a question the others can't. Questioning Strategy Blueprint (75): Voting techniques and response boards belong in your blueprint as questioning moves, because with them you can scale a single question to thirty answers in twenty seconds. Summative Assessment Design (100): If you and a colleague teach the same standard, draft a common scale together before you draft items. Discussion 4 (Assessment Reflection): Ask what kind of assessment your sample is, and whether the format matches the claim the score is making. Discussion 5 (The Grading Debate): Teachers argue about student-generated assessment, which can look like rigor or chaos depending on how the teacher holds students to the scale.
Drag each assessment example into the category that fits its purpose best. When you're done, click Check.
Try a one-sentence answer before you open the reveal. Many teachers draw the line between the two in a different place than Marzano does.
The difference between the two is planning. An informal assessment is improvised in the moment to read the room, like a thumbs vote during a worked example. A formal assessment is one you plan in advance. If you score four students during silent reading with a clipboard and a five-point rubric you wrote last night, the assessment is formal, even though it looks casual. A pop multiple-choice quiz you wrote between classes may feel informal, but you planned it in advance, so it counts as formal.
An informal check stays informal when you record it, because planning in advance is what makes an assessment formal. Once it counts, it has to be reliable, and most thumbs votes aren't. Before you ask the question, decide whether the answers will go in the record.
Type the missing term in each blank, and press Tab to move from one blank to the next. Synonyms count where they fit. When you're done, click Check to see how you did.
Marzano sorts classroom assessment into four buckets. A move you invent on the fly to read the room is an assessment. A move you plan ahead and score against a known scale is a assessment. An assessment two or more teachers give and score together is a assessment. An assessment in which the student proposes how to demonstrate mastery is called student-.
A questioning move that scales one question to thirty answers in twenty seconds belongs in your Questioning Strategy Blueprint as well as your assessment plan. The chapter calls whole-class hand signals such as thumbs up or fist-to-five , and the paper or app-based boards students hold up at the same time are .
For this audit, choose a recent quiz, test, or task, and check off each statement that's true of it. The items you leave unchecked show you where the gaps are, and your progress is saved in this browser.
Chapter 11, Teacher-Made Assessment Strategies, covers planning classroom assessments, selecting items and tasks, building rubrics, providing feedback, and using assessment data to adjust instruction.
Seifert and Sutton describe assessment for learning as a five-step process built into every phase of teaching: clear goals, technique selection, motivation and confidence, instructional adjustment, and communication with families. They also separate assessment, measurement, and evaluation, three terms people in the field often mix up. Formative and informal are different too: formative names a purpose, and informal names a level of planning. Seifert and Sutton want precision from you before you design anything.
You can keep this chapter open as your reference for item-writing and rubric construction. When you defend, in the Assessment Plan, why you used a multiple-choice item here and a performance task there, you can borrow the language from this chapter.
Click two tiles to turn them over. When the tiles match, they stay open, and you'll get a new shuffle each time you play.
Before you reveal the answer, settle on a combination and have a one-sentence justification ready.
The worst combination is reliable but not valid: a test that scores the same way every time, for every scorer, on every day, and measures the wrong thing. Each student would get the same score on a retake or from a second scorer, and every one of those scores describes a skill you didn't teach. When a test is unreliable, the inconsistent results at least make you take a second look. A test that's wrong in the same way every time looks trustworthy, so nobody questions the data, and you base grades and instructional decisions on the scores.
Try this with the next quiz you plan to give: look at the first item and ask which objective it measures. If you can't state the objective in one sentence, the item isn't ready.
A criterion such as "Cites at least two primary sources from the unit and explains how each supports the claim" gives the student far more to work with than "Use evidence" does. The student should know what counts before starting the work.
If you can't see a level in the work, you can't score it. You can't see "Shows deep understanding" on the page, but you can see "Names the cause and explains the mechanism."
When you use a descriptor such as "Is a strong writer," you score the kid instead of the essay. Keep every descriptor on the essay, so you can give the next paper from the same student a different score.
Three sources beat two only when the student handles them better, so the rubric should describe what better handling looks like. When a rubric counts sources, students learn to add sources without using them well.
A colleague can score the same paper as a check on you. If your scores differ by more than a level, you need sharper language in the rubric or anchor exemplars. Inter-rater agreement is the evidence for the rubric's reliability.
If students first see the rubric when it comes back with the grade, they can't use it to plan. If they get it with the prompt, they can plan and check their work against it. Students who plan against the rubric hand in work aimed at its criteria.
Type the missing term in each blank, and press Tab to move from one blank to the next. Synonyms count where they fit. When you're done, click Check to see how you did.
Any assessment depends on three qualities. Whether the interpretation of a score is justified for a use is its . Whether a student gets the same score across forms, days, and scorers is its . Whether the instrument disadvantages some students by design is a question of .
The grid that maps each item to a content topic and an objective level is a table of . A task that asks the student to do the thing the standard names, scored with a rubric, is a .
The builder has six fields, and you can type in any of them. Your text is saved in this browser, and you can print or clear the builder when you're done.
Your Summative Assessment Design needs a rubric a colleague could use without asking you a single clarifying question. Take the draft above and have a teacher you trust score one piece of student work with it. Note where their score and yours disagreed, because the disagreements show you which language to fix.
Open M3 AssignmentsAll four drafts describe the same task, and the descriptors get sharper with each one. Hover over a number to read the note for the draft.
Draft 1: Proficient = good thesis statement.
Draft 2: Proficient = clear, focused thesis statement.
Draft 3: Proficient = thesis statement names a position and the two reasons that support it.
Draft 4: Proficient = thesis statement (1) names the writer's position on the prompt, (2) names two specific reasons drawn from the texts, and (3) appears in the first paragraph in a single sentence.
"Good" depends on whoever scores the paper, so two teachers can grade the same paper at different levels, and the scores aren't reliable.
"Clear" and "focused" are still vague, and the student still has nothing concrete to do.
This draft names a behavior, so you can score it the same way every time, and the student knows what to write.
The draft has three checkable elements, so two scorers will agree, and the student has a clear target.
An assessment can be reliable without being valid. Low reliability weakens validity as well, because you can't interpret unstable results with confidence. You'll miss the worst quadrant unless you check for it.
The bottom-right quadrant (low validity, high reliability) is hard to see. Raters agree on the scores, so the test looks sound, even though every score measures the wrong skill.
Chapter 12, Standardized and Other Formal Assessments, covers norm- and criterion-referenced testing, the terms you need to interpret test scores, bias in standardized tests, and the relationship between standardized testing and your classroom assessments. It builds on the reliability and validity concepts from Chapter 11.
Standardized tests add a second layer of data on top of your classroom assessments. In Chapter 12, Seifert and Sutton give you the vocabulary to explain a score report to a parent at conferences, the math to read a percentile rank without misreading it, and the perspective to know when a state test's results reflect your students and when they reflect a flaw in the test.
You'll get the most from this chapter if you read it alongside Chapter 11. In Chapter 11 you study the assessments you build, and in Chapter 12 you study the ones someone else built and handed to you with consequences attached. You'll use both chapters in the Assessment Plan and in the Grading Debate.
A norm-referenced test reports a student's score relative to a norm group. A 72nd percentile means the student outperformed 72 percent of the norm sample. It doesn't mean the student got 72 percent of items right.
A criterion-referenced test reports performance against a defined standard. A report might say the student met 65 percent of the grade-level criteria, or place the student at "basic," "proficient," or "advanced."
High-stakes tests carry consequences for students, teachers, or schools. The stakes change behavior, sometimes the behavior of the test-taker and sometimes the behavior of the system around the test.
Aptitude tests aim at general cognitive functioning, and achievement tests aim at specific content learned. Diagnoses for learning disabilities often use both.
Percentile ranks, grade equivalents, scaled scores, and standard scores are four different measures. A grade-equivalent of 3.3 for a first grader doesn't mean the student can do third-grade work; it means the student scored as a strong first grader on a first-grade test.
Chapter 11's three qualities apply here too, with bigger samples and tighter technical manuals. You ask the same questions: is the score consistent, is the interpretation justified, and does the instrument itself disadvantage some students?
Assessment Plan (100): Note where standardized data already exists for your students and how you can use it without overweighting it. Triangulate it with your own data rather than deferring to it. Questioning Strategy Blueprint (75): Include practice questions modeled on the format students will see on outside tests, above all for high-stakes content, so the format doesn't become the construct being tested. Summative Assessment Design (100): Borrow the discipline of a test blueprint from large-scale testing, and decide weights and item counts before you write items. Discussion 4 (Assessment Reflection): If your sample is a standardized score report, examine the report itself. What is the percentile claiming, and what would a careful reader push back on? Discussion 5 (The Grading Debate): Should standardized scores influence the grade in your class? From this chapter you get the technical ground to defend either answer.
List two misreadings before you open the answer, because you'll end up correcting the same misreadings at conferences.
The 62nd percentile means one thing: on this instrument, on this day, against the norm group the publisher used, the student outscored 62 percent of test-takers. It isn't a percent correct, a letter grade, or the student's rank in every classroom or every state. It doesn't predict next year's score, and percentile rank carries no information about distance traveled, so it doesn't measure growth. A teacher who walks into the conference with the percentile, the scale score, and the proficiency level side by side gives the parent a fuller read than any one number can.
A score report gives you one piece of evidence about a student. When you read your own classroom data alongside it, you get a fuller picture than either source gives you alone.
Standardized tests can report norm-referenced scores, criterion-referenced scores, or both, and most state accountability tests are criterion-referenced. In Module 4 you'll study the equity and access issues that follow when schools give those scores weight in placement, tracking, and resource allocation.
Each of the five quick checks below takes on a misreading of a score or a test. The first three are questions parents ask at conferences, the fourth is a colleague's question, and the fifth is about a state test whose results don't match your classroom evidence. Answer each one in your head before you open the response.
A parent says, "My son scored a 38 on the state test, but his teacher says he's meeting standards. How can both be true?"
A first-grade child posts a grade-equivalent score of 3.3 on the reading test. The parent asks if the child should skip to third grade. What's the honest answer?
A parent challenges a percentile by saying, "But my child is in the top 10% in this town. The 60th percentile can't be right." What do you say?
A colleague who teaches social studies asks you to explain reliability and validity in two sentences each, without jargon. Try your own version before you open the response.
A state test produces far different scores for two of your students who do similar work in your class. Before you blame the students, what should you check about the test?
The score line below comes from a composite scenario. Settle on a reading of each number before you open the walkthrough.
Maria Lopez · Reading 8 · Scaled Score 415 (range 100–500) · Percentile Rank 52 · Grade Equivalent 8.4 · Performance Level: Proficient · Standard Error of Measurement: ±14
Maria's score line reports five different scores. What can you claim from each one, and at what point would a claim about the same score turn into misuse?
Scaled score 415 (range 100–500). The scaled score is Maria's position on the publisher's continuous scale. It's useful for comparing across years on the same test, and you can't compare it straight across to tests with different scaling.
Percentile rank 52. Maria outperformed 52% of the norm sample on this test, which puts her near the middle of the norm group. It doesn't mean she got 52% of items right.
Grade equivalent 8.4. Maria scored as a typical eighth grader in the fourth month of school would score on this eighth-grade test. The number is useful as a developmental marker, and the misuse would be to read it as a statement of which grade's reading work she has mastered. A grade equivalent compares her score with the typical scores of students in each grade, and it doesn't measure mastery of any grade's curriculum.
Performance level: Proficient. The performance level is the criterion-referenced report. The publisher mapped scaled scores onto cut points (e.g., Below Basic / Basic / Proficient / Advanced), and the label is a categorical claim about meeting a standard.
Standard error of measurement: ±14. Seifert and Sutton don't cover this number, and most score reports print it. If Maria took the test again under similar conditions, her scaled score would be expected to fall within ±14 about two-thirds of the time. The score is an estimate rather than a certainty. When the SEM is ±14, a one-point difference at the cut score can't tell you whether the student met the standard.
Type the missing term in each blank, and press Tab to move from one blank to the next. Synonyms count where they fit. When you're done, click Check to see how you did.
A score reported against a fixed standard is . A score reported against a sample of peers is . The percent of the norm group a student outscored on this instrument is the .
A score that says the student performed like a typical student at a given grade level on this test is the . The chapter stops there, and a score report adds one more number: the wobble around any single score, expressed in scale-score units, is the standard error of . A test with consequences attached for students, teachers, or schools is called testing.
A standardized score report makes a strong sample for Discussion 4. You can model your work on the score-line walkthrough above: put each claim the report supports in one column and each claim it doesn't support in the other. When you name what each number does and doesn't tell you, your reflection gets sharper.
This tab has three chapter walkthroughs, a talk on writing proficiency scales, and two outside talks on assessment. You can watch a walkthrough before or during the chapter reading, whichever helps you process it. For each video, write one sentence connecting something in the video to a specific point in the chapter, and use the sentences to build your Discussion 4 post.
Each walkthrough is a video by another educator that previews a chapter. You can watch it before you read, then connect the chapter to the Assessment Plan, the Questioning Strategy Blueprint, and the Summative Assessment Design.
These two talks go with the module readings.
The tab has eight activities. You'll find one full-cycle branching tree with a running score, three single-decision scenarios, a three-scene assessment-purpose classifier, a nine-item sort across reliability, validity, and fairness, a four-lens reading of a single quiz item, a twenty-five-card vocabulary deck, a five-level proficiency-scale explorer, and a four-paragraph cloze on the cycle. Each activity gives you practice with a task the assignments ask you to do.
The tree has five scenes and twenty-five possible points, and each choice scores 5, 3, or 0. You'll get the most from the tree if you choose the move you'd make in your own classroom, even when another move sounds better on paper.
Tomorrow you start a new unit on persuasive writing. You have a class of twenty-eight ninth graders who came from four different middle schools, and you have ninety minutes of planning time before the first lesson.
Your unit objective: students will analyze a persuasive text and evaluate the strength of its argument using textual evidence. You have one summative slot. The standardized test prep packet keeps showing up in your inbox.
You commit to the essay. Now write the rubric for one criterion: use of textual evidence. Your district has a generic four-point rubric template. A colleague will be co-grading.
It's day six of the unit. You read a one-minute reflection from each student: "In one sentence, name the writer's claim and the strongest piece of evidence for it." Twelve of the twenty-eight students confused the claim with a topic.
At the end of the unit, a student has a 1.5 on the mid-unit essay and a 3.5 on the final essay. The district reports a single number on the standards-based progress report, and you decide which number goes on it.
Each of the three scenarios below asks for one decision. Pick a path and read the consequence, and you can restart a scenario to try the other branches.
A student scores 45 percent on the midterm. The same student scores 95 percent on the final project. The unit grade is your call. Your school has no policy. Your other students will read this grade as a sign of what you value.
What grade do you record?
You go with the average. Before you enter it, ask yourself what the next teacher will learn from this grade about what the student can do today.
You decide on the 95. A parent emails that night and asks how a student who failed the midterm finished with an A. How do you justify it?
You enter the 45.
You'll find support for this decision in both Marzano Chapter 1 and Seifert Chapter 11. If the project measured the same standards as the midterm, the project is the more valid current measure. Under standards-based grading, the grade reports what the student knows now. Document the proficiency scale you used so a parent or admin can follow your reasoning.
Averaging is defensible, and most gradebooks default to it. It also tells the student early failure is permanent and later mastery doesn't erase it. Marzano Chapter 1 separates status from growth, and when you average, you treat the two scores as equal evidence, which they aren't.
You're grading effort and timing instead of learning. In Seifert Chapter 11, validity means your use of a score is justified, and the midterm can't justify a claim about what the student knows now. A unit grade is the wrong tool for addressing procrastination, because the grade should report what the student learned.
You're designing a two-week unit on the Civil War, or your equivalent unit. You have one summative slot. The unit objective is to analyze causes and consequences of a complex historical event using primary and secondary sources. What summative do you choose?
Which summative do you pick?
You hand out the multiple-choice test on Friday.
You commit to the document-based essay. What do you need to add to the rubric before students see the prompt?
You go with the group project and presentation.
You'll find support for this choice in Marzano Chapter 2 and Seifert Chapter 11, because the task matches the verb in the objective. Build the rubric with criteria for source use, argumentation, and historical accuracy, and add an exemplar at the proficient level so students can read the target the same way you read it.
The format fits the analyze verb, but watch for individual accountability inside the group, because a presentation can hide which student did the analyzing and which one designed the slides. Add an individual reflection or a short written component to verify each student's thinking.
Well-written multiple-choice items can probe pieces of analysis, but a 50-item test can't show a student building an argument about causes and consequences from sources. Seifert Chapter 11 calls this validity: your use of the score has to be justified, and a multiple-choice score can't support a claim about the analysis your objective names. The score you get back may be consistent, but it won't tell you whether students can analyze causes and consequences from sources.
A parent emails about their child's standardized test result. The student scored at the 62nd percentile. The parent asks if that means the student got 62 percent of the questions right. How do you respond?
Which reply do you send?
You hit send.
Your explanation of the percentile is correct. What do you add for context before you hit send?
You forward the email and step away.
In Seifert Chapter 12, the percentile is a norm-referenced score rather than a criterion-referenced one. Add the student's scale score and the proficiency level so the parent can see where the student stands against the standard alongside the rank, and offer a 15-minute call. The call gives you a chance to teach the parent how to read the score.
If you said yes, you confirmed the most common standardized-score misunderstanding. In Seifert Chapter 12, percentile rank is a norm-referenced score rather than a percent correct. The parent may carry the misreading into every later score report. If you forwarded the email without an explanation, you missed a chance to show the parent how to read the score, and the parent may feel brushed aside. Test interpretation is part of your professional responsibility, so if you don't know something, say so and look it up.
Below are three classroom moments. For each one, decide which assessment purpose the teacher is using, and read the feedback after you choose.
Scenario 1 of 3. It's the first day of school. The teacher gives a 15-minute reading-comprehension probe modeled on grade-level text. Results are not entered in the gradebook; they shape who gets what reading group for the first six weeks.
Scenario 2 of 3. After 15 minutes of direct instruction on photosynthesis, the teacher pauses and asks students to draw a one-minute concept map of what they just heard. The teacher reads the maps after school and decides whether to reteach a step tomorrow.
Scenario 3 of 3. Two weeks later, students take a 45-minute exam on photosynthesis. The score is recorded against the unit standard at three levels (Below, Meeting, Exceeding) and reported on the progress report.
Drag each scenario into the quality dimension it threatens. A bad item can break more than one, so choose the closest fit, and click Check when you're done.
An eleventh-grade US history teacher writes one short-answer item for the unit summative. Hover over any colored phrase to read the note on it from each of the four lenses.
The item: "Explain the main reason the South seceded from the Union. Cite at least one piece of evidence from the unit readings." The unit objective said students would analyze causes and consequences of the conflict using primary and secondary sources. The teacher attached a generic four-point rubric with adjectives at each level (excellent, good, fair, poor) and graded sixty essays alone over the weekend. A student who transferred mid-semester and a student who reads at the seventh-grade level both took the item under the same conditions as the rest of the class.
The objective is plural (causes, consequences) and asks for analysis, while the item is singular (the main reason) and asks for explanation. Because the verbs don't match, the score will measure something narrower than the standard claims.
Analysis needs more than one piece of evidence, and the item accepts one. A student who cites one quote and stops gets the same credit as a student who weaves in three. In Seifert Ch 11 terms, a student can meet the item without meeting the construct.
Excellent, good, fair, and poor mean different things to different scorers, so two scorers reading the same essay will give it different levels. A rubric needs levels that name observable behaviors, and until it has them, each score reflects how one scorer reads the adjectives.
The transfer student missed a week of source work, and the student who reads at the seventh-grade level can't access the unit readings as written. In Seifert Ch 11, fairness means the instrument doesn't disadvantage a student by design, and giving every student the same testing conditions doesn't fix a problem in the design.
Tap a card to flip it, then mark it Got it or Review again. The cards you mark Review again come back in a second pass, and your progress is saved in this browser.
Marzano's scale has five levels. Click each card to see what the level looks like in student-facing language and what student work at the level tells you. You can build your Summative Assessment Design on these descriptors.
Type the missing term in each blank, and press Tab to move from one blank to the next. Synonyms count where they fit. When you're done, click Check to see how you did.
Marzano's tool that breaks one learning goal into score-point levels written for a student is the . A scoring guide for a single task with criteria and levels is a . The two are different: the first scores a target, the second scores a task. An assessment given before instruction to set a starting point is ; an assessment given during instruction to adjust the next move is ; an assessment given at the end of a unit to report learning is .
Any assessment depends on three qualities. Whether the interpretation of a score is justified for a use is its . Whether the score is consistent across forms, days, and scorers is its . Whether the instrument disadvantages some students by design is a question of . The grid that maps each item to a content topic and an objective level is a table of .
Reporting against a fixed standard is ; reporting against a sample of peers is . A score that says the student outperformed a given share of the norm group is the . The wobble around any single score, expressed in scale-score units, is the standard error of .
When you ask one question and look around the room before calling on a student, the pause that follows is called time. When the student proposes how to show mastery, the assessment is called student-.
The Self-Check cloze below has four short paragraphs and covers the Module 3 vocabulary: proficiency scale, rubric, status vs. growth, formative, summative, wait time, criterion-referenced, norm-referenced, percentile rank, validity, reliability, and table of specifications. You can use it as a last check before you submit the Assessment Plan and the Summative Assessment Design.
Type the missing term in each blank, and press Tab to move from one blank to the next. Synonyms count where they fit. When you're done, click Check to see how you did.
Marzano's tool that breaks a learning goal into score-point progressions is the . Where a proficiency scale levels one learning target, a scores one task across multiple criteria. The Marzano distinction between current performance and learning over time is .
An assessment given during instruction to adjust the next lesson is . An assessment given after instruction to certify learning against the standard is . Mary Budd Rowe's research-backed pause of at least three seconds after a question is called time.
A grading approach that compares each student against a fixed standard is . A grading approach that compares each student against peers in a norming group is . The score that tells you what share of the norming sample a student outscored is the .
The degree to which your use of a score is justified for its purpose is . The degree to which an assessment produces consistent results across time, forms, and scorers is . The matrix that maps assessment items to objectives and Bloom's levels is the table of .
Below are three classroom moments. For each one, pick the frame that best describes the teacher's move, and read the feedback after you choose.
Scenario 1 of 3. A student fails the unit summative. The teacher reviews the formative checks from the unit, sees that the student missed the key concept on day three and never recovered, and offers a reteach plus a retake.
Scenario 2 of 3. A teacher writes a unit test that mirrors the format of the upcoming state assessment, item by item, even using similar reading-load questions. The teacher's stated reason is to prepare students for the format they'll face.
Scenario 3 of 3. After teaching a lesson on idea development in writing, the teacher asks students to underline the strongest sentence in their draft and explain why it's strong. The teacher reads the underlined sentences before the next class to plan tomorrow.
Below are four "answer first, then check" prompts on assessment design. Each one asks a question you'll face as you plan assessments in your first month of teaching.
What's the shortest test of whether an objective will hold up to grading? Try to answer in one sentence.
Suppose you inherit another teacher's assessment plan for a unit. Which element of the plan tends to break down first?
If you read a rubric with care, you can often tell who it was written for. What's the giveaway in the language itself, and what would the student-facing version of the same rubric look like?
Before you report a percentile rank to a parent, ask yourself whether you're about to make a claim the percentile supports or one it doesn't. Which claims fall on each side?
Use these cards for a last check before you submit. The front of each card asks a readiness question, and when you flip it, you'll see what "ready" looks like.
Everything below is optional. The links go to outside sites that support the Module 3 work on classroom assessment, rubric design, and grading systems.
In this book, Wiliam lays out five formative-assessment strategies: clarifying intentions, eliciting evidence, providing feedback, activating learners as resources, and activating ownership. You can use them in the Assessment Plan assignment.
Solution Tree PressThe archive holds several Educational Leadership issues on formative assessment. You'll find them useful when you draft formative checks for the Assessment Plan.
The Association of American Colleges & Universities offers sixteen field-tested rubrics, free to download. Each one describes a learning outcome (writing, critical thinking, quantitative literacy) at four levels, and you can use one as a template for the Summative Assessment Design assignment.
In this practical guide, Brookhart explains how to write rubrics that communicate criteria to students. Read it alongside Seifert Ch 11 on assessment for learning.
Susan M. Brookhart · ASCDMarzano offers practical decision frameworks for choosing among criterion-referenced, norm-referenced, and standards-based approaches. You'll find them useful for the grading philosophy in your Assessment Plan and for Discussion 5 (The Grading Debate).
Rick Stiggins lays out his framework for assessment FOR learning (formative) versus assessment OF learning (summative). You can turn his five keys to quality assessment into rubric criteria.
Pearson