How to Write Alt Text for Students: Examples by Age, Plus Quizzes and Tests

A kindergarten quiz shows a picture of a pond and asks, “How many ducks are in the pond?” A sighted student counts the ducks. A student who uses a screen reader hears the picture’s alt text instead.

If the alt text says “Seven ducks in a pond,” the student hears the answer. If it says “Ducks in a pond,” the student has nothing to count. Either way, the quiz no longer measures counting for that student.

Most alt text advice is written for websites and adult readers. Teachers face two problems that advice does not cover. The first is writing for a student’s age. The second is writing for quizzes and tests, where a description can give away the answer. This post walks through both, with three examples from kindergarten to middle school.

Why classroom alt text is different

Alt text is the short description stored with an image. Screen readers read it aloud in place of the picture. It also shows when an image fails to load.

In a classroom, the person hearing that description might be:

  • A student who is blind or has low vision and uses a screen reader.

  • A young student whose teacher, paraprofessional, or parent reads the description aloud.

  • Any student whose picture did not load on a slow connection.

Perkins School for the Blind points out that most image description guidance is written for adults. Teachers need guidance that works for a 5-year-old and a 15-year-old, and on a quiz as well as a lesson. Two ideas cover most of it: give the same information the image gives, and say it in words that fit the student.

Same information, no more

On a quiz, the goal is parity. Give a student the same information the image gives a sighted student, and no more.

That sounds simple until the image is part of the question. If a diagram labels its parts, the alt text can name them. If the diagram shows the parts unlabeled, the alt text describes what is there and leaves the naming to the student. Alt text gives away the answer only when the image does.

The opposite mistake matters too. “A graph” gives a student nothing to work with. If a sighted student can read the points on a line, the description gives those points too.

Here is what parity looks like on a middle school science quiz. Plant cell parts, including chloroplasts, are part of the grades 6 to 8 science standards in many states.

Example A: Middle school science

Same information, no more

Question: Which part of the cell makes food?

A plant cell with no labels: a thick green outer wall, a large pale blue space near the center, a purple round shape with a dark center, four green oval shapes with stacks of discs inside, three orange oval shapes with folded lines inside, a blue shape of folded layers, a teal stack of curved layers, and many small dots and bubbles.

Version 1: Gives away the answer

Alt text: “A plant cell with the nucleus, cell wall, and chloroplasts.”

The picture has no labels. A sighted student has to recognize the chloroplasts by their shape. This description names them, so the student skips that step.

Version 2: Matches the image

Alt text: “A plant cell with no labels. A thick green outer wall. A large pale blue space near the center. A purple round shape with a dark center. Four green oval shapes with stacks of discs inside. Three orange oval shapes with folded lines inside. A blue shape of folded layers, a teal stack of curved layers, and many small dots.”

It describes what a sighted student sees and names nothing the picture does not name. The student still does the thinking the question measures.

Alt text gives away the answer only when the image does. Describe what is there, and leave the naming to the student.

A quick test: read the question with only your description. Can the student still answer it? Do they still have to do the thinking you are measuring?

Age decides how you say it

Parity decides what goes in the alt text. The student’s age decides how you say it.

Write for your student’s age, and skip any term the lesson has not taught yet. The simplest way to do both is to use the words your lesson already uses. The DIAGRAM Center, an image description project funded by the U.S. Department of Education, gives the same advice.

NWEA’s image description guidelines for assessments give concrete swaps by grade:

  • Graph axes: “bottom” and “side” in lower grades; “x-axis” and “y-axis” in grade 4 and up.

  • Clock hands: “short hand” and “long hand” in lower grades; “hour hand” and “minute hand” in grade 4 and up.

The information stays the same. Only the words change.

Three more habits help younger students:

  • Main idea first. Say what the image is before the details.

  • Short sentences. Many students get alt text as speech, one word after another, with no way to glance back. Short sentences are easier to hold in memory.

  • Keep color when it matters. Children who are born blind learn color words and use them, so do not strip color out of a description for them.

One exception to simpler words: if a word is printed on the image, like an axis label, repeat it exactly. A sighted student reads that word, so your student should get it too.

Here is the same bar graph described two ways for a grade 2 math quiz.

Example B: Grade 2 math

Same facts, different words

Question: How many more students chose apples than bananas?

Bar graph titled Our Class Vote: Favorite Fruit. Along the bottom, labeled Fruit: apples, bananas, oranges, and grapes, each with a picture. Up the side, labeled Number of Votes: 0 to 8. Apples 7, bananas 3, oranges 5, grapes 4.Our Class Vote: Favorite Fruit012345678ApplesBananasOrangesGrapesFruitNumber of Votes

Version 1: Words for grade 4 and up

Alt text: “Bar graph titled Our Class Vote: Favorite Fruit. The x-axis is labeled Fruit and lists apples, bananas, oranges, and grapes. The y-axis is labeled Number of Votes and runs from 0 to 8. Apples have 7 votes, bananas 3, oranges 5, and grapes 4.”

Accurate, but assessment guidelines from NWEA save the terms x-axis and y-axis for grade 4 and up.

Version 2: Words for grade 2

Alt text: “A bar graph called Our Class Vote: Favorite Fruit. The fruits are along the bottom, under the word Fruit: apples, bananas, oranges, and grapes. The numbers 0 to 8 go up the side, next to the words Number of Votes. Apples got 7 votes. Bananas got 3. Oranges got 5. Grapes got 4.”

The same facts in shorter sentences. “Bottom” and “side” follow NWEA’s wording for lower grades.

Both versions give the same information, and neither gives the answer. Only the words change to fit the student. Words printed on the graph, like Number of Votes, stay exactly as written.

Then test it. If a student would need to go through a description twice to follow it, rewrite it. NWEA’s guidelines use the same test.

When the two collide

Most of the time, parity and age work together. Sometimes they cannot.

Go back to the ducks. Counting objects to answer “how many” is a kindergarten math standard. Every way of describing that picture fails in some way. One version gives the number. One leaves the student nothing to count. One turns counting into a test of listening and memory.

Example C: Kindergarten counting

No version works

Question: How many ducks are in the pond?

Seven yellow ducks spread out across a blue pond, with grass, bushes, and two cattails around the edge.

Version 1: Gives the answer away

Alt text: “Seven ducks in a pond.”

The student hears the number instead of counting.

Version 2: Too little

Alt text: “Ducks in a pond.”

The student has nothing to count.

Version 3: A memory test

Alt text: “A duck. Another duck. Another duck. Another duck. Another duck. Another duck. Another duck.”

The student can count, but must keep track of words heard in order. That tests listening and memory, not counting.

The fix: change the question, not the alt text

Give the student seven real objects or counters to count by touch. The student shows the same counting skill without needing to see the picture. If the student has an Individualized Education Program (IEP) or 504 plan, check what it says about test materials, and work with the student’s team to add this support if it is missing.

When every description gives the answer, leaves it out, or tests something else, the question needs another way for the student to show what they know.

When no description works, fix the question. NWEA calls this visual bias: the question depends on seeing in a way words cannot fix. The fix is a different way for the student to show the same skill, such as real objects or counters to count by touch. If the student has an Individualized Education Program (IEP) or 504 plan, check what it says about test materials, and work with the student’s team.

This comes up most in the early grades. Counting, naming shapes, and reading pictures all depend on looking.

Writing alt text is often how these problems come to light. In a workshop with a team of curriculum writers, we worked through alt text for the images in their lessons, which were built in Google Slides. We talked about keeping each description right for the students’ age, and about making sure a student using only a screen reader got the same information as a student looking at the image. Sometimes, writing the alt text led a writer to decide the image was not communicating what they wanted after all.

That may be the most useful part of writing alt text. It is a check on the image itself.

A quick checklist

Use this before an image-heavy lesson or quiz goes to students.

  1. Use the words your lesson already uses.

  2. Say what the image shows first. If an adult will read the description aloud to a young student, starting with “This is a drawing of…” helps. Leave that phrase out when a screen reader will read the alt text, because the screen reader already announces that it is an image.

  3. Give the main idea, then only the details the task needs.

  4. Swap words that depend on sight for everyday ones, like “bottom” and “side” for graph axes.

  5. Keep color when it matters.

  6. Repeat any words printed on the image exactly.

  7. On quizzes, give the same information the image gives, and no more.

  8. Read it aloud once. If it takes two listens, rewrite it.

  9. If no description works at the student’s level without giving away the answer, change the question.

Where the alt text goes depends on your tool. In PowerPoint and Google Slides, look for an alt text field on the image. Alt text added in Canva does not travel with a downloaded image, so add it again wherever you post the image. In Google Forms, it depends on how you add the image. An image inside a question uses its caption as alt text. An image added on its own uses “hover text.” Either way, write it with the same care.

A note on the evidence

The major image description guidelines agree that alt text should fit the reader’s age. The DIAGRAM Center, NWEA, and the Library of Congress’s National Library Service all say so.

I have not found research that tests age-specific alt text with students in kindergarten through grade 2. The closest studies include very few elementary students. So treat this as the best current practice, and watch how your own students respond. Descriptions may not be enough on their own, either. In a study of blind and low vision students in grades 5, 8, and 11, descriptions helped, but the researchers concluded that description alone was not sufficient.

What comes next

The parity rule and the plant cell example come from Accessible by Design, a library of short courses that helps K–12 teachers build accessible quizzes, worksheets, slides, newsletters, and more. The courses will be free for educators.

Want to know when Accessible by Design launches? Sign up for the KShep Creative newsletter to hear when it launches.

Next
Next

The eLearning Ecosystem: What Has to Be in Place Before You Launch