Assessing 2D image recognition for students with CVI

Written by Matt Tietjen, TVI and creator of the 2D image assessment

Visual recognition and form accessibility

Difficulty with visual recognition is a big part of CVI. People with CVI may have difficulty recognizing objects, animals, visual scenes, environmental targets, people, faces, and many other facets of our visual world. Without recognition, the world becomes a barrage of meaningless visual input. CVI has diverse manifestations, and visual recognition skills fall on a spectrum. Some with CVI may read print and chapter books and recognize a wide variety of objects, but they have trouble walking down a hallway or navigating a new, unrecognizable environment. While others with CVI may have difficulty recognizing objects they’ve seen before or processing 2D materials, such as pictures and print text.

Individuals with CVI have varying abilities to visually access objects or learning materials in different forms. Form refers to types of visual materials—real objects, three-dimensional (3D), two-dimensional (2D), abstract, or symbolic. Some individuals with CVI can visually attend to and recognize 3D materials but struggle with interpreting flat, 2D materials. Often, realistic 3D and 2D materials can be more visually accessible than cartoon-like or symbolic materials.

When it comes to CVI, it’s important to know that looking is not understanding. Concept development must be supported with a multisensory approach. For many with CVI, compensatory skills (e.g. color coding, tactile and auditory cues, memory, and context) support recognition.

Students with CVI require comprehensive assessments to determine how they make sense of their world and the most accessible instructional approaches and learning materials. A Learning Media Assessment (LMA)—completed alongside the functional vision assessment (FVA) —is critical in determining the most accessible learning materials and how they are presented. This includes media using all available sensory channels (visual, tactile, auditory, kinesthetic), such as pictures, tactile, print, Braille, auditory, and digital.

The 2D image assessment for students with CVI can be used as part of the LMA and FVA processes to evaluate the student’s visual recognition abilities and the most accessible two-dimensional images for learning. The assessment helps determine whether photographs, drawings, colored forms, iconic pictures, or symbolic pictures are the best recognized.

The CVI 2D image assessment

This 2D image assessment is meant to offer a window into a student’s visual recognition abilities so that their team can design appropriate, accessible instructional materials. This assessment presents 5 categories of images:

Color photographs
Realistic color illustrations
Abstract color illustrations
Realistic black-and-white line drawings
Abstract black and white line drawings

These five categories are meant to encompass the majority of images a child might encounter in instructional materials throughout their school day. For example, many worksheets feature abstract black-and-white line drawings; a read-aloud story may contain color photos or abstract color illustrations; and a student’s choice board might feature realistic color illustrations. Because many with CVI have difficulty with visual recognition, it is critical that teams ask themselves what types of visual media would be most appropriate for a particular student.

In general, students with CVI tend to be more successful in recognizing realistic and color images—and this is supported by research from Manley et al. (2023). Abstract images and black-and-white line drawings tend to be the most difficult. However, while research is showing this general trend, each child is an individual, and it is important to assess each student in this area so that instructional materials may be intentionally designed with their individual visual recognition abilities as a guide.

This 2D image assessment can yield two important types of information about a student:

It gives us a window into the student’s visual recognition abilities: For example, are they able to identify images? What do their answers tell us about how they are interpreting images? Are they mostly going by color or gross form? Do they seem to be seeing only parts of an image and basing their answer on those parts? Or, are they using visual attributes or “salient features” (Roman-Lantzy)—the semantic, defining visual features that make a thing what it is—to identify what they are looking at?
It tells us if there are certain types of images that are easier or harder for the student to interpret: Is the student most successful with realistic, color images, like many students with CVI, or do they show a different pattern? Does the level of realism, the presence of color, or both, seem most important to their ability to recognize images?

Important to know

While this 2D image assessment can provide useful information for some students with CVI, it is important to understand that:

This 2D image assessment is intended to assess a student’s ability to recognize images that feature only one object at a time with a blank background. It is meant to isolate a student’s visual recognition abilities and does not evaluate a student’s ability to handle simultaneous visual information (i.e. crowded worksheets, busy visual scenes, etc.). A student who can recognize a toothbrush as a single image isolated against a blank background on this assessment may struggle to find a real toothbrush amidst the clutter of a bathroom counter or a drawing of a toothbrush on a worksheet.
Even if a student correctly identifies all of the images on the 2D image assessment, they may still have significant difficulties with visual recognition that this limited assessment did not capture. This 2D image assessment is intended as just one piece of information about a student’s visual recognition abilities—one step in an ongoing, dynamic, and individualized discovery process that will not be confined to one assessment or one school year.

Considerations for administering the 2D image assessment

There is no right or wrong way to administer this 2D image assessment. It is not a normed assessment that has to be completed in a specific manner. Rather, it is a tool that can be used as part of the functional vision assessment process and adapted to each individual student and circumstance. The general suggestions for a “protocol” below should only be seen as suggestions.

Show the student one image at a time (on paper or on a backlit screen, depending on what is best for the student). It is okay to adjust the size of the images based on what works best for the student.
Consider starting with the most abstract images first so that the more realistic images don’t give the students clues as to the identity of the more abstract images (again, this is highly individualized, and the evaluator may feel that a particular student may need to start with the more realistic images).
Simply ask, “What do you see?” as you show the student each image and record their results on the data sheet.
Consider saying something like “Thank you” to reach a response rather than telling the student if their answers are correct or incorrect.
Feel free to adjust the pacing and breaks to help the student be as comfortable as possible. Some students may complete the entire assessment in a few minutes, and others may benefit from frequent breaks or even taking parts of the assessment spread out over several sessions or days.

7 ideas to individualize the 2D image assessment

There is no single 2D image assessment for students with CVI. The version provided here is just a suggestion. It can help assess image recognition in some students with CVI, but many will likely benefit from a more individualized assessment approach. Here are some suggestions for individualizing the concept of a visual recognition assessment for different students:

Create a version of the 2D image assessment that features images of objects that are familiar to the child and that the child has had real-life experience with. Google Images can be a great source for such images.
If the 2D image assessment provided here is too long, use just some of the images from each category, provide breaks, and/or split the assessment into different sessions or days.
If asking a student to identify images verbally is not appropriate, there are a number of other ways the student can demonstrate recognition. For example:
- They can select a named image from a field (just ensure that the number of images shown at once is appropriate for the student). At first, the distractor image(s) can be very different from the target image (i.e. asking the student to find an image of a “ball” and the distractor image is a toothbrush) and then making the images increasingly similar if the student continues to answer correctly (i.e. eventually pairing the image of the “ball” with an image of a balloon and seeing if the student can still discriminate the ball from the balloon).
- Images can also be presented one at a time, along with a “yes/no” question. For example, we can show a student an image and ask, “Is this a toothbrush?” Sometimes, the image will be a toothbrush, and other times, it will be something else. Depending on the student, we may decide to start with objects that look very different from a toothbrush (i.e., show the student an image of a truck and ask, “Is this a toothbrush?”) and then work our way toward more similar images, like broom and hairbrush.
The items depicted in the 2D image assessment may not be specific enough for some students, who may do better with real photographs of their own personal items (e.g., their lunchbox, their toothbrush, etc.) presented in a variety of ways depending on each student’s response abilities.
We can integrate opportunities to assess and support recognition into a student’s natural, meaningful daily routines.
- For example, if a student uses an object schedule, in which an object of reference is used to represent each activity in their day (i.e. tambourine for music, paintbrush for art, etc.), we may decide to start pairing the objects with photos of those exact objects over time and then begin fading some of the objects as, or if, the student begins to demonstrate recognition of the photos within the context of their schedule. This kind of long-term diagnostic teaching approach within the context of meaningful routines may be more appropriate for many students than expecting them to take a single “test” of visual recognition.
- It is also important to emphasize that images are not inherently superior to three-dimensional objects. There are a variety of reasons that any given individual may learn best with real objects, 2D materials, or a combination of both. Because it is so common for school programs to gravitate toward two-dimensional materials, it is important to evaluate whether such materials are appropriate for a given student at that particular point in time. Pairing images with familiar objects in the type diagnostic teaching approach described above is just one potential way to gain such information.
We can look for opportunities to assess visual recognition of real, 3D objects using meaningful objects within routines. For example, a student who leans forward to sip from her pink water bottle at snack may or may not be visually recognizing her pink water bottle. Is she leaning forward to take a sip because her instructor is saying “drink” as she moves the cup toward the student, or because the familiar snack-time context gives her a cue that the item moving toward her is her water bottle, or because her water bottle is pink and she sees something pink moving toward her, or because she recognizes the water bottle by its defining visual attributes (i.e. cap and straw)? By moving the cup toward the student silently, doing so in different settings, and sometimes substituting other pink objects for the cup can give us a sense of whether the student is visually recognizing the cup or using other compensatory strategies for recognition. If she still leans forward to take a sip even when her instructor is not saying “drink,” even when in contexts other than a snack, and only when it’s her water bottle rather than other pink objects, she may be showing us that she has some visual recognition of “cup.”
Some students with CVI will correctly identify all or most of the items in the 2D image assessment. This indicates that they have some ability to identify items based on their defining semantic features (since black-and-white line drawings are devoid of most other cues). For these students, we can assess further by adding more difficult images—specifically items with unexpected color, surface patterns, and/or contexts. For example, we can show the student an image of an elephant with zebra stripes (unexpected surface pattern), a toothbrush in a tree (unexpected context), or a blue banana (unexpected color). These kinds of images can be found with a basic Google Images search and/or created using basic editing programs like PowerPoint or Microsoft Word. Correct answers suggest that the student’s ability to use salient (semantic, defining) features is strong enough that they do not need to rely on less reliable cues like color, surface pattern, or context to try to identify what they are seeing.

Ways to apply the results

The results of any 2D image assessment will not prescribe any particular outcome but, instead, can be used as another source of information for the team collaboration and decision-making process regarding accommodations and interventions. Below are some suggested ways the results can be applied or considered.

Instructional materials

The results of a 2D image assessment can give the team valuable information regarding what types of images may currently be appropriate to incorporate into instructional materials and which types may not be appropriate.

If a student demonstrates the ability to identify color photographs or realistic color illustrations well but struggles with black-and-white line drawings, the team may conclude that all classwork for this student should feature realistic color images rather than the line drawings found on most worksheets.

If abstract color images are difficult for the student, the team may find that supplementing abstract images in children’s books with real pictures from the internet (i.e. a real image of a spider against a plain background to compare to the more abstract cartoon spider in an “Itsy Bitsy Spider” book) may help the student understand the books better.

If a student demonstrates difficulty identifying all types of 2D images, it is important for the team to consider whether real objects, in place of (or paired with) images, may be a more effective, accessible way to represent concepts to the student. For other students, detailed image descriptions (verbal or written) in place of (or along with) images can be helpful in making the content accessible.

Compensatory strategies

It’s important to always consider visual fatigue in our students with CVI, where the energy used for visual search and recognition of 2D materials can be incredibly taxing for some. Using a multisensory approach (auditory, tactile, visual, kinesthetic) for learning can support access throughout the school day. For example, using real objects to learn about the details, attributes, and function of an item first before moving to 2D. Even if a student is able to recognize images during the 2D assessment, several factors can disrupt visual recognition and access, such as fatigue, time of day, the environment, visual clutter, other distracting sensory inputs, stress and anxiety, and so much more.

When a student demonstrates difficulty interpreting images based on their salient, defining (semantic) features, the team may consider whether instruction in visual thinking skills could be a helpful compensatory strategy. Specifically for those with CVI, where recognizing 2D doesn’t cause too much visual fatigue or stress. For example, a student who only recognizes brown dogs because their dog is brown is relying on color (a less reliable recognition cue) rather than the more reliable salient, defining features (dogs generally have four paws, a snout, and a tail). The team may decide to try teaching the student to apply systematic thinking strategies as they look at an image (i.e. “This animal has a snout, four paws, and a tail, so I think it could be a dog. It doesn’t have hooves or a mane, so it’s not a horse”). In her approach to CVI, Roman-Lantzy sometimes describes this as teaching a student to be a visual detective, helping them think about what clues to look for in an image.

Considerations for AAC (Augmentative and Alternative Communication)

Many AAC systems incorporate images that a student can select in order to communicate a message. Often, the stock images that come with these systems are abstract/symbolic in nature. It is important to keep in mind that a 2D image assessment like the one described here evaluates how well a student can identify images they have never seen before. It doesn’t give us much information about how well a student could learn a familiar set of images, which are always in the same location, and which they have had repeated, supported opportunities to practice within meaningful language contexts. Therefore, when it comes to selecting types of images for a student’s AAC system, it is important to see the results of a 2D image assessment as just one piece of helpful information in a collaborative process that should involve many team members and be completely individualized to each student.

Download 2D image assessment materials

Read recent research on CVI and 2D recognition: Manley, CE, Walter, K., Micheletti, S., Tietjen, M., Cantillon, E., Fazzi, E., Bex, P., Merabet, L (2023). Object identification in cerebral visual impairment characterized by gaze behavior and image saliency analysis. Brain Dev (2023).