Visual perceptual activities, grouped by the seven skills

Visual perceptual activities in occupational therapy are easier to plan when they are grouped by the skill they load rather than kept as one long list of games. The TVPS-4 splits perception into seven subtests, and that split gives seven headings to sort activities under. The caution worth carrying through the session: motor-free perception explains only a small share of handwriting (Klein and colleagues, Physical & Occupational Therapy in Pediatrics, 2011).

What are the components of visual perception?

There is no single agreed list, and the two lists a therapist meets most often do not match. The Test of Visual Perceptual Skills, fourth edition, scores seven, named in the order Asem and colleagues report them in Children in 2023: visual discrimination, visual memory, visual–spatial relationships, visual form constancy, visual sequential memory, visual figure–ground, and visual closure. All seven are assessed in a multiple-choice response format, so nothing is drawn.

Clinical handouts often compress that. The Nottinghamshire Healthcare NHS Foundation Trust page on visual perception skills names five components for parents: visual discrimination, visual closure, visual memory, visual spatial awareness and visual motor coordination. The last of those is not a perception subtest at all — it is the perception-plus-movement construct, which visual motor integration and what the Beery VMI measures covers separately.

Use the seven-way split when you are writing a plan against test scores, and the five-way split when you are explaining the idea to a family. The granularity differs; the underlying skills do not fall apart differently depending on which handout you print.

Visual perceptual activities for each skill

The table below pairs each TVPS-4 skill with activities and with the neighbouring skill it is most often confused for. The activity column draws on the Nottinghamshire NHS Trust list — hidden pictures, dot to dot, jigsaws, construction games, memory games, sorting objects, matching colour games — plus the standard variations of those games. One thing to be straight about: that page presents its activities as general suggestions and does not assign them to components. The pairing here is ours, made by asking what each game actually requires, and none of these activities has evidence of transfer to handwriting behind it.

Skill	Activities	How it differs from its neighbour
Visual discrimination	matching colour games; sorting objects by one feature; spot-the-difference pairs	the forms stay the same size and orientation, unlike form constancy
Visual form constancy	sorting the same shape across sizes, colours and rotations; finding a triangle in a photo of a roof	the form has been changed before the child has to recognise it
Visual memory	memory and pairs games; show a card, hide it, ask what was on it	one item at a time, with no order to keep
Visual sequential memory	a row of three or four cards reproduced in order; bead strings copied from a hidden model	the order carries the task, not the items
Visual figure–ground	hidden-picture books; finding one named object in a cluttered drawer or toy box	the target is fully visible, just surrounded
Visual closure	pictures half covered by a sheet of paper; a jigsaw piece identified from its visible part	the target is incomplete, not hidden
Visual–spatial relationships	construction games copied from a model; dot to dot; routes on squared paper	the question is where parts sit, not what they are

Two notes on using it. The right-hand column is the one that makes the difference in a session: a child who fails figure–ground items and a child who fails closure items are doing different things when they get stuck, and the game you reach for should change accordingly. And the activities are activities, not treatment — they load the skill, which is not the same as being shown to improve it.

Visual spatial relations activities: position in space, left and right, direction

Spatial relationships is the subtest closest to pencil work, which is why visual spatial relations activities usually involve copying something rather than picking a picture. A block model to reproduce, a bead pattern, a design on squared paper: all three require the child to encode where parts sit relative to each other and then hold that while building.

Direction words are worth separating from direction itself. A child can place a block correctly and still not have the word for what they did, and the two are trained differently: the first by copying, the second by hearing and using position language while they work. Our post on spatial reasoning activities for preschoolers goes through the squared-paper versions and the left–right half in detail, including what it costs to work out someone else’s left while they face you.

A grid is a useful surface here for a reason that has nothing to do with perception testing: it gives countable units. “Three along, two up” is checkable by both of you, so a positional error becomes visible instead of a matter of opinion. That makes grid work good for seeing where a child’s spatial judgement breaks down, which is a different claim from saying grid work remediates it.

Do visual perceptual skills predict handwriting?

Less than perception-focused programmes imply. Two studies set the ceiling.

Volman, van Schendel and Jongmans compared 29 children with handwriting problems and 20 classroom peers in grades two and three in AJOT in 2006 — two decades old now, and still the study this question gets argued from — testing visual perception, visual motor integration, fine motor coordination and cognitive planning. The group with handwriting problems scored significantly lower on all four, visual perception included. But in the regression, visual motor integration was the only significant predictor of handwriting quality in that group; in the control group the only predictor was unimanual dexterity. Perception was lower in the children who struggled, and still not the variable that explained their handwriting.

Klein and colleagues put a number on the whole battery in Physical & Occupational Therapy in Pediatrics in 2011. With 99 students in grades three to six, they measured the Beery VMI fifth edition, the Test of Visual Perceptual Skills-Revised, upper-limb speed and dexterity, and a visual skills appraisal against handwriting legibility and speed. Correlations with legibility ran from .07 to .38. Stepwise regression put the variance explained by the full battery at 20% or less for legibility and 26% or less for speed, and only the Beery VMI scores separated students classified as skilled from unskilled writers.

So the honest framing for a parent asking why perception is being tested at all: a low perception score tells you something about how this child takes in a page, and it will not tell you most of why the handwriting looks the way it does.

Does visual perceptual training carry over to handwriting?

This is the question the evidence answers most directly, and the answer is mixed in a specific, usable way.

Poon and colleagues ran the cleanest test of it in Research in Developmental Disabilities in 2010. It is more than fifteen years old, which matters less here than it might, because it isolates the variable at issue: perception training against conventional teaching, with legibility measured on both. Twenty-six primary-one children with handwriting difficulties were matched on age and randomly assigned to a computerised visual perception and visual-motor integration programme — eight sessions plus a home programme — or to conventional handwriting teaching. The experimental group improved on visual perception skills and on handwriting time, both “On Paper” and “In Air”. On visual motor integration and on handwriting legibility, no significant difference was found between the groups. The training moved what it trained and the speed of writing; it did not move how the writing looked.

Taverna and colleagues got a different result on the construct itself, with a programme that was not perception-only. In IJERPH in 2020 they ran a 10-week intervention based on occupational therapy principles with 55 kindergartners and first graders, testing before, after and a month later. VMI scores improved significantly between the first two time points, particularly for the kindergartners. Manual dexterity scores did not change in either group, though children below the 16th percentile on manual dexterity did improve over time.

Then the two reviews that frame all of this. Kadar and colleagues screened 785 studies of occupational therapy intervention for handwriting in 4-to-6-year-olds for the Australian Occupational Therapy Journal in 2020 and included seven, finding no randomised controlled trial among them; their own conclusion is that the results need to be approached with caution. And Hoy, Egan and Feder, reviewing 11 controlled trials in the Canadian Journal of Occupational Therapy in 2011, reported two findings that cut across treatment type: interventions without handwriting practice were ineffective, and so were interventions with fewer than 20 sessions.

Put together, that is not an argument for dropping perception work. It is an argument against expecting legibility to follow from it, and for keeping the pencil in the session whenever legibility is the goal.

How to pair perception and drawing in one session

The practical consequence of the paragraph above is a session shape rather than a new activity list: perception work for what perception work shows you, then something drawn, on paper, for long enough to count against the 20-session floor.

A graphic dictation is one way to fill the drawn half, and it is worth being precise about what it is. The instruction arrives by ear — “three squares right, one diagonal up, two down” — and the child counts squares in a squared notebook and draws the line. No model is visible at any point, and the picture only resolves at the end of the route. That makes it the opposite case from a copying task: it does not test visual motor integration in the Beery sense, because there is no form on the page to match. What it loads is motor planning, direction and spatial relationships, landing the pencil on the next grid node, and keeping eight directions straight with diagonals among them.

We build Brainy Grid around that exercise, and our page for occupational therapists states the limits in the same terms: the app reads the finished line from a photo taken on the device and scores accuracy, orientation, counting and persistence, and it does not see grasp or pressure at all. Those stay with you in the room.

This week, if you have a child with a flagged perception score, check which norm table the score was read against before you plan around it — Asem and colleagues recorded a 116.60 mean against a US mean of 100 in a sample of 72 preschoolers, and that gap moves who looks impaired. Over the next month, count sessions rather than activities, since 20 is the number Hoy and colleagues attached to anything working for handwriting. And when the perception picture is clean and the writing still is not, the Volman and Klein results point the same way: look at the copy, not the pointing task.