How Assessment & Diagnosis Actually Work: From First Worry to a Written Report

Is this a screening or a full assessment?

A screening is a short, broad check that answers one question: is there enough concern here to look more closely? It never diagnoses anything. A full assessment is that closer look — longer, carried out by a qualified professional, and built from standardised tests, structured observation, developmental history and what parents and teachers report.

Because screens are quick, they accept a trade-off. A good screen is sensitive: it catches most children who genuinely have a difficulty, even at the cost of flagging some who turn out to be developing typically. So a ‘refer’ result is an invitation to investigate, not bad news in itself. Familiar examples include the M-CHAT-R for autism-related concerns and the ASQ as a general developmental screen.

Where a screen gives a yes/no signal, an assessment builds a profile of strengths and difficulties across several areas — usually over more than one session, and often more than one professional. Only an assessment can support a diagnosis.

Laid out end to end, the whole route runs in five stages, and most families move through them over weeks rather than days.

What actually happens on assessment day?

Most of an assessment looks like playing. The therapist talks with you first about your worries and your child’s history, then works with your child through activities — naming pictures, following instructions, building, drawing, moving. With younger children you stay in the room throughout. There is nothing on the day that a child can fail.

A child who feels watched stops showing you what they can do, so a skilled assessor spends as much time observing as testing. We are watching how your child asks for help when a task gets hard, whether they look to you to share something funny, how they cope with being asked to stop — early emotional regulation and wellbeing — and how much they understand when nobody is using gestures to help them.

Bring what you already have — a hearing test result, a school note, an earlier report, a medical letter, and a short list of the words your child uses at home. A phone video of your child at their most typical, not their best, is often worth more than an hour in a clinic room.

At the end you should get the findings explained in plain words, and then in writing. If you leave holding a page of numbers nobody has interpreted for you, the assessment has not finished its job.

What do standard scores, percentiles and scaled scores mean?

A raw score is simply the number of items your child got right, and on its own it means almost nothing. It becomes meaningful only when compared with a large sample of children the same age — the ‘norm’ group.

The most common currency is the standard score. On most major tests the average is set at 100 and the standard deviation (a measure of how spread out scores are) at 15. Roughly two-thirds of children score between 85 and 115, and about 95% between 70 and 130. So 100 is exactly average, 85 sits one standard deviation below, and 70 — two below — is a common threshold for a ‘significant’ difficulty.

A percentile rank says the same thing in everyday language: the percentage of children in the norm group who scored at or below your child. A percentile of 50 is bang in the middle; 16 corresponds to a standard score of about 85. Two cautions. Percentiles are not the percentage of questions answered correctly, and they are not evenly spaced — the gap between the 50th and 60th percentile is far smaller than between the 5th and the 15th, because most children cluster near the middle.

Subtests usually use scaled scores instead: an average of 10, a standard deviation of 3, generally running from 1 to 19. A scaled score of 7 is one standard deviation below average — the same relative position as a standard score of 85. Knowing which scale you are reading, 100/15 or 10/3, is the single most common thing people get wrong.

Why a score is a range, not a pinpoint

Two technical terms appear constantly on reports. The basal is the point at which a child is assumed to pass all easier items, so testing need not start from the very beginning. The ceiling is where the child gets a run of items wrong and testing stops, because harder items would only add failures. Setting them wrongly is a common source of scoring mistakes.

No score is an exact point, and honest reports say so with a confidence interval — for example, ‘standard score 82, 95% confidence interval 76–88’. It means that if your child were tested many times, their true score would fall inside that band on 95% of occasions. So read a score as a range. A child who scored 82 has not meaningfully improved by scoring 85 next time: both sit inside the same band, and a report quoting a single number with no interval is telling you less than it appears to.

Norm-referenced or criterion-referenced — which question is the test answering?

A norm-referenced test compares your child with other children. Standard scores and percentiles are all norm-referenced, and they answer one question: is this child developing differently from most children their age?

A criterion-referenced test compares your child with a fixed skill, ignoring how other children perform. It asks ‘Can this child do this specific thing?’ — can they produce the /s/ sound in the middle of a word, can they follow a two-step instruction, can they hold a pencil well enough to copy a circle. A driving test is an everyday criterion-referenced test: you pass by meeting the standard, whatever anyone else did.

Good assessment uses both. Norm-referenced tests establish whether a difficulty exists and how significant it is; criterion-referenced measures then map the exact skills to teach, which is what turns a report into a therapy plan. Neither is better — a report leaning on only one is giving you half the picture.

Is a standardised test fair to a bilingual child?

Not on its own. Most standardised tests are normed on children who hear one language at home, so a child growing up with Urdu, Saraiki or Punjabi alongside English can score low for reasons that have nothing to do with a disorder. A fair assessment counts what a child can do across all their languages together, not in English alone.

This matters enormously here. A three-year-old in Multan may know the word for water in Urdu, milk in Saraiki and car in English. Counted language by language, that vocabulary looks small in each; counted as one combined total, it may be perfectly typical. Mixing languages in a single sentence is normal bilingual development, not confusion, and growing up bilingual does not cause speech delay.

Informal assessment is how a clinician gets round the limits of a test: observing play, sampling the language your child uses naturally, and taking parent and teacher report seriously as evidence rather than as background.

Dynamic assessment goes further, and it is one of the strongest tools for telling a genuine disorder apart from a difference of experience. It follows a test–teach–retest pattern: assess a skill, teach it properly, then reassess to see how much the child gained. The key idea is modifiability. A child who picks the skill up quickly after brief teaching was mostly short of exposure; a child who makes little progress despite focused teaching is more likely to have an underlying difficulty. That measures learning potential rather than prior opportunity — exactly what a one-off score cannot do.

How does a clinician move from information to a diagnosis?

It begins with a thorough developmental history: pregnancy and birth, early milestones, medical and family history, and how the concerns have changed over time. History carries real weight, because development is a story — when a skill appeared, or disappeared, can be as informative as any test.

Alongside history sit direct observation, ideally in more than one setting, and information from the people who know the child best. In robust services this is a multidisciplinary team effort: a paediatrician, a psychologist, a speech and language therapist and an occupational therapist looking at sensory processing and motor skills each contribute a piece, so no single professional’s view decides the outcome. That matters most for complex presentations such as autism, where good practice calls for a team rather than one clinician working alone.

The heart of the process is differential diagnosis: listing every condition that could explain the profile, then reasoning about which fits best and which can be ruled out. A child who is not talking might have a hearing loss, a developmental language disorder, autism, a global developmental delay, or may be a late bloomer — and a hearing test alone can change the whole picture. Differential diagnosis is the discipline of not leaping at the first plausible label.

Finally the picture is matched against formal criteria. Two systems dominate: the DSM-5 (the American Psychiatric Association’s manual, used widely in research and North American practice) and the ICD-11 (the World Health Organization’s classification, the international standard across the UK and much of the world). They are broadly harmonised but differ in how some conditions are named and grouped, which is why a report may cite one, the other, or both.

Do we need a diagnosis before therapy can start?

No. You do not need a diagnosis, a referral or a hospital letter to ask for help. Therapy targets the difficulty in front of us — the missing words, the unclear speech, the child who cannot settle — and that work can begin while a fuller diagnostic picture is still being put together. A diagnosis often helps with school support and paperwork, but it is not the price of admission. If you are unsure where your child sits, tell us what you are seeing and we will say honestly whether an assessment is the right next step.

Some things do come first, though. A hearing test comes before almost everything else when a child is not talking or not responding to their name, because a treatable hearing problem can look exactly like a language difficulty. And some patterns belong with a doctor rather than a therapist: skills your child had and has now lost, staring episodes or suspected seizures, choking or difficulty swallowing, or a sudden change in behaviour with no obvious cause. If any of those describe your child, see a paediatrician first — we will say the same if you ask us.

What happens after the report?

A report is only useful if it changes what happens on Monday morning. A good one names what your child can already do, describes the difficulty in plain words as well as numbers, and sets out two or three goals specific enough that you would know when they had been met. Vague goals like ‘improve communication’ help nobody.

Three questions are worth asking the person who assessed your child, before you leave: what would you like us to do differently at home, what should we expect to change first, and when do we look at this again? The answers turn a document into a plan.

What typically shifts first is usually smaller and earlier than parents expect. Not sentences, but a child who starts pointing at what they want; not clear speech, but a child who tries a word instead of pulling you by the hand; not calm afternoons, but ten minutes of joint play that used to last two. Those are the changes worth counting, and they are the ones we ask parents to notice between sessions.

A label is a doorway, not the whole child

A diagnosis can arrive as relief — a name that unlocks therapy, school support and a community of families who understand — or as a heavy word attached to someone you love. Both reactions are normal, and most parents feel both in the same week.

A diagnostic category describes a pattern of difficulties your child shares with others. It says nothing about their personality, their potential, their humour, or the particular way they light up. Two children with the same diagnosis can be strikingly different. Use the label as a doorway: it opens access to the right kind of help and gives the team around your child a shared language. Everything that makes your child themselves still has to be discovered person to person, and that is the part no test will ever score.