Walk into many “social skills groups” and you’ll find something that looks a lot like a classroom: scripted conversation starters, turn-taking worksheets, rules about eye contact, and children rehearsing greetings at each other on cue.
Here’s the uncomfortable question to ask about that room: is anyone in it making a friend?
For neurodivergent kids, the scripted approach has a deeper problem than being boring. It teaches that social connection is a performance — a set of neurotypical behaviors to execute correctly — rather than a relationship between two people who enjoy each other. That’s not social skill. That’s masking with a curriculum. And masking, practiced hard enough for long enough, is consistently associated with anxiety and burnout — even when connection was the goal.1
There’s a better frame, and it starts with a piece of research worth knowing.
The two-way street: what “double empathy” changes
For decades, social difficulty in autism was described as a one-way deficit: the autistic child fails to understand others, so we train the child. The double empathy problem — a concept from autism research — reframes it: communication breakdowns between autistic and non-autistic people run in both directions.2 Neurotypical people also routinely misread autistic people.3 Meanwhile, autistic people often communicate with each other quite comfortably.4
Taken seriously, this changes what “social skills” should mean:
- The goal can’t be “act neurotypical,” because the mismatch isn’t a one-sided defect to fix.
- Kids benefit from understanding neurotypical norms — as information about how many people communicate, not as the only right way to be.
- The most valuable social skills are the universal ones: communicating your needs, understanding someone else’s, repairing a conflict, sharing an enthusiasm. Those work in every direction.
In practice, the reframe sounds like the difference between “look at your friend’s eyes when you talk” and “your friend can’t tell you’re listening — what could you do so they know?” One trains a behavior. The other builds a communicator.
What actually builds friendships
Friendship doesn’t grow from rehearsed dialogue. Decades of watching children (and being one) point to the same ingredients:
Shared joy. Children bond over the thing they both love — the train layout, the Minecraft build, the dinosaur taxonomy dispute. Genuine shared interest carries a conversation better than any script, because there’s actually something to talk about. A good group finds the overlap in kids’ passions and builds the session there: if the group loves trains, you build a transit system; if they love space, you become astronauts.
Real conflict, resolved in real time. Scripted groups avoid disagreement; real friendship runs on surviving it. When two kids want the same locomotive, that moment — not a worksheet about sharing — is where negotiation, perspective-taking, and repair get learned, with an adult guiding the process while the stakes are real but small.
Regulation first. A dysregulated child cannot be a social child — no curriculum overrides a nervous system in distress. Sensory-aware environments (movement breaks, quiet corners, fidgets, room to swing or bounce) aren’t perks; they’re what makes the social work possible at all.
The child’s own pace. Some children connect in weeks; others watch from the edge for a month before joining. Watching is participation — it’s how many kids learn a room. Pressure to perform sociability on a schedule mostly teaches kids to avoid the room entirely.
The test of a good group: the kids would come back on Saturday if you told them it was cancelled. If attendance runs on tokens and parental enforcement, the group is teaching compliance in a social costume.
What to ask before enrolling your child in any group
- “What does a session look like?” Listen for play, shared projects, and kid-chosen themes — versus lessons, scripts, and drills.
- “How do you handle conflict between kids?” The affirming answer treats conflict as the curriculum, not a disruption to it.
- “Do you target eye contact or ’expected behaviors’?” You now know why this question matters (the long version).
- “How is the room set up for sensory needs?” Look for movement options and retreat options, both.
- “How are groups composed?” Small, matched thoughtfully — by age and communication style — so kids can actually find each other.
- “Who runs it, and with what oversight?” Facilitators experienced with neurodivergent kids, with clinical oversight, so support is individualized rather than generic.
The goal, restated
The point of social support is not a child who performs politeness convincingly. It’s a child who has found their people — who knows what they need in a social situation and can ask for it, who can weather a disagreement without losing the friend, and who has felt what it’s like to be enjoyed for exactly who they are.
That child will handle the neurotypical world just fine — not because they were drilled on its rules, but because they’re operating from connection and confidence instead of a memorized script.
Yoli runs small, neuro-affirming social skills groups at our Albany playspace for autistic, ADHD, and gifted/2e kids across the East Bay — matched by age and communication style, facilitated by experienced staff with BCBA oversight. Get in touch to ask about the next cohort. Related: supporting your gifted + autistic (2e) child.
Cook, J., Hull, L., Crane, L., & Mandy, W. (2021). Camouflaging in autism: A systematic review. Clinical Psychology Review, 89. doi.org/10.1016/j.cpr.2021.102080; on burnout, Raymaker, D. M., et al. (2020). Defining autistic burnout. Autism in Adulthood, 2(2). doi.org/10.1089/aut.2019.0079 ↩︎
Milton, D. E. M. (2012). On the ontological status of autism: the ‘double empathy problem’. Disability & Society, 27(6). doi.org/10.1080/09687599.2012.710008 ↩︎
Sheppard, E., Pillai, D., Wong, G. T.-L., Ropar, D., & Mitchell, P. (2016). How easy is it to read the minds of people with autism spectrum disorder? Journal of Autism and Developmental Disorders, 46(4). doi.org/10.1007/s10803-015-2662-8 ↩︎
Crompton, C. J., Ropar, D., Evans-Williams, C. V., Flynn, E. G., & Fletcher-Watson, S. (2020). Autistic peer-to-peer information transfer is highly effective. Autism, 24(7). doi.org/10.1177/1362361320919286; Crompton, C. J., et al. (2020). ‘I never realised everybody felt as happy as I do when I am around autistic people’. Autism, 24(6). doi.org/10.1177/1362361320908976 ↩︎