- Research
- Technology-Assisted
- Computer-Based Social Skills Training
Does computer-based social skills training help autistic children and teens?
Programs on a computer or website that teach faces, feelings and social skills. Here’s what 5 sources can and can’t tell you.
A care note before you read. Getting better at a game on a screen is not the same as getting better with people. Your child’s teachers, therapists and your own eyes are the best guide to whether anything is changing in daily life.
Podcast·Two-voice deep dive
Listen to the discussion
Chapters & transcript
This is an AI-generated discussion of this article, transcribed automatically. The written summary below is the reviewed source — if the two differ, go by the summary.
0:00What is this discussion about?
Host A So, like, if you want to teach a teenager how to drive a car, you don't just, you know, hand them an iPad game.
Host B Right. Obviously not.
Host A You don't tell them to steer a digital sports car for a few hours and then just toss them the keys to a real vehicle on a busy freeway.
Host B Yeah, that would be a disaster.
Host A Because the real world is chaotic. It has blind spots, bad weather, and totally unpredictable drivers. You put them in an actual car. But right now, there is this massive push to use computer screens to teach autistic children something infinitely more complex than driving, which is real-world social skills.
Host B Yeah, and I mean, the irony there is hard to miss. We often worry so much that screens are pulling kids away from face-to-face interaction. But in the clinical and educational research space, there is a very active effort to use those exact same screens as a bridge back to the physical world.
Host A Right.
Host B The idea is basically to use technology to solve a problem that technology is often accused of making worse.
Host A Which brings us to the core of today's deep dive. We are looking directly at that bridge to see if it actually holds weight. Like, can screen time genuinely teach real-world social skills?
Host B That's the big question.
Host A Exactly. So specifically, we are looking at computer-based social skills training for autistic children and teens. And to do this, we are pulling from a very specific stack of sources for you. We have an audit document and a findings report from a Spectrum Connect demo run. And this run aggregates a huge volume of systematic reviews and randomized trials to give us like a bird's eye view of the science.
Host B But before we go any further, we have to lay down a really clear baseline for you. What we are doing today is summarizing research findings.
Host A Right. Translating the data.
Host B Exactly. We are just translating the data. This is not medical advice. And absolutely nothing we discuss here is a recommendation for any individual child. We are just looking at what the science currently proves and, perhaps more importantly, what it completely fails to prove.
1:53What does computer-based social skills training look like?
Host A So let's start by painting a picture of what this intervention actually is, because, you know, computer-based training is a massive umbrella term. If a parent or a teacher signs a kid up for this, what does it actually look like in the room?
Host B So based on the sources, we are talking about tutor-style desktop or web software. These are programs with names you might actually encounter in a school or clinic, like FaceSay, Let's Face It, Mind Reading, and Learning Emotions. Got it. And the setup is very direct. A child sits in front of a computer monitor, usually independently, and just interacts with the software.
Host A So we are completely excluding things like virtual reality goggles or those multiplayer iPad games where kids interact with each other digitally?
Host B Entirely excluded. Yes, that is a crucial scope note for this specific data set. The findings we are looking at do not cover mobile apps, tablet apps, virtual reality, or group social skills programs.
Host A Okay, that's an important distinction.
Host B Yet we are strictly looking at traditional computer or web-based software where it's literally just the child in the screen.
Host A Right.
Host B The programs typically display faces. They might zoom in on eyes or play very short video scenes. Then the software asks the child to choose the correct feeling or, you know, the appropriate social response from a multiple choice menu. And the stated goals of these programs are to improve face processing, emotion recognition, and overall social skills.
Host A Okay, if I'm visualizing this, it sounds identical to digital flashcards.
Host B Yeah, that's exactly what it is.
Host A Like a face pops up looking angry, the child clicks the word angry, the screen maybe plays a little celebration animation or gives a digital high five, and then the next face pops up. It's essentially repetitive drilling on emotional cues.
Host B It is highly structured, repetitive practice. I mean, the interface is predictable, the choices are finite, and the environment is entirely controlled by the software parameters.
3:50How much research is there, and on which programs?
Host A Well, if we have these digital flashcards being used in schools and homes, I really want to know how much weight the scientific community is actually throwing behind them. Who is studying this? And like, how big is the evidence base? Because if we are going to judge whether this works, we need to know if we are looking at five kids in a basement or a massive global research effort.
Host B Well, on paper, the volume of research actually looks quite substantial. The sources draw heavily on a 2021 systematic review that examined 17 different randomized control trials.
Host A Okay.
Host B Then we have a newer 2025 meta-analysis that pulls together 25 studies, and that one covers over 1,200 people.
Host A Wow, 1,200.
Host B Yeah. And on top of that, there's a comprehensive 2024 review that evaluated 52 separate experiments.
Host A That is a lot.
Host B It is. We also have detailed data from single-group randomized control trials, such as a trial for the program FaceSay that involved 31 school-age kids, and another trial for Learning Emotions that involved 32 kids.
Host A Wait, I'm looking at those numbers. 52 experiments, over 1,200 kids across multiple meta-analyses. Usually when you see a mountain of data that big, you think the science is settled.
Host B Right, you'd think so.
Host A It feels like we should just pack up and say case closed. But I have a feeling you're about to tell me that a mountain of data doesn't mean much if the foundation is built on sand.
Host B You nailed it. The sheer volume of studies is often a total mirage in clinical research. Having 50 studies doesn't help if 45 of them are measuring the wrong things or using deeply flawed methodologies.
Host A Wow.
Host B Yeah. And the audit document also highlights some major retrieval quirks in this specific run of data.
Host A What do you mean by retrieval quirks?
Host B Well, for instance, specific clinical trials for the programs Let's Face It and Secret Agent Society couldn't even be retrieved because of mismatched identifiers in the databases.
Host A Oh, man.
Host B Yeah. And the program Mind Reading? The data captured in this run only looked at adult populations, which falls completely outside our target demographic of school-age children and teens.
Host A Okay. So our mountain of evidence has some pretty massive caves hollowed out inside of it.
6:05What did the studies actually measure?
Host A To figure out if this intervention is actually effective, we have to look past the raw number of studies and look at what they were actually measuring.
Host B Exactly. So the populations being studied are school-aged autistic children and adolescents, generally between the ages of 7 and 15.
Host A Okay.
Host B And the data includes children both with and without intellectual disabilities. But the hinge point of this entire deep dive, the absolute most critical distinction the sources make, is the difference in the specific outcomes they chose to measure.
Host A I think I know where this is going. It's the split between getting a high score on the game versus actually having a better conversation at the lunch table.
Host B Exactly. In the scientific literature, they categorize this as near transfer versus far transfer.
Host A Okay, near transfer versus far transfer.
Host B Right. So near transfer evaluates whether the child improved at the exact, specific task they practiced. If they practiced identifying emotions on a computer screen, near transfer basically asks, can they now accurately identify emotions on a computer screen when we test them?
Host A Which is pretty straightforward.
Host B Right. Far transfer, on the other hand, asks the infinitely more difficult question. Did this screen-based practice translate into a totally different context?
Host A Like real life.
Host B Exactly. Did it result in actual real-world social behavior and skills? Does the child interact differently with their peers on a chaotic playground? Can they read a dynamic, turning, moving human face while someone is speaking to them in a noisy room?
Host A Right, because identifying a static, two-dimensional photo of someone smiling in a perfectly lit studio is fundamentally different from noticing the subtle smirk on a friend's face while three other kids are yelling in the background.
Host B Exactly.
Host A The cognitive load of the real world is entirely different from the sterile environment of the software. But wait, did they also look at maintenance? Like, if you stop playing the game, do you just forget the skills a week later?
Host B They attempted to measure maintenance, but the data is incredibly sparse.
Host A Really?
Host B Yeah. The intervention programs themselves were often quite brief. The trial for Learning Emotions, for example, involved only 12 hours of total computer time.
Host A That's not very long.
Host B No, it's not. And the vast majority of these studies failed to conduct any meaningful follow-up after the training period ended. They would test the children immediately after the 12 hours were up, record the data, and stop. Oh, wow. They didn't check back 3, 6, or 12 months later to see if the skills became a permanent part of the child's toolkit.
Host A Let's get to the results then. We have the near transfer, the far transfer, and this very shaky data on maintenance. What actually happened when these kids used the software?
8:45Do children get better at naming feelings on a screen?
Host A Let's start with the near transfer, the digital flashcards.
Host B So the near transfer data is genuinely positive. The 2021 review by Zhang demonstrated a significant increase in emotion recognition.
Host A Okay, good.
Host B To give you the specific metric, they noted a medium effect size of 0.59 on their statistical scale. The 2025 review by Liu also found large benefits specifically for desktop delivery. Children clearly improve at naming feelings on faces when they are looking at a screen.
Host A Okay, so for you listening, let's put that .59 effect size into plain English. It means that if you take a kid who used this software and a kid who didn't, the one who used the software will noticeably and measurably outperform the other on a screen-based test. It's a solid, undeniable bump in skill. They beat the digital game.
Host B Yes. The intervention successfully teaches the specific task it was designed to teach.
9:43Do the skills carry over to real friendships and daily life?
Host A But here is where the rubber meets the road. What happened with the far transfer? When these kids stepped away from the desktop and walked out into the world, did it actually help them navigate human beings?
Host B The evidence for far transfer is where the entire premise starts to completely collapse.
Host A Uh-oh.
Host B Yeah. Zhang's 2021 review found absolutely no general improvement in real social skills. None at all.
Host A Wow.
Host B And Nejati's 2024 review, that's the one that evaluated 52 experiments, found that 14 of those studies showed zero transfer to real life.
Host A Wait, zero?
Host B Zero. The review went as far as explicitly noting that traditional face-to-face social training outright beats digital training when it comes to actual social transfer.
Host A Okay, I have to pause you because I'm looking at my notes on the single group trials and some of the authors claim they had major breakthrough successes. The trial for FaceSay claimed it improved mentalizing and broader social skills. And the trial for Learning Emotions claimed it achieved something called distant generalization. Now, distant generalization sounds an awful lot like far transfer to me.
Host B I know. The terminology is incredibly deceptive if you don't look at the methodology.
Host A How so?
Host B Well, when the Learning Emotions Study claimed they achieved distant generalization, it did not mean the child used this skill with a real human being. All it meant was that during the practice sessions, the children were looking at cartoon pictograms of faces.
Host A Okay.
Host B But during the test, they were shown photographs of real human faces.
Host A You've got to be kidding me. Swapping a cartoon for a JPEG on the exact same computer monitor is what they call distant generalization?
Host B Yes. The task remained entirely on-screen, static, and two-dimensional. It was not a real human interaction. The generalization was literally simply moving from an illustration to a photograph, not from a computer screen to a playground interaction.
Host A Going back to our driving analogy from the very beginning, that's like saying a student has achieved distant generalization in driving because they practiced on a digital car that was painted blue, but they passed the test on a digital car that was painted red.
Host B That's perfect. Yes.
Host A They are still just playing a simulator. They are nowhere near a real highway.
Host B The analogy holds up perfectly.
11:57Do the gains last, and is it safe?
Host B And when we look at the other outcomes, like maintenance and potential harms, the picture remains very bleak. Right. The Zhang Review reported that the initial gains made on the screen, the near-transfer skills, were not maintained over time. The skills faded once the intervention stopped.
Host A Which is pretty standard if you're just memorizing flashcards.
Host B Exactly. And regarding safety, while no adverse effects were actively reported in the studies, the audit document points out a glaring caveat.
Host A What's that?
Host B No safety data was ever captured in the first place.
12:30Why is the evidence rated low?
Host A If the skills fade and the real-world results are essentially non-existent, how much faith are the authors of these reviews actually putting in the science? What is the official grade for this body of evidence?
Host B The audit document issues a very harsh verdict. The overall confidence in this body of research is rated as critically low.
Host A Critically low.
Host B Yes. And when looking specifically at the certainty for real-world transfer, the far transfer we've been discussing, the rating is very low, which in their grading framework means the evidence is legally and scientifically insufficient to make any sort of clinical claim.
Host A It makes me wonder how the researchers got it so wrong. Let's dig into the design weaknesses, because clearly the scientific method broke down somewhere. We already mentioned that the samples are tiny, 31 kids here, 32 kids there. Right. If you have a sample size that small, the results can be completely skewed just because two kids happen to have a really great week or a really terrible week.
Host B Exactly. The small sample sizes introduce massive volatility into the data. Beyond that, the larger reviews made the mistake of inappropriately mashing different devices together.
Host A Alright, what do you mean?
Host B The 2025 review pooled data from traditional computers, mobile phones, virtual reality headsets, and even robotic interventions all into one massive bucket. That makes no sense. It doesn't. You cannot draw a clean conclusion about desktop software when it's mixed with a child interacting with a physical robot. But perhaps the most damning flaw across the board involves the control groups.
Host A Right, because if you're running a good scientific trial on computer software, you need your intervention group to use the software for, let's say, 12 hours. Then your control group needs to spend the exact same 12 hours on a computer doing something else, maybe a neutral math game or a reading puzzle. That is the only way to know the child improved because of the social skills software and not just because you gave them a fun interactive computer screen to stare at. So what did these studies do instead?
Host B In many of these trials, the control groups received what was vaguely termed their usual curriculum. They received absolutely no extra computer time and no matched attention from researchers.
Host A If you're listening right now and you've got kids or you've ever been in a classroom, you know exactly what usual curriculum looks like. It's often just sitting at a desk.
Host B Yeah.
Host A Comparing a flashy, interactive, novel video game against a kid just sitting in a chair doing nothing isn't a fair fight. Of course the kid on the computer looks more engaged and performs better. That doesn't prove the software works. It just proves kids like novel technology.
Host B The novelty effect is a massive confounding variable here. And it gets even worse when you look at the measurement problems.
Host A How so?
Host B Well, the scientific community has no core outcome set for this intervention. Every single study basically decides for itself what success looks like and how to measure it.
Host A Which brings us back to the most glaring issue, how they actually tested the kids at the end of the 12 hours. You mentioned the near-transfer success earlier, the fact that kids got better at the screen tests. How were those tests designed?
Host B In many instances, the studies tested the children using the exact same kind of picture tasks they had just spent weeks practicing on within the software.
Host A Wow. It's the ultimate example of teaching to the test. It's like giving a student the actual answer key to study all week, giving them the exact same test on Friday, and then throwing a party when they get an A. It artificially inflates the success rate to the point of being totally meaningless.
15:58What do we still not know?
Host B Because of these foundational design flaws, tiny samples, bad control groups, and teaching to the test, the sources highlight massive gaps in what we actually know. The open questions completely overshadow the findings.
Host A So what are the biggest black holes in the data? What do we still not know?
Host B We do not know if these programs help a child make or keep real friends. That outcome is simply not measured.
Host A That's crazy.
Host B Yeah. And we do not know if any of the on-screen gains last beyond the immediate post-test because long-term follow-up data is almost universally missing.
Host A Right.
Host B Furthermore, we lack validated data for specific named programs for children, partly due to the retrieval errors and the adult-only studies we mentioned earlier.
Host A I want to circle back to the safety issue, too. If the sources state that safety data is completely missing, I want to clarify what that actually means. Does it mean the scientists rigorously tested it and proved it is 100% safe with no side effects? Or does it mean literally no one bothered to track things like screen fatigue, frustration, or negative behavioral changes after the kids turned the software off?
Host B It is definitively the latter. The researchers did not prove the intervention was safe. They simply didn't look. That's wild. No adverse events were reported because no one systematically captured or measured data on potential harms. The absence of evidence is not evidence of absence.
Host A If you are a clinical professional, like a therapist, a doctor, someone trying to help a family navigate their options, how do you look at all this missing data, all these tiny flawed studies and still make an informed decision?
Host B Clinicians reviewing this kind of evidence employ a very specific, cautious reasoning process. They weigh the hard reality we've been unpacking. Getting better at a video game is not getting better with people.
Host A Exactly.
Host B They look at the near-transfer success and acknowledge that, yes, the child can now name the emotion on the digital flashcard, but they remain highly skeptical of the intervention's overall value due to the glaring, consistent lack of far transfer.
Host A To actually earn the trust of the clinical community, what would a study need to look like? What would strengthen a clinician's confidence in this technology?
Host B Well, the clinical community would need to see a very tightly designed, rigorous trial. First, they would demand blinded raters. The people measuring the child's real-world social skills, such as a teacher or a parent, cannot know whether the child was in the software group or the control group.
Host A Because human nature is powerful. If a parent knows their kid is spending 12 hours in a highly touted social skills trial, they are going to subconsciously look for improvements. Yes. They might rate the child's behavior higher just because they expect it to be better.
Host B Yeah. The placebo effect by proxy is a real phenomenon in clinical trials. Second, clinicians would need to see measurements of actual real-world social outcomes.
Host A Like playground stuff.
Host B Right. Are they interacting differently at recess? Are they initiating conversations? They need to see data that extends far beyond mere screen-based picture tests.
Host A Makes sense.
Host B Third, as we discussed, they would need an active control group that spends the exact same amount of time on a computer doing a neutral task to rule out the novelty effect. Until the research meets those stringent criteria, clinicians will remain cautious.
Host A This has been an incredibly eye-opening journey through the sometimes muddy waters of diagnostic and intervention research. We've covered a lot of ground today, from effect sizes to digital flashcards.
19:32What are the key takeaways?
Host A Let's pull this whole deep dive together into clear, actionable takeaways for you, the listener.
Host B Sounds good.
Host A Here is your first key finding. Computer programs that teach faces and feelings do, in fact, help autistic children improve at naming emotions on screen-based tasks. The near transfer works. They can master the digital flashcards.
Host B The second point is a major limitation you have to remember, though. The studies backing up that first finding are mostly very small, and they frequently test the child on the exact same kind of picture tasks they just practiced on. This methodology artificially inflates the appearance of success.
Host A The third point is another finding, and it's the reality check. When researchers look at whether these digital skills actually carry over into real-world social skills, the far transfer, the reviews show mixed or completely null results. Real life is simply too dynamic and messy compared to a controlled screen.
Host B And the fourth point is a massive open question you need to keep in mind. We still do not know if any of those on-screen gains are maintained over time because the long-term follow-up data simply does not exist in these sources.
Host A Looking at all of this, it leaves you with a profound question to chew on. If digital interfaces are currently only proven to be good at teaching digital skills, maybe we need to ask ourselves what environment we actually want our kids to thrive in. Do we want them to master the screen or do we want them to master the playground? Because right now, practicing on one doesn't seem to prepare you for the other.
Host B That is the core tension of this entire field of research. As we wrap up, it is essential to reiterate that decisions about any individual child belong in a conversation with that child's pediatrician or care team. This deep dive cannot make that call. Research on this topic is constantly evolving, and Spectrum Connect will update its coverage as new evidence appears.
Host A Spectrum Connect translates autism research into plain language for parents and checks findings for accuracy against misinformation. If you are a credentialed expert, MD, PhD, PsyD, or DO, and would like to help validate our findings, email research@spectrumconnect.us.
Key Takeaways
Children seem to get better at naming feelings on faces on a screen. A 2021 review of 17 fair tests (randomized trials) found a medium gain on average (a score of 0.59 on a common scale). Most tests are small, and the studies differ a lot from each other, so we are not very sure.
Real-life social skills are the open question. The same review found no general improvement in social skills and said the gains were not kept over time. It also said 9 of 14 studies reported some carry-over of the skill, but the summary we read does not say carry-over to what. A 2024 review of 52 studies of social training found no carry-over in 14, and face-to-face programs did better than digital ones.
No named program has a result we can lean on. FaceSay was tested in one fair test with 31 school-aged children, who did better on feelings, on understanding what others think, and on social skills. Learning Emotions was tested in one fair test with 32 children aged 7 to 15 over 12 hours, with better results on picture-based feeling tasks and no later check. We could not find readable trials of Let’s Face It! or the Secret Agent Society computer game, and the only Mind Reading study we found was in adults.
Safety and long-term results are not known. No side effects were reported in what we read, but we found no safety data either. We read abstracts and summaries, not full papers, and some counts are still pending.
If you’re thinking about a program, a few questions help. What exactly will my child practise? How will we know if it helps outside the screen? How long will we try it before checking?
What this means for you
This page is about programs a child uses on a computer or a website, often alone, like a tutor. They show faces, eyes and short scenes, and ask the child to pick the feeling or the right thing to do. Some well-known names are FaceSay, Let’s Face It! and Mind Reading. Some programs mix a computer game with group sessions and parent sessions, such as Secret Agent Society. This page does not cover apps for phones and tablets (see our report on therapeutic apps), virtual reality, or group social skills classes. Those are covered in other reports.
The research suggests children get better at what they practise: naming the feeling on a face in a photo or picture. A 2021 review of 17 fair tests found a medium gain. A 2025 review of 25 studies with 1,226 people found a large average benefit for feeling skills, but it mixed computers with phones and tablets, virtual reality and robots, so it can’t tell us about computer programs alone. Most of this research is small, and the studies differ a lot from each other.
The harder question is whether that carries over to real friendships and everyday life. The 2021 review found no general improvement in social skills and said the gains were not kept over time. The 2024 review of social training found no carry-over in 14 of 52 studies. The main thing tested is often the same kind of picture task the child practised, and fewer tests checked real social life. Those results do not agree: one small test of FaceSay reported better social skills, while the 2021 review found none overall. The choice stays with your family and your child’s care team.
Who was studied. Autistic school-aged children and teens. The only age range we could confirm for a single trial is 7 to 15 years, with and without intellectual disability (Learning Emotions, 12 hours); FaceSay was tested in “school-aged” students. Ages and countries for the reviews were not captured. Most studies tested whether children could name the feeling on a face in a photo or picture, often right after the program ended, and few looked at how children got on with others months later. Mind Reading was first tested in adults, not children. Whether these results apply to a particular child or program is not established.
Where the studies landed
Promising but contestedAll 5 sources we could read are shown below. Most point toward gains on the practised task; what carries over to real life is where they split.
3 Points toward it helping, mostly on the practised task
1 Mixed: feelings better, real life not shown
1 Points against skills carrying over
Tap any tile to read it. Five more records turned up but are not counted: a Mind Reading study in adults, two reviews we saw only as catalogue entries, and two trials (Let’s Face It! and Secret Agent Society) we could not retrieve. The 2021 and 2025 reviews probably share some of the same trials, and the 2025 review is also used in our phone and tablet apps report.
Points toward it helping, mostly on the practised task3
Mixed: feelings better, real life not shown1
Points against skills carrying over1
Spectrum Connect reviews published research on interventions parents are exploring for their autistic children — so you can see where the evidence actually stands. No agenda, no selling, no cherry-picking. Just the studies, our method, and what it means for you.
See the research behind this
For anyone who wants to check our work · 5 sources
Where we looked
This was a web-based scoping search, not a full multi-database search: 2 web searches plus 10 page-read attempts. All 10 returned something, but only 5 gave readable content (pages from MDPI, ASHA Evidence Maps, Nature and two from Springer). 3 gave catalogue records with no abstract, and 2 returned unrelated articles because the identifiers we tried pointed to other papers. No database was searched directly, and PubMed and PMC pages were not used (they blocked automated reading), so there are no per-database counts. The search of record for a full run covers PubMed/MEDLINE, Embase, APA PsycINFO, ERIC, Cochrane CENTRAL, IEEE Xplore, the ACM Digital Library, Epistemonikos and trial registers (ClinicalTrials.gov, ISRCTN, ANZCTR). Here is the PubMed search of record.
("Computer-Assisted Instruction"[MeSH] OR computer*[tiab] OR "web-based"[tiab] OR software[tiab] OR multimedia[tiab] OR FaceSay[tiab] OR "Let's Face It"[tiab] OR "Mind Reading"[tiab] OR "Secret Agent Society"[tiab] OR "Junior Detective"[tiab]) AND ("Emotions"[MeSH] OR "Facial Expression"[MeSH] OR "Social Skills"[MeSH] OR "emotion recognition"[tiab] OR "social skill*"[tiab] OR "face processing"[tiab]) AND ("Autism Spectrum Disorder"[MeSH] OR autis*[tiab] OR asperger*[tiab]) NOT ("Virtual Reality"[MeSH] OR "Mobile Applications"[MeSH])
Run on PubMed →
Counts for every database, and the dates searched, are pending. No protocol was registered in advance. This search is not reproducible from the record we kept.
What we did with what we found
This is a scoping count, not a formal screening flow. No full text was assessed; works were judged by title with no second reviewer. The 17 distinct works less 11 not selected leaves 6, and the other 4 of the 10 cited came from looking up identifiers directly.
What the strongest evidence says
A medium gain on average in a 2021 review of 17 fair tests (0.59). The tests are small and results differ a lot between studies. The task is often the same kind of picture the child practised.
One review found no general improvement in social skills; one small FaceSay test reported some. Who rated social skills in that test, and whether they knew the group, was not captured.
Rests on one sentence in a summary of the 2021 review: gains were not maintained over time. Learning Emotions had no later check.
One small fair test each (31 and 32 children), both reporting gains. One test each is not enough to say the program itself works.
We could not retrieve a trial of Let’s Face It! or of the Secret Agent Society computer game. The only Mind Reading study we found tested adults. Secret Agent Society mixes a game with group and parent sessions, so a result for the whole package can’t show what the computer part does alone.
No side effects were reported in what we read, but we found no safety data either, so this is not the same as none.
Keep reading
What would you like to do?
Want to talk to a parent who’s been there?
We match you with families whose kid is like yours. Free, takes two minutes.
Heard of something we haven’t covered?Suggest an article
If another parent, a clinic or a video mentioned a therapy you can’t find here, tell us. We’ll read the research and review it next.
Suggest an articleThink we got something wrong?Report a factual error
We publish the whole record so it can be checked, and that only counts if we act on what you find. If a number looks wrong, a study is missing or has been retracted, or we’ve read a finding in a way the evidence doesn’t support, tell us.
You don’t need a research background to file one. “This doesn’t match what our doctor told us” is a useful report. Every one reaches a person: we reply within seven days, and within thirty we have either corrected the page or told you when we will. Substantive reports send the affected steps back through the protocol and need fresh sign-off before anything here changes.
Report a factual errorAre you a clinician or researcher?Review this article
If you have read this and the evidence holds up, you can sign off on it. We send you a personal review link where you confirm your credentials, and our team checks every review before anything is published. Once approved, your name and a link to your profile appear on this article.
Ask to review this