Don't Tell Me What to Do!

Note: We’re going to talk about this from the point of view of a parent-kid interaction but it could be between two adults as well.

Number 8 in my Eight Biases article was psychological reactance, and the example was simple: tell a teenager to clean their room and watch the room get less clean.

That's true for most kids. It isn't true for all of them… and for some kids it’s a big emotional reaction. Have you ever heard:

• "You're not the boss of me!"
• "You can't make me."
• "I'll do it later." (yeah right…)
• "Why do I have to? You never make my brother do it!"
• "I was GOING to do it!"
• "My stomach hurts."
• "I can't, I'm too tired."
• "Wait, can I tell you something about sharks first?"
• "Fine!" (then the door slams)
• "I don't want to go anymore." (about the trip they begged for yesterday)

We’re going to talk about when this is normal-kid-behavior and when it might be something else.

Before you label the next "no," ask one question: does this kid also avoid the things they WANT to do?

Two Kids Say No to the Same Request

Picture two kids, both about nine, both asked to clean their room.

The first one argues, stalls, and then does something close to the opposite (the clothes go from the floor to the bed). Ask again, louder, and it gets worse. Say "it's your call, but I'd love for the room to be done before the pool trip" and it may get done.

The second one also refuses. They change the subject, invent a stomachache, start a game, or ask you a question about sharks. Push, and it can turn into a meltdown that looks way out of proportion to a messy floor. Then the next morning they refuse the pool trip they begged for yesterday… because it's on the calendar now!

The easy explanation is that the second kid is the first kid, turned up louder. Isn't that what most of us would guess?

Then why would that kid skip the pool trip they begged for?

Reactance Protects a Freedom

Psychological reactance is the push to get back a freedom you feel is being taken away. Jack Brehm named it in 1966, and the core idea hasn't changed much since: when someone threatens a choice you expected to have, you're motivated to restore it (Brehm, 1966). The more important the freedom, and the bigger the threat, the harder you push back.

Why would we have this at all? Because it's useful most of the time. Reactance is the reason a pushy pitch makes you want to walk out, and walking out is often the right call.

That lab coat one has research behind it. Sagarin and colleagues trained people to spot ads using illegitimate authority, and afterward they resisted those ads and stayed open to legitimate ones (Sagarin et al., 2002).

The "truth" anti-smoking campaign is a big example of the same idea (the papers don't use the word reactance, so that part is my interpretation). Instead of telling teenagers not to smoke, it told them the tobacco companies were lying to them. Agreement that not smoking "expresses independence" went up 22 points! Meanwhile Philip Morris's own "Think. Don't Smoke" ads went with MORE intention to smoke (Farrelly et al., 2002). A later analysis credited "truth" with about 22% of the drop in youth smoking from 1999 to 2002, roughly 300,000 fewer young smokers (Farrelly et al., 2005).

Van Petegem and colleagues found that controlling parenting led to reactance, and reactance went with more problems, both acting out and internal ones (Van Petegem et al., 2015). Wanting to be your own person is healthy – in that study the problems showed up when defiance was the main way a teenager got there.

Where it doesn’t work is when nobody is actually taking a freedom away… like when you asked for the advice in the first place, or when a parent just wants the room clean.

What does the push feel like from inside? Rains pooled 20 studies with 4,942 people and found the best fit was reactance as anger mixed with counterarguing (Rains, 2013). You get mad, and you start building the case against whatever you were told to do.

Everybody has it. Some people carry more of it as a trait, and there's a scale for that, the Hong Psychological Reactance Scale, refined on 3,085 people (Hong & Faedda, 1996). A team in Portugal tested it on about 1,600 teenagers (average age about 14 to 15), found it works for that age group, and found that higher reactance went with lower wellbeing (Moreira et al., 2020).

About that teenager line… I couldn't find a solid study showing teenagers have a higher reactance than everyone else. It's a reasonable guess, but it may not be true (most of us would think it is).

Miller and colleagues found that controlling language made young adults resist a health message, and that adding a line giving the choice back helped the message work (Miller et al., 2007). A 2025 meta-analysis pooled the language studies and found freedom-threatening wording raised anger and negative thoughts, and both of those predicted less persuasion (Li & Shi, 2025).

An Aside… Worth DOUBLE

Giving the choice back can double a yes!

One original study was literally on the street. Guéguen and Pascual asked strangers for bus money, and adding "but you are free to accept or to refuse" took the yes rate from 10% to 47.5% (Guéguen & Pascual, 2000).

Carpenter's meta-analysis of 42 effects and 22,333 people found it roughly doubled compliance overall, and the exact wording didn't matter much as long as it made the freedom clear (Carpenter, 2013). It worked better in person than at a distance, and Carpenter thinks part of that is people not wanting to look like a jerk in front of someone (so it may be more than reactance).

Are there other phrases? Yes, and they don't all work the same way.

A similar application is in an enrollment, saying "That's ok, Karate isn't for everyone." In an enrollment. That gives the choice back to the family enrolling. We don’t have a study I can cite on this (if someone doing a dissertation wants my help on this, I’d volunteer)! This can change a conference from resistance to a new member and future Black Belt!

In the terms from Part 3 of the Regulating Mind series, the freedom you expect to have is the set point. A demand creates error between what you predicted (my choice) and what you sensed (the request). The controller outputs a feeling. Anger for example. The anger drives a behavior (arguing, petulance, resistance behaviors, opposition).

That lines up with Rains (2013), where reactance was anger mixed with counterarguing. Give the freedom back and the error usually closes, so the anger (and the arguing) tends to drop off.

Demand Avoidance Protects Certainty

With demand avoidance, the demand itself is the problem, whoever makes it.

What does that look like at home?

Most kids do a few of these some of the time. The demand-avoidant kid does a lot of them, even for things they like, and many parents describe it as worse at home (some of these kids hold it together at school and let it out when they get home).

A kid pushing back on YOU (reactance) usually argues. A demand-avoidant kid is more likely to change the subject, turn into a cat, or disappear… anything so the demand doesn't happen.

What the Names Mean

PDA gets called a lot of things, and the names aren't interchangeable.

None of the demand avoidance (leaving out ODD) terms are in the DSM or the ICD. Kildahl and colleagues reviewed 13 studies and found the evidence that PDA is either a separate disorder or a stable subtype of autism is currently lacking (Kildahl et al., 2021).

A 2026 review went further. Company and Rotella rated 12 studies and found every one at high risk of methodological bias, and they raise the problem of circular logic in the questionnaires used to find it (Company & Rotella, 2026). Woods has argued for years that the PDA label itself can create the category it claims to find (Woods, 2017).

ODD is a different thing. It's defined by angry mood, defiance and vindictiveness toward people (Mars et al., 2024). That's much closer to reactance (angry, and aimed at a person) than to demand avoidance, and it's one more reason not to label a kid from one incident.

Where PDA Came From

Elizabeth Newson started describing these kids at her Nottingham clinic in the 1980s and argued in 2003 that PDA was a separate profile within the autism spectrum (Newson et al., 2003). Her description was 150 children who avoided ordinary demands using social strategies – distraction, excuses, role play, charm, and then escalation when the demand stayed on the table.

When O'Nions and colleagues compared parent reports on three groups of kids, the PDA group looked like the autism group on autistic traits, close to the conduct-problem group on antisocial behavior, and higher than both on emotional symptoms (O'Nions et al., 2014).

That points at anxiety…the emotional reaction to the expectation of a threat. In this case the threat would be change.

When is that functional? Most of the time. We all have some bias toward keeping things the way they are, and it's in my Eight Biases article as status quo bias. Samuelson and Zeckhauser found that just labeling an option as the current one made people more likely to pick it, and only 28% of TIAA-CREF members had EVER changed their retirement allocation (Samuelson & Zeckhauser, 1988).

Change causes anxiety because it might (that's the expectation not a certainty... anxiety) end up with a bad outcome. It's not unhealthy to have some resistance to change. Carleton, reviewing the research on fear of the unknown, puts it this way: enough fear to approach unknowns with caution would be adaptive, as long as it doesn't get in the way of living (Carleton, 2016).

We also don't have growth and development and new experiences if we don't change. So we need change. PDA is when it's an irrational amount.

Stuart and colleagues surveyed parents of 214 kids, ages 4 to 17, with diagnosed or suspected PDA. Both anxiety and intolerance of uncertainty predicted extreme demand avoidance, and the authors describe the avoidance as a way of making things more certain (Stuart et al., 2020).

My position: this next part is me exploring model fit, and nobody has tested it yet.

A demand means something is about to change. For a kid whose system treats every unknown as a threat, that's an error signal whether it came from a parent, a teacher, a calendar… or their own plan from yesterday.

In Part 3 terms, that looks like very high observer gain (every demand weighted as a strong signal) paired with high controller gain (a small error produces a big action). Part 3 also says where those settings may come from: volatility teaches observer gain and consequence teaches controller gain. I don't know if that's the origin for these kids. It's a testable guess.

The self-imposed demand is the piece reactance theory can't explain. Reactance needs somebody on the other side taking a freedom away. When the kid wanted the pool, planned the pool, and then can't go to the pool, there's nobody to push back against.

Telling Them Apart: The Want Test

Reactance needs someone else's pressure, so it doesn't make a kid refuse their own birthday cake.

Demand avoidance often does. A kid refuses the trip they asked for because it became an expectation. An adult can't start the project they picked and were excited about. The avoidance fires with nobody pushing, and that's what the want test is looking for.

Autism Is Not PDA

My son had autism. He wasn't speaking until he was about four years old.

We used 1-2-3 Magic at home. That's one, wait five seconds, that's two, wait five seconds, that's three, and three meant a timeout equal to his age. He got to the point where, if he was trying to get my attention while I was on the phone, I'd just put my finger up (indicating “one”)… and he'd know he had to stop.

What was key is that the path was always the same. One meant two was coming, two meant the time out was coming, and so the two became the consequence. The time out wasn't the consequence. I hardly ever had to enforce it.

An autistic kid who did BETTER with a clear, predictable path is the opposite of what the demand-avoidant profile would predict. I don't know if that count would have worked the same way on a kid with the PDA profile (I'd guess not, since the count itself is a demand). It may have worked on my son partly because it took the uncertainty out: he always knew exactly what came next.

How common is the profile inside autism? Gillberg and colleagues screened autistic kids across the whole population of the Faroe Islands and found that 9 of 50 autistic kids (about 1 in 5) had features of extreme demand avoidance, which worked out to about 0.18% of the whole population, and only one of the nine still met the criteria at follow-up in late adolescence or adulthood (Gillberg et al., 2015).

Does it work the other way around? Yes. When O'Nions and colleagues screened 153 people referred for possible autism, 26 of the 27 who scored high on PDA features met autism criteria (O'Nions et al., 2016).

What To Do When It's Reactance

Give the freedom back, and keep the standard.

1. Offer a real choice inside the expectation. "Room before dinner or after dinner?" The room still gets cleaned.
2. Say the freedom out loud. "It's your call." Carpenter found the "but you are free" line roughly doubled compliance, mostly with strangers (Carpenter, 2013), and it's reasonable to expect some of that at home.
3. Correct in private. From the Eight Biases article – any correction delivered in front of your staff, no matter how right, will cost you more compliance than a private one (and kids are probably the same).
4. Less talk, less emotion. We teach this in instructor training with one phrase: "the emotion I express needs to be the emotion that they need, not what I feel."
5. Enforce the consequence you already announced, every time. An announced consequence isn't a new threat to their freedom. They already knew.

What To Do When It Might Be Demand Avoidance

You still have the expectation… you're changing how it gets presented, with less uncertainty and fewer demands hitting at once.

1. Make the next step predictable. Visual schedules, the same order every time, warnings before transitions. This is where my son's raised finger fits, with the caveat above.
2. Use the PDA Society's PANDA approach as a starting point. It stands for Prioritise and compromise, Anxiety management, Negotiation and collaboration, Disguise and manage demands, and Adaptation (PDA Society, 2026).
3. Say it as information. "The shoes are by the door" instead of "put your shoes on."
4. Watch the praise. For some of these kids, "great job!" is one more expectation to live up to next time. This may feel counterintuitive, however, understand there are infinite individual differences between people. The idea is to pay attention to what’s happening and if you identify it, you can strategize.
5. Get a clinician who knows autism involved. I'm a licensed psychotherapist, and this is one where the evaluation matters. Anxiety treatment is a real option either way.

One note: The PANDA strategies make sense, but given the 2026 review, nobody has tested them the way we'd want (Company & Rotella, 2026).

One Question Before the Label

Most of the "no's" you hear are probably reactance (the Faroe study put the demand-avoidant profile at about 0.18% of the whole population), and that fix is cheap – give the choice back and keep the standard!

If you're not sure, go back to the want test – is this kid skipping things they WANT to do? If the answer is yes, more than once, and across settings, that's the kid to talk to a clinician about. You can always reach out to us too.

(Also see The Hard Case: Autism and Stimming for the other place the Regulating Mind model runs into autism.)