Operant Conditioning in Parenting

The four quadrants at home

The rows are not equal. Positive reinforcement builds behavior, and parent-training programs spend most of their time on it. Negative punishment is for the short list of behaviors that must stop. Positive punishment has the weakest evidence and the worst side effects. Extinction is the tool everyone uses accidentally, in reverse. Examples for every quadrant ›

Catch them being good: the attention economy of a household

A parent's most powerful reinforcer is attention, and a household has a fixed budget of it. Think of the budget as a payroll. Misbehavior is paid immediately, reliably, and in full: the shout, the lecture, the parent crossing the room. Good behavior — playing quietly, sharing unasked, shoes by the door — mostly goes unpaid, because a quiet child is a chance to do something else. Over a thousand repetitions the child learns exactly which behaviors the household pays for.

Catching them being good reverses the payroll. Notice the behavior you want and name it, specifically and at once: "You waited until I was off the phone — that was patient," not a global "good girl" an hour later. Parent-training programs treat this labeled praise as a skill to be modeled and rehearsed in session; Parent–Child Interaction Therapy, for instance, coaches parents until they deliver ten labeled praises in five minutes of play, because that is what it takes to overturn the existing allocation.^[1][19]

The reinforcement trap: Patterson's coercive family process

Gerald Patterson's observations of families of aggressive children in their homes found a pattern that is the most important idea on this page. The parent makes a request; the child whines, argues, or screams; the parent, worn down, withdraws the request or hands over what the child wanted. Two things were just learned. The child's aversive behavior was negatively reinforced: it made the demand disappear. The parent's giving in was also negatively reinforced: it made the screaming stop. Each has trained the other, and neither meant to.^[3]

The cycle escalates because it also involves shaping. A parent who holds out through the whine and gives in at the scream has reinforced screaming; next time the child starts closer to the scream. A parent who gives in only sometimes has put the tantrum on an intermittent schedule, the schedule that produces the most persistent behavior of all. Patterson called the result the coercive family process, and his group's longitudinal work traced where it leads: coercive exchanges in early childhood predict conduct problems, rejection by ordinary peers, school failure, drift toward deviant friends, and delinquency in adolescence.^[4] More on negative-reinforcement traps ›

What parent management training teaches

The remedy Patterson's group developed became Parent Management Training (PMT), now a family of programs with a shared core.^[5] Alan Kazdin's version, refined over decades of trials with children referred for oppositional and aggressive behavior, teaches parents a small set of skills through modeling, role-play, and rehearsal — not lectures — and sends them home with practice assignments.^[1] Reviews applying the strictest criteria for evidence-based treatment consistently put parent training at the top of the list for childhood disruptive behavior; the Oregon model was among the first to meet the "well-established" standard.^[2] The order matters: praise and attending come first and take most of the program's time, commands next, time-out and point systems last.

Time-out done correctly

Time-out is short for time-out from positive reinforcement, and the full name is the instruction. It works only if the environment the child leaves — the time-in — is warm, engaged, and reinforcing; a child removed from a dull room to a bedroom full of toys has been rewarded, not punished. Done properly it is immediate, brief (a few minutes; roughly a minute per year of age for young children), calm, free of lectures, and ended when the child has been quiet for a moment rather than while she is still protesting.

Time-out has been studied with parents for more than fifty years and is part of every major evidence-based parenting program.^[7] The claim that it damages attachment has been examined directly: a 2019 review concluded that, implemented as designed, the evidence does not support the concern, and that steering parents away from time-out risks pushing them toward tools with far worse evidence.^[8] The American Academy of Pediatrics, which advises against spanking and recommends positive reinforcement and limit-setting in its place, teaches time-out in its guidance for parents.^[9] The step-by-step time-out procedure ›

Sticker charts and token systems that work

A sticker chart is a token economy, and it obeys the same rules as the ones on hospital wards and in classrooms.^[10] Most home charts fail on one of four details.

1. Immediacy. The sticker goes on within seconds of the behavior, with praise — not at bedtime for "being good today." For a young child the sticker is the consequence; the prize it buys is a bonus.
2. Small steps. "A clean room for a week" is a goal, not a behavior. Start with "clothes in the hamper before dinner" and shape upward once that is reliable.
3. Cheap, quick exchanges. Three stickers should buy something today — choosing dinner, ten extra minutes before bed, a game with a parent.
4. A planned fade. Once the behavior is steady, raise the price, space the exchanges, and let praise and natural consequences take over. A chart still running unchanged six months later has become the reason for the behavior.

Two cautions. Do not fine stickers off the chart for misbehavior; a child at zero has nothing to work for. And do not pay for things the child already loves: tangible rewards for an activity that was already interesting can reduce interest in it once the rewards stop, whereas praise, and rewards for behavior the child would not otherwise do, generally do not.^[11] Charts are for building behavior that is not happening, not for decorating behavior that is.

Bedtime and the extinction burst

Bedtime is where most parents meet extinction for the first time, usually by accident. A child who cries when the parent leaves, and whose parent returns, has been reinforced for crying; stop returning, and the crying goes through an extinction burst — louder and longer on the first nights — before it declines. The classic 1959 case followed exactly this course, including a relapse when a relative went back in.^[12]

Sleep researchers have tested both the unmodified version and graduated extinction, in which the parent checks in briefly at lengthening intervals rather than not at all. A review of 52 treatment studies found that the large majority reported clinically significant improvements, with unmodified extinction and preventive parent education best supported and graduated extinction close behind.^[13] A randomized trial that followed infants for a year found graduated extinction reduced the time to fall asleep and the number of night wakings, with no rise in stress hormones — infant cortisol was, if anything, slightly lower — and no effect on attachment or later emotional and behavioral problems;^[14] a five-year follow-up of a larger trial found neither lasting harms nor lasting benefits.^[15]

Spanking: what the evidence says

Spanking is positive punishment, and it has been studied more than any other parenting consequence. The most careful meta-analysis, restricted to ordinary open-handed spanking rather than abuse, pooled 75 studies covering 160,927 children and found spanking significantly associated with 13 of the 17 outcomes examined — more aggression, more antisocial behavior, more mental-health problems, worse parent–child relationships — every one in the harmful direction and none in the beneficial direction.^[16] The data are mostly correlational, and difficult children are spanked more; but the associations held in the longitudinal studies the review examined, and no comparable evidence finds benefits. In 2018 the American Academy of Pediatrics advised parents not to spank, hit, or otherwise physically punish children, not to use words that shame or humiliate, and to rely instead on positive reinforcement, limit-setting, and brief time-out.^[9]

The operant analysis predicts the result. Spanking as practiced is delayed, inconsistent, and escalating; it delivers a burst of parental attention; it models the aggression it is meant to stop; and it makes the parent an aversive stimulus, which produces avoidance and lying. The corporal-punishment evidence in detail ›

Consistency beats severity

Parents whose consequences are not working tend to make them bigger. The laboratory says to make them more reliable instead. In Azrin and Holz's classic review, punishment delivered intermittently was far less effective than punishment delivered every time, and punishment introduced mildly and then escalated produced adaptation: the organism learned to tolerate each new level.^[17] In practice consistency means three things. Immediacy: the consequence follows within seconds or minutes, which is why "wait until your father gets home" changes nothing except the child's feelings about the front door. Contingency: it follows the behavior and nothing else — not withheld because the parent is busy, not delivered because the parent is tired. Agreement: both parents, the grandparents, and the babysitter enforce the same short list of rules, because a rule enforced by one adult in three is a rule on a variable schedule. A rule you cannot enforce every time is better dropped than announced.

The Premack principle at home

"First homework, then screen" is the Premack principle: a more probable behavior can reinforce a less probable one.^[18] Grandmothers knew it as "first your vegetables, then dessert." Its power at home is that it needs no prizes: the reinforcers are the things the child already does when free — playing outside, watching a show, being read to — made contingent on the things he does not. The preferred activity comes after, promptly, and in an amount small enough that the child could not have had it anyway. "You can play now if you promise to do homework later" reverses the order and reinforces promising.

Screens and the variable-ratio problem

Screens complicate every contingency in the house for a specific reason: the device delivers its own reinforcement on a variable-ratio schedule, the arrangement that produces the highest rates of behavior and the greatest resistance to extinction. A game or a feed pays out unpredictably, so "one more minute" is not defiance; it is the burst a pigeon shows when the key stops paying. Ending screen time is therefore a loss, the transition off a screen is the most reliable trigger of protest in the modern household, and a parent who hands the tablet back to end the protest has reinforced it.

• Put screens on the "then" side of every first–then, never the "first."
• Signal the end with a timer the child can see, so the timer rather than the parent is the cue for stopping.
• Never hand over a screen to stop a tantrum; that is the supermarket candy with a battery.
• Set the daily amount when everyone is calm; a limit renegotiated during a protest has just been placed on an intermittent schedule. How apps and games use operant conditioning ›

What not to do

• Bribing before the behavior. "If you stop screaming I'll buy you the toy" delivers the reinforcer for screaming. Reinforcement follows the behavior you want; a bribe precedes it and is triggered by the one you don't.
• Delayed consequences. A privilege lost next weekend for something done on Tuesday is felt as arbitrary, and by Saturday it is punishing Saturday's behavior.
• Punishing after the fact. The mess discovered an hour later cannot be connected to the act, for a puppy or a toddler. A consequence discovered late is better skipped than delivered.
• Rewards that satiate. Candy after every good act stops working by mid-afternoon. Attention, activities, and choices satiate far more slowly, and they are free.
• Threats you will not carry out. An unenforced warning teaches that warnings are noise, so the next real one is ignored too.

A worked example: the supermarket tantrum in ABC

The ABC model page analyzes a supermarket tantrum and shows how to intervene at the antecedent, the behavior, and the consequence. Here is what that analysis leaves out: how the tantrum was built, and why there are two contingencies in the aisle rather than one.

Play it forward. On the first Saturday the parent gives in at the whine. On the second, resolved to be firmer, she holds out through the whine and gives in at the scream — which shapes the scream. On the third she holds out through the scream and gives in when the child hits the floor. By the fourth Saturday the child starts on the floor, because that is the response that has been paid, and the parent gives in at once, because that is the response that ends it fastest. Nobody in this story is weak or bad. Two ordinary learners have shaped each other on an intermittent schedule, and the aisle now controls both of them.

The way out is the one the ABC page lays out: change the antecedent, teach and reinforce a replacement, and make sure candy never again follows a scream. The honest addition is that the next scream will be the loudest, that giving in to it will make the one after worse, and that the parent who has decided in advance what she will do — with a snack and a job for the child already in her bag — rarely reaches the floor at all. All applications of operant conditioning › · Operant conditioning in the classroom ›

Key takeaways

• Every household runs on consequences whether or not anyone planned them. Positive reinforcement builds behavior and takes most of a parent-training program's time; negative punishment is for the short list of behaviors that must stop; positive punishment has the weakest evidence and the worst side effects.
• Attention is the household's most powerful reinforcer, and misbehavior is usually paid first and in full. Labeled praise, specific and immediate, reverses the allocation; a scolding is still attention, so a behavior that survives a hundred scoldings is probably being paid by them.
• In the coercive family process, the child's whining is negatively reinforced when the demand disappears and the parent's giving in is negatively reinforced when the noise stops. Holding out and then giving in shapes a louder version and puts it on an intermittent schedule, the most persistent of all.
• Time-out works only when time-in is warm and reinforcing, and done as designed it does not damage attachment. A sticker chart is a token economy: immediate delivery, small steps, cheap same-day exchanges, and a planned fade.
• Spanking is associated with worse outcomes on 13 of 17 measures and better outcomes on none, and the operant analysis predicts why. Consistency beats severity: immediacy, contingency, and agreement among every adult, and a rule that cannot be enforced every time is better dropped than announced.

Check yourself

A four-year-old keeps drawing on the wall even though he is scolded every time. The parent concludes the punishment is not strong enough. What does the behavior say?

A scolding is attention, and attention reinforces. If the behavior keeps happening after it has been scolded a hundred times, the scolding is not punishing it; it is probably paying for it. Watch what the behavior does, not what the consequence was meant to do, and move the attention to the behavior wanted instead.

During a dull afternoon a child hits her brother and is sent to her toy-filled bedroom for two minutes. Is this time-out?

Not in function. Time-out is short for time-out from positive reinforcement, so it works only if the time-in the child leaves is warm, engaged, and reinforcing. Leaving a boring room for a room full of toys is an escape and a reward, not a loss, and hitting is likely to go up.

A parent who used to give in at the first whine resolves to be firmer. She now holds out through the whine and gives in when the child screams. Has she made progress?

No. She has shaped screaming: the louder response is the one that was paid, so next time the child starts closer to the scream. Because she gives in only sometimes, the tantrum is now on an intermittent schedule, the schedule that produces the most persistent behavior of all. The way out is decided in advance, not during the standoff.

A parent stops returning to a crying child at bedtime. On the first two nights the crying is louder and longer than ever. Is the procedure failing?

No. That is the extinction burst, and it is the procedure working: crying that returning had reinforced gets louder before it declines. Giving in on night three would be intermittent reinforcement of the loudest crying yet, which is why the decision to hold the line, and every caregiver's agreement, has to come before night one.

Explain it to a friend. Explain how a tantrum trains a parent, using a standoff from your own household or childhood as the example.