Self-Control in Dogs: The Role of the Frontal Cortex, Stress, and Training

1. What the "Prefrontal Cortex" Means in a Dog

The frontal cortex is the forward portion of the cerebrum, conventionally described as the brain's executive region because it integrates sensory and limbic input and modulates behavior accordingly. In dogs, that description is broadly appropriate — but it carries assumptions worth unpacking before building training advice on it.

1.1 Core Executive Functions

Across mammals, frontal executive systems are associated with four capacities that matter directly for dog behavior: inhibiting a prepotent response, weighing likely outcomes before acting, regulating emotional responses so that fear or excitement does not fully determine behavior, and shifting strategy when circumstances change (which is the flexibility side of the same system).

Put simply, this is what allows a dog to pause rather than react. Olsen (2018) provides the fullest review of executive function in dogs, covering inhibitory control, cognitive flexibility, and working memory — and her central argument is that the field's methods need substantial overhaul before strong claims can be made.

1.2 A Necessary Caveat About Homology

The canine frontal lobe is proportionally much smaller than the primate frontal lobe, and its subdivisions do not map cleanly onto the human dorsolateral and ventromedial prefrontal regions that most "prefrontal cortex" claims are built on. Notably, Cook, Spivak and Berns (2016) — the study most often cited for canine prefrontal function — describe their result as localizing a frontal brain region, not as identifying a canine homologue of the human PFC.

This is not a reason to discard the framework. Frontal-limbic regulation is a deeply conserved mammalian arrangement, and the behavioral phenomena it predicts in dogs are real and measurable. It is a reason to treat anatomical specificity in dog training literature with caution: statements about which prefrontal subregion is doing what in a reactive dog are almost always imported from human neuroscience rather than derived from canine data (the same caution applies across canine neurobiology generally).

1.3 The One Canine Imaging Study of Inhibition

Cook, Spivak and Berns (2016) trained thirteen dogs to hold still in an MRI scanner and complete a go/no-go task, and separately tested them on an A-not-B task outside the scanner. A frontal region showed elevated activity during successful inhibition in all subjects, and dogs with greater mean activation in that region produced fewer false alarms — that is, they were less likely to respond when they should have withheld.

This is a genuine and important result: it links a canine brain measure to a canine behavioral measure of control, and it establishes individual differences at the neural level. It is also, as far as the canine imaging literature goes, essentially the only direct evidence of its kind, with thirteen scanner-trained dogs. Much of what follows in this article rests on behavioral evidence rather than on brain measurement, and the distinction is worth holding.

1.4 Why This Matters for Training

When a dog can access frontal control, learning is efficient, alternative behaviors can compete with impulses, and the dog can offer regulated behavior in mildly distracting settings. When that control is compromised — by arousal, by chronic stress, or by developmental stage — well-established cues stop working, and no amount of repetition in that state repairs them.

2. Impulse and Control: Two Systems in Competition

Behavior in any given moment reflects competition between fast, subcortically driven responses and slower regulatory processes.

2.1 The Speed Difference

Limbic structures, including the amygdala, generate defensive and appetitive responses within milliseconds of a trigger being detected. Frontal regulation is slower and conditional: it requires that the dog has previously learned an alternative, that baseline stress is not already high, and that arousal sits within a range the individual dog can work in.

That asymmetry is why the sequence matters so much in practice. By the time a handler has registered that another dog has appeared, the limbic response in their own dog has already begun; the question is only whether regulation catches up.

2.2 Why Frontal Control Fails Under Stress

The mechanistic account here is well documented — in rodents, monkeys, and humans. Arnsten (2009) reviews evidence that even mild uncontrollable stress rapidly impairs prefrontal function through excessive catecholamine signalling, and that sustained stress produces structural change in prefrontal dendrites. Glucocorticoids compound this over longer timescales.

None of this has been demonstrated in the dog brain. There is no canine study showing cortisol-driven dendritic loss or measured prefrontal volume reduction, and claims to that effect in training literature are extrapolations. What canine research does show is compatible: chronically stressed dogs behave in ways consistent with reduced regulatory capacity, and anxious dogs differ measurably in brain network organization (see §3.2) (the fuller account of chronic stress and cortisol in dogs).

2.3 What Happens During a Reactive Episode

The conventional description runs: trigger detected, limbic response initiated within milliseconds, sympathetic activation with catecholamine release, cardiovascular and respiratory changes, frontal regulation outcompeted, behavior produced without effective inhibition (as covered in detail for reactive dogs).

Read as a functional description of what the dog can and cannot do, this is well supported by behavior. Read as a claim about measured canine neural events in that moment, it is inference. The practical conclusion does not depend on the stronger reading: whichever mechanism is operating, a dog in that state is not withholding cooperation (behavior in that moment reflects a state, not a decision).

3. Why Dogs Lose Control

3.1 Arousal Beyond the Individual Optimum

The relationship between arousal and performance is usually invoked as a universal inverted U, attributed to Yerkes and Dodson (1908) — whose original result in mice was in fact conditional on task difficulty rather than general. The canine evidence is more interesting than the popular version. Bray, MacLean and Hare (2015) compared 106 pet dogs and assistance dogs bred and trained for low arousal on a detour inhibition task. Artificially increasing arousal improved inhibition in the low-baseline assistance dogs and impaired it in the higher-baseline pet dogs.

So the useful statement is not that arousal is the enemy of control. It is that each dog has a working range, that the range differs substantially between individuals, and that the encouragement which sharpens one dog pushes another past the point where regulation is available (which is what arousal regulation training is actually managing).

3.2 Chronic Stress

Prolonged, unpredictable, or uncontrollable stress is associated with reduced regulatory capacity. Xu et al. (2023) performed resting-state fMRI on 25 healthy and 13 anxious dogs, assessed with the C-BARQ, and found differences in functional network topology, including stronger amygdala connectivity within the anxiety-related network. Two limits apply: the design is cross-sectional and correlational, so it cannot establish that anxiety caused the network differences, and the paper carries a published 2025 correction to its methods section.

Where stress becomes not merely intense but uncontrollable, the relevant framework is the revised account of learned helplessness: passivity is the default mammalian response to prolonged uncontrollable adversity, and what is learned through frontal circuitry is the presence of control, which inhibits that default (Maier & Seligman, 2016) (the full revision of that account).

3.3 Incomplete Learning and Missing Generalization

Dogs are context-specific learners, and this is measurable rather than anecdotal. Bray, MacLean and Hare (2014) found that inhibitory control performance did not transfer cleanly across superficially similar tasks, and Brucks et al. (2017) reported that different measures of inhibitory control in dogs do not correlate with one another. A dog that withholds a response in one setup may not do so in another that looks equivalent to a human observer.

The practical implication is unforgiving: a behavior trained in the kitchen is a behavior trained in the kitchen (as the learned-behavior versus emotional-response distinction makes explicit).

3.4 Environmental Load

Sensory input that exceeds what the dog can process — density of dogs, noise, unpredictable movement, confinement on a short leash — reduces the resources available for regulation regardless of how well the behavior is trained. This is a management variable, not a training variable, and it is the one most often left unaddressed.

4. Development: How Self-Control Matures

4.1 What the Canine Data Actually Show

Two studies carry most of the weight here. Lazarowski et al. (2020) tested candidate detection dogs across the first year on four problem-solving tasks and found age-related improvement in measures related to inhibitory control, attention, and spatial cognition between 3 and 12 months. Bray et al. (2021) followed 160 candidate assistance dogs longitudinally, testing at 8–10 weeks and again at roughly 21 months, and found that performance improved with age with the largest effects on executive function measures.

So the canine evidence establishes clear developmental improvement in inhibition across the first two years. It does not establish a single point at which development stops.

4.2 Adolescence

The best canine evidence for an adolescent phase comes from Asher et al. (2020), who found reduced responsiveness to commands from the caregiver around eight months of age relative to five months, with the effect specific to the caregiver rather than generalized to an unfamiliar person. That is a behaviorally documented adolescent conflict phase in dogs, and it matches what owners report: a young dog that appears to have forgotten training, particularly with the person it is most attached to.

The popular explanation — a surge in limbic activation that outpaces frontal maturation, mirroring the human adolescent pattern — is plausible and consistent with the behavior, but it has not been measured in developing dogs.

4.3 The 2.5 to 3.5 Year Figure

A widely circulated claim holds that the canine prefrontal cortex is not fully mature until roughly 2.5 to 3.5 years of age, varying with breed and size. This figure is worth handling carefully, because it appears throughout training literature without a source.

It traces to veterinary behaviour commentary rather than to a neurodevelopmental study. No published imaging or histological work establishes a maturation endpoint for canine frontal cortex, and the peer-reviewed developmental data reach to about 21 months (Bray et al., 2021). The figure is best treated as a clinical rule of thumb from behavioral practice — one that usefully counters the assumption that a one-year-old dog is a finished adult, but which should not be presented as a measured neurological fact.

There is also reason to doubt that a single endpoint is the right model at all. Foraita, Howell and Bennett (2023), using a validated owner-report executive function scale across the canine lifespan, found that different components followed different age trajectories: working memory and owner-directed attention rose and later declined, while forms of inhibition moved in different directions from one another. Executive function in dogs does not appear to mature as one thing on one schedule.

4.4 Experience-Dependent Change

Repeated successful regulation, reinforced, is associated with improved control — and Foraita et al. (2023) found that dogs with more training history and working dogs scored higher on executive function independent of age.

At the neural level, Deshpande et al. (2024) scanned awake detection dogs at three points across training and identified a stable core network whose baseline connectivity predicted which dogs would succeed, alongside a second, more flexible network whose connectivity changes tracked behavioral change over training. The sample is small and the measure is functional connectivity rather than structural change, so "training-related plasticity" here is an inference from correlated change. It is nonetheless the closest thing available to canine neural evidence that training reshapes the systems involved (early experience shapes these systems too).

5. Building Self-Control Through Training

Effective training does not impose control; it creates the conditions under which control is possible and then reinforces it.

5.1 Train Below the Individual Threshold

Work where the dog can still process. Because the optimum differs between individuals (Bray et al., 2015), threshold has to be established empirically for the dog in front of you and re-established when fatigue, pain, or earlier events have shifted it.

5.2 Reinforce Regulated Choices

Reward the moment the dog disengages, checks in, or withholds a response, rather than only rewarding cued behaviors. This makes regulation itself the reinforced class of behavior (and how it is reinforced determines how durable it becomes).

5.3 Predictable Structure

Routines and consistent contingencies reduce uncertainty, which reduces baseline stress load. Predictability is not a comfort measure; unpredictability is specifically what the stress literature identifies as most costly.

5.4 Gradual Exposure and Explicit Generalization

Raise difficulty in steps, and train generalization deliberately across locations, distances, distractions, handlers, and body positions rather than assuming transfer. Given Bray et al. (2014) and Brucks et al. (2017), transfer should be treated as the exception rather than the default.

5.5 Emotional State Before Cognitive Demand

A dog carrying high stress load cannot reliably access regulation, so managing state comes before asking for control. That may mean environmental change, routine change, addressing pain, or veterinary involvement (pain in particular lowers behavioral tolerance).

5.6 Why Many Failures Are Not Technique Failures

A large share of training failure is not incorrect technique but a demand for cognitive control in a state where it is not available. No cue overrides a dog whose regulation is already outcompeted — asking for a sit from a dog that cannot take food is not a cue problem, it is an expectation problem.

This is why skilled trainers spend as much effort on arousal and environment as on behaviors. Self-control is not a command; it is a capacity that has to be protected while it is built (closely related to how frustration and failed impulse control operate).

6. Why Aversive Methods Undermine Self-Control

6.1 Stress Physiology and Frontal Function

Aversive methods raise stress hormones, and the mechanism by which sustained glucocorticoid and catecholamine elevation degrades prefrontal function is well characterized (Arnsten, 2009) — in rodents, monkeys, and humans. Applied to dogs, the expected consequence is a self-reinforcing loop: reduced regulation produces more reactive behavior, which prompts more correction. The loop is a reasonable inference; the canine version of the neural mechanism has not been measured.

6.2 What the Dog Learns Instead

The canine evidence is strongest at the behavioral level. Schilder and van der Borg (2004) compared guard dogs recently trained with electric shock against controls trained by similarly harsh non-electric means and found persistent differences in sessions where no shock was delivered, concluding that the shocked dogs had learned that their handler's presence and commands predicted shock. Herron, Shofer and Reisner (2009) surveyed 140 owners at a veterinary behavior referral service and found that confrontational techniques each drew aggressive responses from at least a quarter of the dogs on which they were used — a retrospective, self-reported study in a referral population, but the most direct data available.

What aversive methods reliably build is a predictive relationship between the handler and something unpleasant. That is the opposite of the condition self-control requires (and it can push behavior into shutdown rather than regulation).

6.3 The Welfare and Learning Evidence

Vieira de Castro et al. (2020) found aversive-trained companion dogs slower to approach ambiguous locations in a judgement bias test, indicating a more negative mood state. Casey et al. (2021) replicated the pattern with a matched-pair design and larger sample. Ziv's (2017) review of 17 studies concluded that aversive methods carry welfare costs without demonstrated effectiveness advantages, while noting the field's small samples and missing effect sizes.

Both judgement bias findings are correlational: owners who choose aversive methods may differ systematically, and dogs' full histories are rarely known (and stress interacts with anxiety in ways that compound this).

7. Practical Example: The Reactive Dog on Leash

A dog barks, lunges, and spins at the sight of another dog. The behavior is not a verdict on the dog's character or the handler's authority (nor is it a dominance problem). Functionally: a trigger is detected, a fast limbic response is generated, regulation is outcompeted because arousal is beyond this dog's working range, and the behavior occurs before an alternative can be selected.

7.1 What Does Not Work

Shouting or leash corrections raise arousal further and add an aversive event to a situation the dog already found difficult. Requiring a sit while the trigger closes demands regulation that is not currently available. Repeated exposure without changing the emotional response risks sensitization rather than habituation — and behavior that is merely suppressed tends to return (as the extinction literature predicts).

7.2 What Does

Lower arousal first, primarily through distance and by reducing baseline stress load. Change what the other dog predicts, pairing its appearance at sub-threshold intensity with something the dog values. Then build alternative responses — a check-in, a scatter search, a turn away — at a distance where the dog can still select them, so that a regulated option is available and reinforced.

Over time the reactive behavior diminishes, and it does so for two reasons at once: the emotional response to the trigger has changed, and a competing regulated response has been strengthened to the point where it can win (the same logic that applies to separation-related panic).

8. Summary: Self-Control at a Glance

Work below threshold — Rationale: regulation is only available within an arousal range that differs by individual (Bray et al., 2015). In practice: establish the range empirically, and re-check it when the dog is tired, sore, or has had a difficult day.

Reinforce regulated choices — Rationale: what gets reinforced is what becomes probable, and disengagement is a behavior like any other. In practice: mark and reward check-ins, disengagement, and withheld responses, not only cued actions.

Predictable structure — Rationale: unpredictability is the stressor dimension most associated with regulatory cost. In practice: consistent routines and contingencies, so the dog is not continuously resolving uncertainty.

Explicit generalization — Rationale: transfer across contexts is weak in dogs (Bray et al., 2014; Brucks et al., 2017). In practice: retrain across locations, distances, handlers, and distraction levels rather than assuming it carries.

State before demand — Rationale: a dog under high stress load cannot reliably access regulation. In practice: address environment, routine, and pain first (Mills et al., 2020).

Avoid aversive methods — Rationale: they raise stress in the systems control depends on, and condition the handler as a predictor of unpleasant events (Schilder & van der Borg, 2004; Ziv, 2017). In practice: reward-based methods, with management covering what training cannot yet handle.

Patience with development — Rationale: inhibition improves measurably across the first two years (Lazarowski et al., 2020; Bray et al., 2021), and adolescence brings a documented dip in responsiveness to the caregiver (Asher et al., 2020). In practice: expect regression around adolescence and keep criteria achievable.

9. Research Gaps and Critical Appraisal

Confidence varies sharply across this material.

The prefrontal framework is largely borrowed. Subdivision-level claims about canine prefrontal cortex are imported from human and rodent neuroscience. The canine frontal lobe is proportionally smaller and its homologies with primate prefrontal regions are not established. Cook et al. (2016) themselves report a frontal region, not a canine PFC homologue.

One imaging study carries most of the neural weight. Cook et al. (2016) scanned thirteen dogs. It is a careful study and the only direct canine neuroimaging evidence linking a brain measure to behavioral inhibition. Nothing in the canine literature replicates it at scale.

No canine evidence for stress-induced prefrontal damage. The dendritic and volumetric findings (Arnsten, 2009) are rodent, monkey, and human. Canine training studies measure behavior, cortisol, and judgement bias — not brain structure. Statements that aversive training shrinks a dog's prefrontal cortex are not supported by canine data.

"Inhibitory control" may not be one thing in dogs. Measures do not correlate across tasks (Brucks et al., 2017), performance is context-specific (Bray et al., 2014), higher inhibition helps on some tasks and hurts on others (Müller et al., 2016), and Olsen (2018) argues the field needs methodological overhaul. Treating self-control as a single trainable quantity oversimplifies.

The 2.5 to 3.5 year maturation figure lacks a primary source. It derives from veterinary behaviour commentary, not from neurodevelopmental measurement, and Foraita et al. (2023) suggest different executive components follow different trajectories rather than converging on one endpoint.

The connectivity studies are correlational and small. Xu et al. (2023) is cross-sectional with 13 anxious dogs and carries a published correction; Deshpande et al. (2024) infers plasticity from correlated connectivity change in a small working-dog cohort. Neither establishes causation.

Individual and breed variation is substantial. Baseline arousal, threshold, and developmental pace differ enough between dogs that population findings map loosely onto any individual (as breed-behavior research also shows).

10. Conclusion

Self-control in dogs is a capacity, not a character trait, and it is available only under certain conditions. When a dog is below its own arousal threshold, not carrying a heavy stress load, and has an alternative behavior worth producing, frontal regulation can outcompete a fast impulse. When those conditions fail, the same dog with the same training history cannot produce the same behavior — and this is a matter of state, not willingness. The detailed neuroscience behind that picture is largely borrowed from other species and should be held as such; the practical conclusions do not depend on it, because the canine behavioral evidence points the same way. Build control by protecting the conditions it requires: manageable arousal, low uncertainty, reinforced regulation, explicit generalization, pain ruled out, and methods that do not make the handler a predictor of trouble. That is slower than suppression and it is the only version that holds.

Key Insights (Takeaways)

• Self-control depends on frontal brain systems, but the detailed prefrontal model is borrowed. Dogs inhibit behavior and a frontal region tracks that inhibition (Cook et al., 2016, n = 13), yet canine frontal subdivisions have not been mapped onto human prefrontal regions, and most anatomical specificity in dog training literature is imported rather than measured.

• Arousal has an individual optimum, not a universal one. Raising arousal improved inhibition in calm assistance dogs and impaired it in more excitable pet dogs (Bray et al., 2015), so threshold must be established for each dog rather than assumed.

• Inhibition improves measurably across the first two years (Lazarowski et al., 2020; Bray et al., 2021), and adolescence brings a documented dip in responsiveness specifically toward the caregiver (Asher et al., 2020). The widely repeated "2.5 to 3.5 years" maturation figure comes from clinical commentary, not from neurodevelopmental measurement.

• "Inhibitory control" behaves less like a single trait than the training vocabulary suggests: measures do not correlate across tasks, performance is context-specific, and more inhibition is not uniformly better (Brucks et al., 2017; Bray et al., 2014; Müller et al., 2016; Olsen, 2018).

• Aversive methods work against the systems self-control depends on. Shock-trained dogs learned that the handler predicted pain (Schilder & van der Borg, 2004), confrontational techniques provoked aggression in a substantial minority of dogs (Herron et al., 2009), and aversive-trained dogs show a more pessimistic judgement bias (Vieira de Castro et al., 2020; Casey et al., 2021) — though the claim that this shrinks a dog's prefrontal cortex has no canine evidence behind it.

References

Arnsten, A. F. T. (2009). Stress signalling pathways that impair prefrontal cortex structure and function. Nature Reviews Neuroscience, 10(6), 410–422. https://doi.org/10.1038/nrn2648

Asher, L., England, G. C. W., Sommerville, R., & Harvey, N. D. (2020). Teenage dogs? Evidence for adolescent-phase conflict behaviour and an association between attachment to humans and pubertal timing in the domestic dog. Biology Letters, 16(5), 20200097. https://doi.org/10.1098/rsbl.2020.0097

Bray, E. E., Gruen, M. E., Gnanadesikan, G. E., Horschler, D. J., Levy, K. M., Kennedy, B. S., Hare, B. A., & MacLean, E. L. (2021). Dog cognitive development: A longitudinal study across the first 2 years of life. Animal Cognition, 24(2), 311–328. https://doi.org/10.1007/s10071-020-01443-7

Bray, E. E., MacLean, E. L., & Hare, B. A. (2014). Context specificity of inhibitory control in dogs. Animal Cognition, 17(1), 15–31. https://doi.org/10.1007/s10071-013-0633-z

Bray, E. E., MacLean, E. L., & Hare, B. A. (2015). Increasing arousal enhances inhibitory control in calm but not excitable dogs. Animal Cognition, 18(6), 1317–1329. https://doi.org/10.1007/s10071-015-0901-1

Brucks, D., Marshall-Pescini, S., Wallis, L. J., Huber, L., & Range, F. (2017). Measures of dogs' inhibitory control abilities do not correlate across tasks. Frontiers in Psychology, 8, 849. https://doi.org/10.3389/fpsyg.2017.00849

Casey, R. A., Naj-Oleari, M., Campbell, S., Mendl, M., & Blackwell, E. J. (2021). Dogs are more pessimistic if their owners use two or more aversive training methods. Scientific Reports, 11(1), 19023. https://doi.org/10.1038/s41598-021-97743-0

Cook, P. F., Spivak, M., & Berns, G. (2016). Neurobehavioral evidence for individual differences in canine cognitive control: An awake fMRI study. Animal Cognition, 19(5), 867–878. https://doi.org/10.1007/s10071-016-0983-4

Deshpande, G., Zhao, S., Waggoner, P., Beyers, R., Morrison, E., Huynh, N., Vodyanoy, V., Denney, T. S., & Katz, J. S. (2024). Two separate brain networks for predicting trainability and tracking training-related plasticity in working dogs. Animals, 14(7), 1082. https://doi.org/10.3390/ani14071082

Foraita, M., Howell, T., & Bennett, P. (2023). Executive functions as measured by the Dog Executive Function Scale (DEFS) over the lifespan of dogs. Animals, 13(3), 533. https://doi.org/10.3390/ani13030533

Herron, M. E., Shofer, F. S., & Reisner, I. R. (2009). Survey of the use and outcome of confrontational and non-confrontational training methods in client-owned dogs showing undesired behaviors. Applied Animal Behaviour Science, 117(1–2), 47–54. https://doi.org/10.1016/j.applanim.2008.12.011

Lazarowski, L., Krichbaum, S., Waggoner, L. P., & Katz, J. S. (2020). The development of problem-solving abilities in a population of candidate detection dogs (Canis familiaris). Animal Cognition, 23(4), 755–768. https://doi.org/10.1007/s10071-020-01387-y

Maier, S. F., & Seligman, M. E. P. (2016). Learned helplessness at fifty: Insights from neuroscience. Psychological Review, 123(4), 349–367. https://doi.org/10.1037/rev0000033

Mills, D. S., Demontigny-Bédard, I., Gruen, M., Klinck, M. P., McPeake, K. J., Barcelos, A. M., Hewison, L., Van Haevermaet, H., Denenberg, S., Hauser, H., Koch, C., Ballantyne, K., Wilson, C., Mathkari, C. V., Pounder, J., Garcia, E., Darder, P., Fatjó, J., & Levine, E. (2020). Pain and problem behavior in cats and dogs. Animals, 10(2), 318. https://doi.org/10.3390/ani10020318

Müller, C. A., Riemer, S., Virányi, Z., Huber, L., & Range, F. (2016). Inhibitory control, but not prolonged object-related experience appears to affect physical problem-solving performance of pet dogs. PLoS ONE, 11(2), e0147753. https://doi.org/10.1371/journal.pone.0147753

Olsen, M. R. (2018). A case for methodological overhaul and increased study of executive function in the domestic dog (Canis lupus familiaris). Animal Cognition, 21(2), 175–195. https://doi.org/10.1007/s10071-018-1162-6

Schilder, M. B. H., & van der Borg, J. A. M. (2004). Training dogs with help of the shock collar: Short and long term behavioural effects. Applied Animal Behaviour Science, 85(3–4), 319–334. https://doi.org/10.1016/j.applanim.2003.10.004

Vieira de Castro, A. C., Fuchs, D., Morello, G. M., Pastur, S., de Sousa, L., & Olsson, I. A. S. (2020). Does training method matter? Evidence for the negative impact of aversive-based methods on companion dog welfare. PLoS ONE, 15(12), e0225023. https://doi.org/10.1371/journal.pone.0225023

Xu, Y., Christiaen, E., De Witte, S., Chen, Q., Peremans, K., Saunders, J. H., Vanhove, C., & Baeken, C. (2023). Network analysis reveals abnormal functional brain circuitry in anxious dogs. PLoS ONE, 18(3), e0282087. https://doi.org/10.1371/journal.pone.0282087 (Correction published 2025, PLoS ONE, 20(3), e0320459. https://doi.org/10.1371/journal.pone.0320459)

Yerkes, R. M., & Dodson, J. D. (1908). The relation of strength of stimulus to rapidity of habit-formation. Journal of Comparative Neurology and Psychology, 18(5), 459–482. https://doi.org/10.1002/cne.920180503

Ziv, G. (2017). The effects of using aversive training methods in dogs — A review. Journal of Veterinary Behavior, 19, 50–60. https://doi.org/10.1016/j.jveb.2017.02.004