Learned Behavior vs. Emotional Response: Why Dog Training Sometimes Fails

1. Two Learning Systems, One Brain

Dogs learn through two parallel processes that run simultaneously and continuously influence each other. Almost every training failure that looks mysterious becomes legible once the two are told apart.

1.1 Operant Conditioning: Learning Behaviors

Operant conditioning, systematized by B. F. Skinner, is the process by which consequences change the future probability of a behavior. Reinforcement makes a behavior more likely; punishment makes it less likely. This is the system that most conventional training explicitly targets: the dog sits, food follows, sitting becomes more probable in that context.

Operant learning works best when the dog is regulated enough to notice the consequence, connect it to what it just did, and repeat the successful variant. It is highly effective for building new skills in low-distraction settings — and its efficiency depends on parameters that are themselves worth understanding (how reinforcement schedules shape persistence).

1.2 Classical Conditioning: Learning Emotions

Classical conditioning, described by Ivan Pavlov, is the process by which a previously neutral stimulus acquires emotional and physiological significance through repeated pairing with something that already matters. The result is a conditioned emotional response: automatic, involuntary, and not subject to the dog's decision.

A clicker paired repeatedly with food becomes a conditioned reinforcer that elicits anticipation. A person, a place, or a piece of equipment paired with pain or fright becomes a conditioned fear stimulus that elicits a defensive response on sight. This happens whether or not the trainer intends it, and it does not require the dog to notice, choose, or agree (how fear associations are formed, stored, and reopened).

1.3 Why the Two Systems Cannot Be Separated

In practice the two run together. While the trainer attends to the behavior being reinforced, the dog is simultaneously forming associations with the handler, the equipment, the location, and the cue itself. Every reinforcer and every correction carries a classical side-effect.

The clearest canine demonstration of this comes from shock-collar research. Schilder and van der Borg (2004) compared guard dogs recently trained with electric shocks against controls trained by similarly harsh but non-electric means. The shocked dogs showed lower ear carriage and more stress-related behavior, and — critically — the differences persisted in sessions and park walks where no shock was delivered. The authors concluded that these dogs had learned that the handler's presence and commands predicted shock, outside the original training context. The operant lesson was one thing; what the dogs learned emotionally was that their handler predicted pain.

This is also why a behavior can be fluent in the kitchen and absent on the street. The operant learning is intact. The emotional context has changed (behavior reflects an internal state, not a verdict on the dog).

2. Emotional Interference: Why Fear Overrides Training

The most common reason training fails is not a gap in the dog's knowledge. It is that fear, frustration, or over-arousal has made stored behavior temporarily inaccessible.

2.1 The Neural Account — and Where It Comes From

The standard account runs as follows. On perceiving a threat, the amygdala drives sympathetic activation within milliseconds; catecholamines are released, heart rate rises, and the body prepares for defensive action. At the same time, prefrontal regions supporting impulse control, flexible decision-making, and retrieval of trained responses lose functional capacity.

The prefrontal half of this is well documented. Arnsten (2009) summarizes evidence that even mild uncontrollable stress produces a rapid loss of prefrontal cognitive function, with prolonged exposure producing structural change in prefrontal dendrites — the mechanism being excessive catecholamine signaling that disconnects prefrontal networks. This work is from rodents, monkeys, and humans. It has not been replicated in fearful dogs, and the popular label "amygdala hijack" is a lay term rather than a technical description of a measured canine process. Canine neuroimaging exists but is small-sample and has largely addressed reward, voices, and faces rather than fear circuitry.

Treated with that caveat, the model is still the most parsimonious explanation available for a familiar clinical picture: a dog that performs fluently at home and cannot respond at all three metres from a trigger (the prefrontal basis of canine self-control). Asking for a sit at that moment is not a discipline problem. The circuits the behavior depends on are not available.

2.2 The Arousal–Performance Relationship

The relationship between arousal and performance is usually presented as an inverted U and attributed to the Yerkes–Dodson law. The attribution deserves precision. Yerkes and Dodson (1908) worked with Japanese dancing mice and electric shock, and their actual finding was conditional: the optimal stimulus intensity shifted with task difficulty, being lower for harder discriminations. The universal inverted U is a later simplification of a narrower result.

The directly relevant canine test is Bray, MacLean and Hare (2015). Comparing 106 pet dogs and assistance dogs bred and trained for low arousal on a detour inhibition task, they found that artificially increasing arousal improved performance in the low-baseline assistance dogs and impaired it in the higher-baseline pet dogs. The practical reading is not "arousal is bad." It is that each dog has an optimum, that the optimum differs between individuals, and that the same encouragement that sharpens one dog tips another over the edge (how arousal regulation works in practice). Threshold is an individual quantity, not a fixed line (as temperament and coping style research would predict).

2.3 Emotional States Are Not Obedience Problems

Fear and anxiety are physiological states that constrain attention, learning capacity, and behavioral options. A frightened dog is not electing to misbehave. Training built on the assumption that unwanted behavior is a choice — that the dog is ignoring, testing, or challenging — misidentifies the problem at the first step, and inherits a framework that the evidence no longer supports (why dominance fails as an explanatory concept).

Obedience cues do not resolve fear, because they operate on a different system. Emotional change requires emotional learning.

3. Why Punishment Makes Fear-Based Behavior Worse

When a dog fails to perform under stress, the intuitive response is to add pressure. For fear-driven behavior this is not merely ineffective; the evidence indicates it makes matters worse.

3.1 What the Dog Actually Learns

For a confident dog, a mild correction may suppress an unwanted behavior in the moment. For an anxious dog, the correction adds a second aversive event to a situation the dog already found threatening — confirming rather than correcting the appraisal. The functional lesson is that the trigger predicts bad outcomes, that the handler does not make things safer, and that low-level warning signals did not work.

Herron, Shofer and Reisner (2009) surveyed 140 owners attending a veterinary behavior referral service about techniques they had already tried. Confrontational interventions — striking or kicking, growling at the dog, physically forcing an item from the mouth, alpha rolls, staring the dog down — each drew an aggressive response from at least a quarter of the dogs on which they were used, and dogs presenting for aggression toward familiar people were significantly more likely to respond aggressively to alpha rolls and shouted corrections. Two limits travel with this finding: it is retrospective owner recall, and the sample was a referral population already selected for behavior problems. It is nonetheless the most direct data available on what happens when confrontation meets an already-anxious dog.

This is the ordinary route by which fear becomes aggression — not through dominance, but because an animal that cannot escape and whose quieter signals have failed escalates to what remains (the neurobiology behind escalation and frustration).

3.2 Fallout: Avoidance, Aggression, Shutdown, Compulsion

Fallout is the term for the unintended consequences of aversive methods, which typically surface well after the session and quietly reshape behavior and relationship.

Avoidance. Dogs learn to avoid not only the punished behavior but the context, the equipment, the location, and the person present. Schilder and van der Borg's finding that shocked dogs remained altered around their handler outside the training ground is precisely this pattern.

Aggression. Where avoidance is blocked — a leash, a corner, a confined space — defensive aggression becomes the available option, directed at the handler or redirected onto whatever else is present.

Shutdown. Some dogs stop offering behavior altogether, becoming subdued and easy to mistake for calm. This is usually described as learned helplessness, and the description needs updating. Overmier and Seligman (1967) originally interpreted such passivity as a learned expectation of futility. Maier and Seligman (2016) revised the account after five decades of neuroscience: passivity is the default mammalian response to prolonged uncontrollable adversity, and what is actually learned — via prefrontal circuitry detecting control — is the presence of control, which then inhibits that default. The dog has not learned to be helpless; it has failed to learn that it has any influence (the full revision of the learned-helplessness account). This reframing matters practically: the therapeutic target is restoring demonstrable control, not merely removing the stressor.

Compulsive behaviors. Repetitive tail-chasing, pacing, spinning, or fixation on lights and shadows are reported in animals kept under inconsistent or unpredictable aversive conditions, and are best treated as a signal of a chronic state rather than as a habit to be interrupted (how chronic stress reshapes the system).

3.3 The Welfare Evidence

Three lines of canine evidence converge. Vieira de Castro et al. (2020) compared companion dogs from reward-based and aversive-based training schools and found the aversive-trained group slower to approach ambiguous locations in a judgement bias test, indicating a more negative underlying mood state. Casey et al. (2021) replicated this with a matched-pair design and a larger sample: dogs whose owners reported using two or more aversive techniques showed the more pessimistic pattern. Cooper et al. (2014) and China et al. (2020) found no consistent training advantage for electronic collars over reward-based instruction, alongside welfare costs; Schalke et al. (2007) found the stress response depended heavily on whether the dog could predict and avoid the stimulus. Rooney and Cowan (2011) add a distinct and often-overlooked result: dogs of owners reporting more punishment were less likely to engage with a stranger and performed worse on a novel learning task — punishment did not merely damage welfare, it reduced learning ability.

Ziv's (2017) review of 17 studies reached the same conclusion while noting the field's genuine limitations: small samples, missing effect sizes, and possible observer bias (the full neurological picture of aversive fallout).

4. The Illusion of Suppression

One of the more consequential misconceptions in training is that visible obedience indicates emotional stability.

4.1 Masking

A dog can be taught to withhold the behaviors that communicate distress — growling, backing away, freezing and staring — while the underlying state is unchanged or worsening. What remains is a dog that complies and shows nothing, which is easy to read as progress. Practitioners call this masking. It is a clinical model rather than a directly tested canine construct, but it is consistent with the shock-collar data, in which dogs performing normally still carried measurable stress markers.

The cost is diagnostic as well as ethical: a dog that has learned not to warn removes the early signals that make a bite predictable and preventable (which is why objective behavioral measurement matters).

4.2 Why Suppressed Behavior Returns

Suppression holds a behavior below the surface without altering what elicits it. When the suppressing conditions weaken — the trigger comes closer, the dog is tired or in pain, the handler is absent — the behavior tends to return, often more intensely than before. The same principle applies to behavior reduced by non-reinforcement rather than by punishment: extinguished responses recover with time, with context change, and after a single reminder of the original association (why extinguished behavior comes back).

Durable confidence comes from predictable outcomes, available distance from threats, and the repeated experience that the dog's own behavior changes what happens to it. None of these is produced by enforcing compliance.

5. Changing the Emotion, Not Just the Behavior

If the driver of a behavior is an emotional state, an intervention aimed only at the behavior addresses the wrong variable.

5.1 Classical Counterconditioning and Desensitization

Counterconditioning changes what a stimulus predicts, pairing it with something the dog values so that the conditioned response shifts from fear toward anticipation or neutrality. Systematic desensitization controls intensity, presenting the trigger at a level low enough that the fear response is not triggered at all, then raising it gradually.

The canine evidence is real and bounded. Butler, Sargisson and Elliffe (2011) ran a controlled within-subjects trial of systematic desensitization for separation-related behavior in eight dogs, finding significant reductions in both frequency and severity, with near-complete resolution at three-month follow-up in the six dogs followed up. Shnookal et al. (2024) conducted a PRISMA systematic review of counterconditioning-based interventions and found mostly positive outcomes, particularly for kennel behaviors and aggression toward other dogs — while flagging inconsistent definitions across studies, heavy reliance on expert implementation, small samples, and separation-related behaviors proving comparatively resistant to change. Riemer's (2023) practitioner review of noise fears reaches a compatible conclusion, noting that counterconditioning to real-life noises and relaxation training performed particularly well.

The practical limitation is the one owners actually hit: the protocol requires reproducing the trigger at a controlled intensity, which for thunderstorms, strangers, or traffic is difficult and slow.

5.2 Operant Learning for Choice and Agency

Where classical procedures change the emotional valence, operant procedures restore influence. Targeting, station training, cued interaction, and consent-based husbandry give the dog a clear and repeatable way to affect what happens next.

This is not a soft addition. If the revised learned-helplessness account is right, the detection of control is the active ingredient in recovery from uncontrollable stress rather than a pleasant extra (how behavioral flexibility is rebuilt). Note also that the mechanistic basis of that claim is rodent work; the canine version is inference from a consistent clinical picture.

5.3 Working Below Threshold

Working below threshold means keeping the dog at an arousal level where learning is still possible. Concretely: reduce exposure intensity, increase distance, pair triggers with safety rather than pressure, allow the dog to leave, and teach coping responses rather than only cues.

Bray et al. (2015) is the reason to treat threshold as individual rather than universal. There is no single correct distance or duration; there is a level at which this dog can still process, which has to be established empirically and will shift with fatigue, pain, and prior events that day.

5.4 What Happens After the Session

Emotional context does not only affect performance during training; it affects what is retained afterwards. Reicher et al. (2024) trained 24 family dogs under a controlling and a permissive style, then recorded non-invasive sleep EEG and retested. Dogs that experienced the more rewarding-than-expected condition — a positive expectancy violation — showed improved performance after sleeping. This is a single study with a modest sample and should be held provisionally, but it points in a consistent direction: the affective quality of a session is part of what is consolidated, not a side-effect of it (how prediction error drives what gets learned).

6. Common Reasons Training Fails Even When the Dog Knows the Behavior

6.1 Emotional Interference

The dog is too frightened, frustrated, or aroused to access what it has learned. This is the most frequent cause and the one most often mistaken for disobedience (particularly in reactive dogs).

6.2 Lack of Generalization

Dogs are context-specific learners, and this is measurable rather than anecdotal: Bray, MacLean and Hare (2014) found that inhibitory control performance in dogs did not transfer cleanly across superficially similar tasks. A behavior trained in a quiet room is a behavior trained in a quiet room. Generalization has to be built deliberately across locations, distractions, handlers, and body positions.

6.3 Poisoned Cues

A cue that has been followed by unpleasant consequences can come to predict them, producing hesitation, conflict, or refusal. The term is a practitioner concept without a dedicated experimental literature in dogs, but the underlying process is exactly what Schilder and van der Borg documented at the level of the handler and the command.

6.4 Unclear Cues and Delayed Feedback

Cues diluted by repetition, emotional tone, and running commentary carry less information. Feedback that arrives late is weaker feedback: the standard practice rule of marking within roughly one to two seconds derives from operant research in rats and pigeons plus field observation of owners training dogs, and controlled canine work on reinforcement delay remains sparse and largely reported in conference abstracts and unpublished formats. Treat it as a well-motivated convention rather than a canine-validated constant.

6.5 Skipping Steps

Advancing before the current step is fluent produces plateaus and regression. The failure is usually visible one criterion earlier than where it shows up.

6.6 Underlying Medical Problems

Pain, sensory decline, and neurological disease lower tolerance and raise reactivity, and the association between pain and problem behavior is well documented in the veterinary literature (Mills et al., 2020). Medical assessment belongs before behavior modification, not after it fails (particularly for musculoskeletal pain and aggression).

7. From Compliance to Emotional Change

The most useful shift available to a trainer or owner is a change of question: from how to get the dog to comply, to why the dog does not currently feel safe. Safety is not something a dog earns through obedience; it is the condition under which learning becomes possible at all.

7.1 Principles of Emotion-Focused Training

Change the emotional response first, using counterconditioning and controlled exposure, and let behavior follow. Keep the dog below its individual threshold, established empirically rather than assumed. Give the dog real influence over outcomes through choice-based and consent-based procedures. Train generalization explicitly across contexts. Use reward-based methods, which the comparative evidence supports on both welfare and effectiveness grounds. Attend to the handler's own state: Parr-Cortes et al. (2024) found that the odour of an unfamiliar stressed person shifted dogs' responses in a cognitive bias test, and secure attachment to the handler is associated with better cortisol regulation in dogs during stressful separations (Schöberl et al., 2016), which makes the handler part of the environment being trained (how emotional states transfer between human and dog). Finally, rule out medical contributors before concluding that a problem is behavioral.

8. Summary: Behavior and Emotion at a Glance

Operant conditioning — Mechanism: behavior followed by a consequence that changes its future probability. What it changes: the likelihood of a specific action. Works when: the dog is below threshold and can process feedback. Limits: does not directly alter the emotional response, and is overridden at high arousal.

Classical conditioning — Mechanism: a neutral stimulus repeatedly paired with a significant event acquires an automatic emotional response. What it changes: what a stimulus means to the dog. Works when: applied systematically below threshold as counterconditioning and desensitization. Limits: requires controlled, repeatable exposure, which many real-world triggers do not permit.

Emotional interference — Mechanism: high arousal or fear with amygdala activation and impaired prefrontal function, extrapolated from rodent, primate, and human work. What it causes: trained behavior becomes temporarily inaccessible despite being intact. Prevention: work below the individual threshold, manage arousal, prioritize predictability.

Fallout from aversive methods — Mechanism: aversive events condition fear to the handler, the context, and the equipment, not only to the targeted behavior. What it causes: avoidance, defensive aggression, shutdown, compulsive behavior, and measurably reduced learning ability. Prevention: reward-based methods, and treating the emotional state as the target.

9. Research Gaps and Critical Appraisal

Confidence varies considerably across this material, and the divisions matter.

The neural mechanism is extrapolated, not canine-measured. The amygdala–prefrontal account rests on rodent, primate, and human work (Arnsten, 2009). No study has imaged the prefrontal cortex of a fearful dog under a real-world trigger. The model is well-founded and explanatory, but it is inference.

"Amygdala hijack" is a popular label. It is not a technical term for a described canine mechanism, and its confident use in training literature outruns the evidence considerably.

The inverted-U is a simplification with one good canine test. Yerkes and Dodson (1908) reported a task-difficulty-dependent effect in mice, not a universal law. Bray et al. (2015) is the direct canine test and its result complicates the popular version: the effect of raising arousal reversed sign depending on the dog's baseline.

The welfare findings are robust in direction but correlational. Vieira de Castro et al. (2020) and Casey et al. (2021) both establish association rather than causation; owners who choose aversive methods may differ systematically, and dogs' histories are rarely fully known. Casey et al. state this limitation explicitly.

Judgement bias tests are not a clean read-out of mood. Krahn et al. (2024) showed that prior discrimination training alters dogs' subsequent performance in a judgement bias test, meaning learning history contributes independently of welfare. This qualifies, without overturning, the cognitive-bias evidence on training methods.

The counterconditioning evidence base is positive but thin. Shnookal et al. (2024) identified only 14 eligible studies, with inconsistent definitions, small samples, and dependence on expert implementation; Butler et al. (2011) had eight dogs. Effectiveness is supported; effect sizes and generalizability are not well established.

Masking and poisoned cues are clinical models. Both are coherent and consistent with the shock-collar data, but neither has a dedicated experimental literature in dogs.

Individual variation dominates. Threshold, arousal optimum, and recovery rate differ substantially between dogs, so population-level findings map only loosely onto any individual animal.

10. Conclusion

Training fails when it addresses what a dog does while ignoring how the dog feels. A dog can hold a behavior perfectly and still be unable to produce it when fear, frustration, or arousal has taken the relevant systems offline. That is not defiance, and adding pressure at that moment only adds a second aversive event to a situation the dog already found intolerable. Lasting change comes from altering what the trigger predicts, working at an intensity the individual dog can still process, restoring the dog's influence over outcomes, and ruling out pain first. The underlying neuroscience is borrowed and should be held as such — but the practical conclusions do not depend on it, because the canine behavioral and welfare evidence points the same way on its own.

Key Insights (Takeaways)

• A learned behavior and an emotional state are different things, produced by different processes and stored differently. A dog can know a cue perfectly and be unable to perform it under fear or high arousal, and the canine anchor for this is the shock-collar work showing that aversive training conditions a response to the handler and context rather than only to the targeted behavior (Schilder & van der Borg, 2004).

• The amygdala–prefrontal explanation for why fear blocks trained behavior is drawn from rodent, primate, and human research (Arnsten, 2009) and has not been measured in fearful dogs. It remains the best available account, but "amygdala hijack" is a lay label, not a described canine mechanism.

• Threshold is individual, not universal. Raising arousal improved inhibitory control in calm assistance dogs and impaired it in more excitable pet dogs (Bray et al., 2015), and the original Yerkes–Dodson result was conditional on task difficulty rather than a general law.

• Punishment worsens fear-based behavior and reduces learning capacity. Confrontational techniques drew aggressive responses from a substantial minority of dogs in a referral population (Herron et al., 2009), aversive-trained dogs show a more pessimistic judgement bias (Vieira de Castro et al., 2020; Casey et al., 2021), and dogs of owners using more punishment performed worse on a novel learning task (Rooney & Cowan, 2011) — though these are associations, not demonstrated causes.

• Emotional change requires emotional learning. Counterconditioning and systematic desensitization have controlled support in dogs (Butler et al., 2011; Shnookal et al., 2024), with the honest caveats being small samples and expert dependence. Practically: change what the trigger predicts, stay below the individual threshold, give the dog real control rather than only cues, train generalization deliberately, and rule out pain first (Mills et al., 2020).

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

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

Butler, R., Sargisson, R. J., & Elliffe, D. (2011). The efficacy of systematic desensitization for treating the separation-related problem behaviour of domestic dogs. Applied Animal Behaviour Science, 129(2–4), 136–145. https://doi.org/10.1016/j.applanim.2010.11.001

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

China, L., Mills, D. S., & Cooper, J. J. (2020). Efficacy of dog training with and without remote electronic collars vs. a focus on positive reinforcement. Frontiers in Veterinary Science, 7, 508. https://doi.org/10.3389/fvets.2020.00508

Cooper, J. J., Cracknell, N., Hardiman, J., Wright, H., & Mills, D. (2014). The welfare consequences and efficacy of training pet dogs with remote electronic training collars in comparison to reward based training. PLoS ONE, 9(9), e102722. https://doi.org/10.1371/journal.pone.0102722

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

Krahn, J., Azadian, A., Cavalli, C., Miller, J., & Protopopova, A. (2024). Effect of pre-session discrimination training on performance in a judgement bias test in dogs. Animal Cognition, 27(1), 66. https://doi.org/10.1007/s10071-024-01905-2

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

Overmier, J. B., & Seligman, M. E. P. (1967). Effects of inescapable shock upon subsequent escape and avoidance responding. Journal of Comparative and Physiological Psychology, 63(1), 28–33. https://doi.org/10.1037/h0024166

Parr-Cortes, Z., Müller, C. T., Talas, L., Mendl, M., Guest, C., & Rooney, N. J. (2024). The odour of an unfamiliar stressed or relaxed person affects dogs' responses to a cognitive bias test. Scientific Reports, 14(1), 15843. https://doi.org/10.1038/s41598-024-66147-1

Reicher, V., Kovács, T., Csibra, B., & Gácsi, M. (2024). Potential interactive effect of positive expectancy violation and sleep on memory consolidation in dogs. Scientific Reports, 14(1), 9487. https://doi.org/10.1038/s41598-024-60166-8

Riemer, S. (2023). Therapy and prevention of noise fears in dogs — A review of the current evidence for practitioners. Animals, 13(23), 3664. https://doi.org/10.3390/ani13233664

Rooney, N. J., & Cowan, S. (2011). Training methods and owner–dog interactions: Links with dog behaviour and learning ability. Applied Animal Behaviour Science, 132(3–4), 169–177. https://doi.org/10.1016/j.applanim.2011.03.007

Schalke, E., Stichnoth, J., Ott, S., & Jones-Baade, R. (2007). Clinical signs caused by the use of electric training collars on dogs in everyday life situations. Applied Animal Behaviour Science, 105(4), 369–380. https://doi.org/10.1016/j.applanim.2006.11.002

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

Schöberl, I., Beetz, A., Solomon, J., Wedl, M., Gee, N., & Kotrschal, K. (2016). Social factors influencing cortisol modulation in dogs during a Strange Situation Procedure. Journal of Veterinary Behavior, 11, 77–85. https://doi.org/10.1016/j.jveb.2015.12.003

Shnookal, J., Tepper, D., Howell, T., & Bennett, P. (2024). Counterconditioning-based interventions for companion dog behavioural modification: A systematic review. Applied Animal Behaviour Science, 276, 106305. https://doi.org/10.1016/j.applanim.2024.106305

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

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