Social Referencing and Social Buffering in Dogs

1. Two Ways a Person Changes a Dog's Response

1.1 Information and Support

People influence how dogs respond to difficult situations in two distinguishable ways. They can supply information — the dog looks to a person and uses what it sees or hears to decide how to treat something uncertain. And they can supply support — the mere presence of a familiar person changes how strongly the dog reacts to a challenge, whatever that person does.

The first is called social referencing, the second social buffering. Both terms come from developmental psychology, both have been tested in dogs, and both are regularly invoked in training advice with more confidence than the studies support.

1.2 Why They Belong Together

In everyday situations the two can overlap. A dog meeting something unfamiliar with its owner beside it is being buffered by the owner's presence and can reference the owner's reaction at the same time. Separating them matters because they imply different things for practice: one depends on what the person expresses, the other on who the person is.

1.3 How to Read the Evidence

The canine literature on both is small, mostly experimental and mostly based on short laboratory encounters. That makes the designs clean and the conclusions narrow. Each section below states what was manipulated and what was measured, because in this area the wording of popular summaries tends to drift furthest from the data (why the measure determines the claim).

1.4 Why the Topic Matters for Trainers

Much everyday handling advice rests implicitly on one or both mechanisms. "Stay calm so your dog stays calm" touches on whether dogs use human signals. "Don't leave your dog alone at the vet" touches on buffering. "Walk past confidently" assumes that the handler's movement can carry information. Each is plausible; the question this article asks is which parts of each have been tested and with what result.

The value of the evidence lies in being more specific than general advice. It identifies which signal carried the effect, which person produced it, and in which kind of situation it was observed.

1.5 Terms Borrowed From Child Development

Both concepts were developed to describe human infants and their caregivers, and the canine studies deliberately copied infant procedures. That makes comparisons possible and invites overinterpretation. A dog that behaves in a way that resembles infant referencing has shown a functional similarity, not that the underlying processes are the same.

The articles discussed here are careful about that distinction. Popular summaries of them often are not, which is why the wording in the following sections stays close to what was measured.

2. Social Referencing

2.1 The Concept

Social referencing describes using another individual's emotional information about an ambiguous object or situation to guide one's own behavior toward it. In human infants it has three components: looking to the other person when confronted with something uncertain, adjusting behavior according to the emotional message received, and learning something lasting about the object from it.

Each component can be present without the others, which is why the dog studies test them separately.

2.2 The First Test

In the first canine study, adult dogs were confronted with a potentially frightening object — an electric fan with ribbons attached — while their owner was present. Most dogs, 83 percent, looked referentially to the owner after looking at the object, which is consistent with referential looking interpreted as information seeking (Merola, Prato-Previde & Marshall-Pescini, 2012b).

When owners then delivered a positive or negative message using only voice and facial expression, the dogs' behavior differed little between the two groups. When owners also moved — approaching the fan or retreating from it — the difference in behavior was clear: dogs in the positive group moved closer to the fan and dogs in the negative group moved away, mirroring the owner's action (Merola et al., 2012b).

2.3 What That Pattern Means

The first study found evidence for referential looking and for behavioral adjustment following the owner's actions, and weak evidence for adjusting behavior on the basis of vocal and facial emotion alone. That is a more modest result than "dogs read our emotions and act on them", and it is the one the data support.

The distinction matters for practice. A person's body movement toward or away from something was a far stronger signal in this study than the tone of voice or the expression on the face.

2.4 Owner Versus Stranger

A second study refined the design: the object was switched off during the message, following the procedure used with infants, and either the owner or a stranger acted as the informant, delivering a positive or negative message. Most dogs looked referentially to the informant regardless of who it was; gaze alternation between the fan and the informant was shown by 62 percent of dogs with the owner and 52 percent with the stranger (Merola, Prato-Previde & Marshall-Pescini, 2012a).

Behavioral regulation, by contrast, depended on the relationship. With the owner as informant, dogs receiving a positive message looked at the owner more often and spent more time approaching and close to the object, while dogs receiving a negative message took longer to approach and interact with it. Fewer differences appeared when the stranger delivered the message (Merola et al., 2012a).

2.5 How Strong the Evidence Is

Two studies from one research group, one type of ambiguous object, adult pet dogs in a laboratory room, and effects that were clearer for some measures than for others. The findings are consistent with each other and with the idea that dogs use human emotional information, especially from familiar people. They do not establish how general the ability is, how it varies between individuals, or how it operates in the far noisier situations of daily life.

2.6 Why Movement Might Carry More Than Voice

An owner approaching an object does two things at once: expresses an attitude and demonstrates, in the most concrete way possible, that the object is safe to be near. A dog watching that has direct evidence about the situation, not just an emotional cue. Voice and facial expression convey attitude only, and dogs may attend to them less reliably, particularly when the object itself is noisy or moving.

That is an interpretation rather than a finding. It is consistent with the results and it explains why "just talk to your dog in a happy voice" is a weaker intervention than walking toward the thing yourself (Merola et al., 2012b).

2.7 Referencing Is Not Obedience

Looking to a person in an uncertain situation is sometimes confused with checking in for instructions. The two may overlap in trained dogs, but referencing as tested in these studies concerns information about the object, not about what the dog is expected to do. A dog can reference its owner and still decide for itself how to respond.

2.8 What the Dog Does Next

The practical value of referencing lies in what happens after the look. A dog that glances at its handler and sees a calm approach has received information that makes approaching easier. A dog that glances and sees a startle, a sudden tension on the lead or a retreat has received the opposite. Most handlers are not aware of producing either signal, which is why the first practical step is simply noticing the glance (how reactive moments unfold).

2.9 Developmental Questions

In human infants, social referencing emerges at a particular point in development and changes as children grow. Whether the same is true in dogs — whether puppies reference differently from adults, and whether adolescence changes the pattern — has not been tested in the studies discussed here, which used adult dogs.

The question is practically interesting because puppies meet more new things than adult dogs do, and the information they receive from people during that period could shape later responses. It remains a plausible hypothesis rather than a finding (how early development is organized).

2.10 Individual Differences in Looking

Not every dog looked to its owner: 83 percent did in the first study, which leaves a substantial minority that did not (Merola et al., 2012b). Dogs differ in how much they attend to people generally, and a dog that rarely references its handler receives less information from that source, whatever the handler does.

For such dogs, the handler's influence may run more through buffering than through referencing, and through shaping the situation rather than through signalling about it (how individual differences are described).

2.11 Positive and Negative Messages Are Not Symmetrical

In the second study, the owner's positive message produced the clearest differences: dogs looked at the owner more, approached the object and spent more time near it. The negative message produced slower approach and less interaction (Merola et al., 2012a). Both directions were present, and the design cannot say whether dogs are generally more responsive to one kind of message than the other.

For practice the asymmetry that matters is different. A handler rarely delivers a deliberate negative message; what happens is an unintended one — a flinch, a tight lead, a hurried retreat. Because dogs reference their handler's movement, those involuntary signals are the ones most worth managing.

3. Social Buffering

3.1 The Concept

Social buffering describes the reduction of an individual's stress response by the presence of a social partner. It does not require the partner to do anything in particular, which distinguishes it from comforting or from active reassurance.

3.2 Human Versus Canine Companion

In a study of adult kennel-housed dogs, removing a long-standing kennel mate from the home run for four hours had no effect on the remaining dog's behavior or plasma glucocorticoid levels. In a novel environment, dogs showed increased activity and elevated glucocorticoid levels — and these responses were as large when the kennel mate was present as when the dog was alone. When the dogs were in the novel environment with their human caretaker, activity and glucocorticoid levels were not elevated (Tuber, Hennessy, Sanders & Miller, 1996).

The dogs also spent more time near the human and solicited social behavior from the human more often than from the kennel mate. For these dogs, the familiar person buffered the response to novelty and the familiar dog did not.

3.3 The Owner in the Face of a Threat

A later study tested thirty dogs approached by a threatening stranger, once while separated from the owner and once in the owner's presence, in balanced order, with telemetric heart-rate recording. The threatening approach raised heart rate, and the rise was significantly less pronounced when the owner was present (Gácsi, Maros, Sernkvist, Faragó & Miklósi, 2013).

The order of the two encounters also mattered: dogs that met the stranger first with the owner present showed an attenuated heart-rate increase when they later met the stranger alone. The authors interpreted this as a safe-haven effect analogous to the one described in human attachment research.

3.4 Separation Without a Heart-Rate Change

The same study reported something easily overlooked. Separation from the owner produced characteristic distress vocalizations in many dogs but did not change mean heart rate; it did increase heart rate variability (Gácsi et al., 2013). Behavior and cardiac measures therefore pointed in different directions during separation, which is a reminder that no single indicator captures the whole response (why stress measures often disagree).

3.5 Buffering Frustration

Social buffering has also been tested against a different kind of challenge: the frustration that follows when an expected reward stops coming. Dogs trained to gaze at an experimenter's face to obtain food then experienced extinction, with food no longer delivered. Frustration responses increased when food was withheld, and the presence of the owner reduced them; dogs with the owner present also persisted longer in gazing at the experimenter (Dzik, Jousset, Vietti & Bentosela, 2025).

That extends the concept from fear and novelty to reward omission, which is exactly the situation that arises whenever a behavior is deliberately no longer reinforced (how frustration is described).

3.6 Unfamiliar People

Familiarity is not the only route. In dogs newly admitted to a shelter, thirty-minute sessions with a person — passive presence, petting or play — produced a substantial and near uniform reduction in plasma cortisol compared with control conditions, along with less vocalizing and panting (Shiverdecker, Schiml & Hennessy, 2013).

That population and setting differ considerably from pet dogs at home, and the finding is included here only for its mechanism: human contact as such can lower the physiological response of a stressed dog, even when the person is not an attachment figure.

3.7 Frustration Is Worth Taking Seriously

The extinction procedure used in the buffering study builds on earlier work from the same research tradition, which showed that when a previously reinforced gaze toward a human stops producing food, dogs move away, orient sideways to the person, lie down, walk around, sniff and vocalize more (Jakovcevic, Elgier, Mustaca & Bentosela, 2013). Those are ordinary signs of frustration, and they appear whenever a behavior is deliberately no longer rewarded.

That makes the buffering result more than a curiosity. Extinction is a routine part of many training plans, and the evidence that a familiar person's presence reduces the frustration it causes is directly relevant to how such plans are run (Dzik et al., 2025).

3.8 Why a Familiar Dog Did Not Help

The finding that a long-standing kennel mate did not buffer the response to novelty, while a human caretaker did, is striking and easy to overgeneralize (Tuber et al., 1996). The dogs were adult laboratory-housed animals with a specific history of human care, and the result says something about those relationships rather than about the value of canine company in general.

Canine companions can matter in other ways and in other populations. What the study shows is that for these dogs, in this situation, the person carried the effect that the dog did not.

3.9 What Buffering Changes in the Body

The buffering studies measured glucocorticoids, heart rate and heart rate variability, and behavior. Each captures activation rather than how the dog experienced the situation, and they do not always move together (what these measures capture). The consistent direction across the studies — lower responses with a familiar person present — is more informative than any single number.

Oxytocin is often proposed as the mechanism behind social buffering. That is a reasonable hypothesis with support from other species, and none of the canine studies discussed here measured it (what is known about oxytocin in dogs).

3.10 The Order Effect

The most practically interesting detail in the safe-haven study is the order effect: dogs that first met the threatening stranger with the owner present showed a smaller heart-rate increase when they later met the same stranger alone (Gácsi et al., 2013). If that holds, a first buffered encounter might shape how a later unbuffered one is experienced.

It is one finding in thirty dogs, with a single type of stimulus and a short interval between encounters. It is worth knowing about as a lead and not strong enough to build a protocol on.

3.11 What Counts as a Familiar Person

The buffering studies used the dog's owner or long-term caretaker. Whether other people the dog knows well — a regular dog sitter, a trainer, a family member who is not the main carer — produce similar effects is not established by these studies. The attachment framework predicts a graded effect with the strength of the relationship, which is plausible and untested.

For practice it suggests that, where the owner cannot be present, the next best option is a person the dog already knows and trusts rather than a stranger, while remembering that even unfamiliar people reduced cortisol in the shelter study (Shiverdecker et al., 2013).

3.12 Human Interaction in Stressful Settings

The shelter study found that passive human presence was associated with reduced cortisol, with effects comparable to some active interaction conditions (Shiverdecker et al., 2013). That is a notable detail for anyone designing interventions: the results suggest that, in this situation, not only the kind of contact but the mere presence of a calm person may play a role.

It fits the definition of buffering as a presence effect, and it suggests that in difficult situations a quiet, available person may help more than an attempt to distract or cheer the dog up.

4. What the Two Have in Common

4.1 The Relationship Matters in Both

In referencing, dogs adjusted their behavior more when the owner rather than a stranger delivered the message (Merola et al., 2012a). In buffering, the human caretaker reduced the response to novelty where the kennel mate did not (Tuber et al., 1996), and the owner reduced frustration in the extinction study (Dzik et al., 2025). A recurring pattern is that the identity and familiarity of the person can influence the effect in the tested contexts.

4.2 The Link to Attachment

Both phenomena are predicted by attachment theory: an attachment figure serves as a source of information in uncertain situations and as a safe haven under threat. The canine findings fit that framework without proving that the dog–owner relationship is attachment in the full developmental sense (how attachment is studied in dogs).

4.3 Where They Differ

Referencing depends on what the person does: the message, and above all the person's movement toward or away from something. Buffering depends mainly on the person being there. For practice that means two different levers: what a handler expresses, and whether the handler is present at all.

4.4 When the Owner Is the Problem

Both mechanisms assume that the person is a source of safety. Where the dog has learned to associate its handler with unpredictable or aversive treatment, the same presence and the same glance may carry the opposite information. That has not been directly tested in these studies and remains a hypothesis based on broader attachment theory (what aversive handling costs).

4.5 Stranger Effects Are Not Zero

Strangers also received referential looks, and dogs responded to some extent to a stranger's message (Merola et al., 2012a). In shelter dogs, contact with unfamiliar people reduced cortisol (Shiverdecker et al., 2013). Familiarity strengthens both effects; it is not the only route to them.

4.6 Two Mechanisms, One Situation

In real encounters the two effects are hard to separate. A dog meeting a new object with its owner present is buffered by that presence and can reference the owner's reaction at the same time. The laboratory designs separate them by holding one constant while varying the other; everyday life never does.

For practice the separation is still useful, because it shows which lever is being pulled. Staying close supplies presence. Moving calmly toward something supplies information. Both can be used deliberately, and they reinforce each other when they point the same way.

4.7 The Place of Controllability

Both mechanisms can influence how a dog responds to uncertainty. Referencing supplies information about what something is; buffering supplies a stable element in an unstable situation. That connects them to the wider evidence that predictability and control reduce the impact of challenges, and it explains why a calm, predictable handler is doubly useful (why predictability matters).

5. What the Evidence Does Not Show

5.1 Emotions Are Not Transmitted Wholesale

The first referencing study found little effect of voice and facial expression alone (Merola et al., 2012b). Claims that dogs automatically take on their owner's emotional state in every situation go considerably beyond this literature (what is known about emotional contagion).

5.2 Presence Is Not a Treatment

Buffering reduces a response during a challenge. None of the studies shows that repeated buffered exposure resolves a fear, and a dog that copes only when the owner is present may simply be managing rather than learning that the situation is safe. The order effect in the safe-haven study is suggestive in this respect, and it is a single finding in thirty dogs (Gácsi et al., 2013).

5.3 Not Every Dog, Not Every Owner

All results are group effects. Individual dogs differ in how much they look to people, and owners differ in how much reassurance their presence provides. None of the studies was designed to predict the effect for an individual dog–owner pair.

5.4 Buffering Is Not Desensitization

A dog that tolerates the vet with its owner beside it has received support during a difficult event. That is valuable, and it is different from a dog whose response to the vet has changed. Changing the response requires the ordinary work of gradual exposure and pairing with something good, which buffering can support but does not replace (how fear responses are changed).

5.5 A Note on Separation

If a familiar person reduces a dog's stress response, the absence of that person removes the reduction. For dogs that cope poorly when left alone, heavy reliance on presence-based support in all difficult situations may make absences harder rather than easier. That is a reasoned concern rather than a demonstrated effect, and it argues for balance rather than for withholding support (what separation-related behavior involves).

5.6 Lead Tension as Information

On a walk, the lead transmits the handler's movement directly. A sudden tightening when another dog appears is, from the dog's side, a movement signal from the person it references most, delivered at the moment the dog is uncertain. None of the studies tested lead tension specifically, and it follows directly from the finding that owner movement was the strongest referencing cue (Merola et al., 2012b).

Practising a loose-lead routine for the moment something unexpected appears is therefore not only a handling skill. It changes what the dog learns from the handler in exactly the situations where it is looking.

5.7 Why Pretending Is Not the Answer

A common version of the referencing advice tells handlers to act calm regardless of how they feel. The studies do not test acted calm, and whether dogs distinguish performed from genuine composure is unknown. What the data do support is that the handler's movement matters most (Merola et al., 2012b).

That gives a more workable target than managing one's emotions on demand: rehearse what the body does when something unexpected appears — keep the lead loose, stop or walk on steadily, avoid sudden retreats — until it happens without having to think about it. A practised routine is more reliable than an effort to feel differently in the moment.

6. What Follows for Practice

6.1 Your Movement Is the Strongest Signal

In the first referencing study, the owner approaching or retreating from the object changed the dogs' behavior far more than tone of voice or facial expression (Merola et al., 2012b). A handler who walks calmly toward something the dog finds uncertain, rather than talking about it, is using the signal that actually worked.

6.2 Your Reaction Is Being Watched

Most dogs looked to their owner when confronted with something uncertain (Merola et al., 2012b). A startle, a sudden tightening of the lead or a retreat is information to the dog in that moment. That is not a reason for pretending to be calm, which dogs may or may not detect, but for practising routines that make the handler's own reaction predictable.

6.3 Presence Helps in Difficult Moments

Veterinary visits, first encounters with something new, and situations that are unavoidably unpleasant are the obvious applications of buffering. The data from threat and novelty suggest that a familiar person's presence reduces the response (Tuber et al., 1996; Gácsi et al., 2013), which is an argument for owners staying with their dogs where that is possible and safe (how handling can be arranged).

6.4 Buffering During Extinction

Where a behavior is being deliberately no longer reinforced, the owner's presence reduced frustration in the one study that tested it (Dzik et al., 2025). That is a small but useful point for anyone running an extinction-based plan: it goes better with a calm familiar person nearby than with the dog left alone to work it out (how extinction works).

6.5 Build Independence Too

Because buffering is presence-dependent, a dog that meets every challenge only with the owner beside it may struggle when the owner is not there. The practical balance is to use presence for the hard moments and to build confidence in easier ones where the dog can succeed on its own.

6.6 Situations in Practice

Something new on the walk. A dog stops and looks from the object to the handler. The handler walks a few calm steps toward it, without urging, and waits. That uses the signal the referencing study found strongest, and it leaves the decision to approach with the dog.

The veterinary visit. Staying with the dog where the clinic allows it uses the buffering effect directly. Combined with predictable handling and short procedures, it addresses both the presence and the information the dog receives.

An extinction-based training plan. Running the plan with the owner present and calm, rather than leaving the dog alone to discover that a behavior no longer works, reflects the one study that tested buffering against frustration (Dzik et al., 2025).

The dog that won't move without its owner. Presence helps; it can also become the only condition under which the dog copes. Here the plan shifts toward building confidence in easy situations with the owner at gradually greater distance (how fear develops and changes).

6.7 What to Tell a Household

Four sentences cover it. Your dog watches you when something is uncertain, and what you do with your body tells it more than what you say. Walking calmly toward something helps more than reassuring words. Staying with your dog through difficult moments reduces its stress. And your presence is a support, not a cure, so build confidence in easier situations too.

6.8 What Not to Do

Forcing a dog toward something it is uncertain about removes the choice that makes referencing work: the dog is not using information to decide, it is being moved. Scolding a dog for hesitating adds an unpleasant event to an already uncertain one, from the very person the dog was looking to. Neither helps, and both work against the mechanisms described here (why the emotional response matters).

6.9 Adolescent Dogs

Adolescence is a period in which responses to the main carer can become temporarily less reliable. How that interacts with referencing and buffering has not been studied, but it is a reason to protect the relationship during that phase rather than to rely on it without thought (what changes in adolescence).

6.10 Arousal and Referencing

It is plausible that a highly aroused dog references less effectively, looking to its handler less and using what it sees less well. That has not been tested in the studies discussed here. If it holds, keeping arousal manageable in situations where the dog will meet something uncertain is therefore not only about calmness for its own sake, but about keeping the channel through which the handler's information reaches the dog open (how arousal is regulated).

6.11 Group Settings

In classes and group walks, dogs have several people to reference and several possible sources of support. The studies suggest that the owner remains the most influential of them (Merola et al., 2012a), which is a reason for trainers to coach owners' movement and positioning rather than to take over the dog in uncertain moments.

7. Summary at a Glance

Most dogs looked to their owner — 83 percent looked referentially to the owner when faced with an uncertain object (Merola et al., 2012b).

Movement mattered more than voice — Voice and facial expression alone had little effect; the owner approaching or retreating had a strong one (Merola et al., 2012b).

The owner's message counted more than a stranger's — Dogs regulated their behavior more with the owner as informant (Merola et al., 2012a).

A familiar human buffered novelty; a familiar dog did not — In kennel-housed dogs in a novel environment (Tuber et al., 1996).

The owner's presence dampened the heart-rate response to a threat — In thirty dogs approached by a threatening stranger (Gácsi et al., 2013).

The owner reduced frustration during extinction — In a reward-omission task (Dzik et al., 2025).

7.1 How the Findings Fit Together

Taken together, the studies describe a relationship in which the familiar person matters in two ways: as a reference point when something is uncertain and as a stabilizer when something is difficult. Neither role depends on special technique. Both depend on the dog having learned that this person is predictable and safe, which is the one thing a handler can build deliberately over time.

7.2 Why the Evidence Is Small but Useful

Seven studies is not a large literature, and several of the findings rest on a single experiment. What makes them useful is that their designs isolate specific variables: the informant's identity, the kind of message, the presence of a person versus a dog, the order of encounters. That precision is what allows practical conclusions to be drawn at all, even if each needs replication.

The honest way to use them is as a guide to which signals and conditions are worth attending to, not as proof that any particular handling routine works.

8. Research Gaps and Critical Appraisal

The referencing literature is small. The key studies come from one research group and use one ambiguous object (Merola et al., 2012a, 2012b).

Buffering studies use varied populations. Kennel-housed laboratory dogs, pet dogs in a laboratory and newly admitted shelter dogs are not interchangeable (Tuber et al., 1996; Gácsi et al., 2013; Shiverdecker et al., 2013).

Long-term effects are unknown. No study shows whether repeated buffered exposure changes a dog's later response when alone.

Individual differences are unexplored. Which dogs reference or are buffered most, and why, has not been systematically studied.

Everyday settings are untested. Laboratory encounters are brief and controlled; walks, clinics and homes are not.

8.1 What Would Move the Field

Four studies would add most. A replication of the referencing paradigm with different objects and in outdoor settings. A test of whether repeated buffered exposure changes a dog's later response when alone. A comparison of how referencing and buffering vary across individual dogs and dog–owner pairs. And studies in puppies and adolescent dogs, to see whether either mechanism is especially influential during development.

Everyday relevance is extrapolated. The practical recommendations in this article follow from laboratory findings and have not been tested as interventions in homes, clinics or training settings.

Mechanisms are unmeasured. No canine study discussed here measured oxytocin or other candidate mechanisms alongside referencing or buffering, so the physiological basis remains a hypothesis borrowed from other species.

9. Conclusion

People change how dogs respond to difficult situations in two distinguishable ways, and both have been demonstrated in dogs with more nuance than training lore suggests. Dogs look to people when confronted with something uncertain — most dogs did so with their owner (Merola et al., 2012b) and many with a stranger (Merola et al., 2012a) — but what they do with the information depends on the relationship and on the kind of signal: the owner's message influenced behavior more than a stranger's, and the owner's movement toward or away from an object had a far stronger effect than voice and expression alone. Separately, a familiar person's presence reduced physiological and behavioral responses to novelty where a familiar dog did not (Tuber et al., 1996), dampened the heart-rate response to a threatening stranger (Gácsi et al., 2013), and reduced frustration when an expected reward stopped coming (Dzik et al., 2025). The evidence is small, mostly laboratory-based and silent about long-term effects, and it gives no licence for claims that dogs simply absorb their owners' emotions. What it supports is modest and practical: move calmly toward what you want your dog to approach, recognize that your reactions are being watched, stay with your dog through difficult moments, and build independence in the easier ones.

Key Insights (Takeaways)

• Social referencing and social buffering are different mechanisms: one depends on what a person expresses, the other on the person being there.

• Most dogs looked to their owner when uncertain; the owner's movement influenced them far more than voice or expression (Merola et al., 2012b).

• Dogs regulated their behavior more strongly on the owner's message than on a stranger's (Merola et al., 2012a).

• A familiar person buffered the response to novelty where a familiar dog did not (Tuber et al., 1996).

• The owner's presence reduced the heart-rate response to a threat and reduced frustration during extinction (Gácsi et al., 2013; Dzik et al., 2025).

• Presence helps during a challenge; it is not by itself evidence that a fear is being resolved.

References

Dzik, M. V., Jousset, M., Vietti, J., & Bentosela, M. (2025). Effects of the presence of a human on frustration in domestic dogs: Is there a social buffering phenomenon? Applied Animal Behaviour Science, 283, 106503. https://doi.org/10.1016/j.applanim.2025.106503

Gácsi, M., Maros, K., Sernkvist, S., Faragó, T., & Miklósi, Á. (2013). Human analogue safe haven effect of the owner: Behavioural and heart rate response to stressful social stimuli in dogs. PLOS ONE, 8(3), e58475. https://doi.org/10.1371/journal.pone.0058475

Jakovcevic, A., Elgier, A. M., Mustaca, A. E., & Bentosela, M. (2013). Frustration behaviors in domestic dogs. Journal of Applied Animal Welfare Science, 16(1), 19–34. https://doi.org/10.1080/10888705.2013.740974

Merola, I., Prato-Previde, E., & Marshall-Pescini, S. (2012a). Dogs' social referencing towards owners and strangers. PLOS ONE, 7(10), e47653. https://doi.org/10.1371/journal.pone.0047653

Merola, I., Prato-Previde, E., & Marshall-Pescini, S. (2012b). Social referencing in dog-owner dyads? Animal Cognition, 15(2), 175–185. https://doi.org/10.1007/s10071-011-0443-0

Shiverdecker, M. D., Schiml, P. A., & Hennessy, M. B. (2013). Human interaction moderates plasma cortisol and behavioral responses of dogs to shelter housing. Physiology & Behavior, 109, 75–79. https://doi.org/10.1016/j.physbeh.2012.12.002

Tuber, D. S., Hennessy, M. B., Sanders, S., & Miller, J. A. (1996). Behavioral and glucocorticoid responses of adult domestic dogs (Canis familiaris) to companionship and social separation. Journal of Comparative Psychology, 110(1), 103–108. https://doi.org/10.1037/0735-7036.110.1.103