Animal behaviour

The actions and responses of animals, studied through their mechanisms, development, functions and evolutionary history.

Collective vigilance is the surveillance supplied by a group of animals, for example when attention by any member can reveal an approaching predator. Group coverage and each individual's time spent vigilant are different quantities. Several members can jointly provide substantial coverage even if each watches for only part of the time.

An original independent-attention model makes that distinction precise. Suppose each of $n$ individuals is vigilant at a randomly chosen instant with probability $p$, independently of the others. The probability that at least one is vigilant isFor $p=0.2$, one individual gives0.2 coverage, while five give $1-0.8^5=0.67232$. The model specifies attention, not guaranteed detection: visibility, sensory ability and information transfer still matter.

Independence is a substantive assumption. If every individual watches during exactly the same20percent of instants, coverage remains0.2 however large the group. In an ideal five-member schedule with perfectly nonoverlapping20-percent attention intervals, coverage can instead be one. Thus group size and individual attention fractions alone do not determine the collective result.

A field study of Dolichotis patagonum found that larger communal warrens had more adult attendance and greater total vigilance despite less sitting-alert time per pair. The reported portion of the day with at least one vigilant pair reached90percent at larger crèches. This supplied an observational example of individual and group measures changing in different directions.^[ref]

The field comparison does not randomize warren membership or establish that the independent-attention model describes those animals. Simultaneous behavioral observations and records of predation risk can test which temporal correlations and detection mechanisms actually produce group coverage.

A playback experiment presents a recorded signal to an animal and measures its response under a specified procedure. In animal behaviour, controlled playback can test the effect of an acoustic cue while avoiding the uncontrolled behavior of a live signaler. The recording, loudspeaker position, playback level, timing and response criterion are parts of the experimental manipulation.

A field example compared recordings of one female lion roaring with three females roaring together. The recordings simulated unfamiliar intruders in territories of 21 lion prides. Defending females were less likely to approach the three-intruder condition and approached it more cautiously; the defenders' own group composition also mattered.^[ref]

For an original design exercise, imagine testing eight animal groups with one recording, then testing eight groups with eight independently obtained recordings. Both designs observe eight groups, but only the second samples variation among those recordings. Replaying a single recording many times does not make its acoustic details independent examples of a whole signal category. This is one route to pseudoreplication.

A playback experiment also requires checking that a response difference tracks the intended cue rather than an unintended loudness or position difference. The observed approach or call is a behavioral result; interpreting it as a particular cognitive process requires an argument connecting the manipulation, controls and alternative explanations.

A camera trap is a camera arranged to record animals at a location without a person continuously operating it there. The trigger and sampling procedure determine what becomes a recorded encounter. Images can reveal occurrence, activity or recognizable individuals, but a photograph is not automatically an independent animal count.

In a 1995 tiger study, cameras along travel routes produced 31 photographic captures of ten identified animals over nine sampling occasions. Researchers analyzed the resulting encounter histories with mark and recapture models, including variation in detection probability among individuals.^[ref]

An original toy example has three images of animal A, two of B and one of C. There are six photographs, three detected animals and an unknown number of undetected animals. Counting images answers the first question. Recognizing individual markings answers the second. Estimating total abundance additionally requires assumptions about the observation process.

Sampling effort matters too. Two cameras operated for three nights provide six camera-nights of nominal effort, but that equality alone does not make them equivalent to six cameras operated for one night: location and temporal coverage differ. State how encounters are defined and how repeated images are handled before interpreting a photographic detection rate as a biological difference.

In animal behaviour, a social group is a set of animals whose association and interactions are studied as a unit. Its operational definition specifies membership, the observation period and what counts as association. Sharing a location once need not establish persistent affiliation, and a group assembled for an experiment need not represent every naturally occurring grouping.

Group structure includes more than social group size. With $n$ members, there are $n(n-1)/2$ possible unordered pairs. A group of four therefore permits six distinct pairs, whereas a group of ten permits 45. These are possible relationships, not observations that every pair actually interacts. Recording which animals associate preserves information that a headcount discards.

Group membership also affects experimental interpretation. A study of neophobia compared 70 Bengalese finches in seven groups with 58 white-rumped munias in six groups. Feeding responses were recorded under familiar and novel-object conditions.^[ref] The counts of birds and groups describe different levels of replication. If animals share a setting and influence one another, treating every bird as an independent replicate of a group-level treatment risks pseudoreplication. The appropriate experimental unit follows the treatment assignment and measurement design, rather than simply the largest available count.

Social group size is the number of individuals assigned to a social group under a stated membership rule and observation period. A headcount at one moment, a count of regular members and the number present during an experiment can differ. Report which quantity was measured before comparing groups.

The way groups are sampled changes their apparent typical size. Consider two groups containing 2 and 8 animals. Their arithmetic mean size is $(2+8)/2=5$. If an animal is instead sampled uniformly from the ten individuals, the expected size of its group isMore animals experience the larger group. For group sizes $n_1,\dots ,n_k$, this individual-weighted mean group size is ${\sum }_i{n}_i^2/{\sum }_i{n}_i$, rather than ${\sum }_i{n}_i/k$. These are original illustrative calculations, not measured animal data.

Social group size also affects the interpretation of replication. A finch experiment used 70 Bengalese finches in seven groups and 58 white-rumped munias in six groups. The groups contained ten birds each except one eight-bird munia group. Those totals describe birds and groups, not interchangeable counts of independent treatment assignments.^[ref] The experimental unit must follow the design; a larger headcount alone does not solve pseudoreplication.

The individual-weighted mean group size is the expected size of the social group containing an individual chosen uniformly from all individuals in the groups being studied. Larger groups receive more weight because they contain more potential sampled individuals. For positive group sizes $n_1,\dots ,n_k$, the probability of selecting someone in group $i$ is $n_i/{\sum }_j{n}_j$, giving the weighted meanThe focal individual is included; its expected number of group companions is $G-1$.

For an original example, groups of sizes $1,1,4$ contain six individuals. Their group-based arithmetic mean is 2, but $G=(1+1+16)/6=3$. Four of the six individuals experience a group of four, so an individual-based observation gives the larger group more influence.

Write $\mu =k^{-1}{\sum }_i{n}_i$ and ${\sigma }^2=k^{-1}{\sum }_i(n_i-\mu {)}^2$, using a divisor of $k$ for this finite collection. Expanding the squares gives ${\sum }_i{n}_i^2=k({\sigma }^2+{\mu }^2)$ and thereforeEquality holds precisely when all groups have equal size. The difference is determined by group-size variance, rather than a contradiction between two legitimate averages.

An experiment reporting numbers of animals and numbers of groups supplies both levels of description. For instance, a finch study used 70 Bengalese finches in seven groups and 58 white-rumped munias in six groups.^[ref] Choosing the appropriate average depends on whether the scientific question concerns typical groups or the social setting experienced by typical individuals.

Neophobia is avoidance of, or inhibited response to, unfamiliar stimuli. In animal behaviour, it is investigated through specified responses such as approach or feeding in a novel-object test. An unfamiliar object near food and an unfamiliar food itself are different stimulus manipulations, so the task needs to be stated.

One experiment compared domesticated Bengalese finches with white-rumped munias, their wild ancestor. All groups fed in control conditions; with novel objects present, six of seven Bengalese-finch groups and none of six munia groups fed during the observation period.^[ref] The observation measures behavioural disruption by novelty. It does not directly read an animal's subjective experience or isolate a unique inherited cause of the difference.

A constructed feeding latency example shows why matched conditions help. One animal feeds after $30\mathrm{s}$ in familiar conditions and $90\mathrm{s}$ with a novel object, a $60\mathrm{s}$ increase. Another changes from $120$ to $150\mathrm{s}$, a $30\mathrm{s}$ increase. The second has the longer raw delay but the smaller novelty-associated change.

Observation duration matters too. If an animal has not fed when recording ends, its latency is subject to censoring; it is not a measured infinite delay. Prior exposure, hunger and social group context can also affect the response. These details make the behavioural measurement interpretable and reproducible.

A novel-object test measures an animal's response to an unfamiliar object under a specified procedure. In studies of neophobia, the object can be placed near familiar food, and the observer records approach, whether feeding occurs, or feeding latency. Keeping the food familiar helps distinguish response to the object from response to an unfamiliar food itself.

One controlled comparison used Bengalese finches and white-rumped munias. All groups fed in control conditions. With novel objects present, six of seven finch groups and none of six munia groups fed during the observation period.^[ref] This measures a difference in the task response; it does not uniquely identify its inherited or developmental cause.

For an original illustrative design, compare a bird's first feeding time with familiar food alone against its first feeding time when an unfamiliar object is added nearby. Define the start of each trial and the maximum observation time beforehand. Recording both approach and actual feeding can distinguish a bird that investigates without eating from one that never approaches.

Prior exposure can make an object familiar through habituation, and a companion's behavior can affect the result. Thus object identity, trial sequence and social group size belong in the procedure, alongside the numerical outcome. A novel-object test is an operational measurement, rather than a direct reading of subjective fear.

Feeding latency is the elapsed time from a defined starting event to a defined feeding event. An experiment might measure from food presentation to an individual's first peck or to the first bird feeding in a social group. Those endpoints answer different questions. The start, the feeding criterion and the observation limit must all be specified.

In an original example, an animal feeds after $20\mathrm{s}$ without a novel object and $80\mathrm{s}$ with one nearby: the novelty-associated increase is $60\mathrm{s}$. Another changes from $100$ to $130\mathrm{s}$, an increase of $30\mathrm{s}$. Comparing only the two novel-condition times reverses the ordering obtained by comparing their changes.

A group-level first-feeding time is $T_{\mathrm{group}}=\min (T_1,\dots ,T_n)$, if $T_i$ denotes each member's time under that same group trial. This identity requires no independence assumption. It also shows why a group endpoint cannot be treated as though it were a measurement of every member: one early feeder determines the recorded minimum.

If recording ends after $180\mathrm{s}$ before feeding occurs, the observation is subject to right censoring: the time exceeds the recorded limit. Replacing it by an observed time of exactly $180\mathrm{s}$, or by infinity, changes what was measured.

A finch study combined whether groups fed at all with feeding latency comparisons in familiar and novel-object conditions. Keeping those outcomes separate preserves information that an average of observed feeding times alone could lose.^[ref]

Cooperative breeding is a reproductive system in which additional group members help care for offspring beyond the care supplied by the breeding parents alone. Additional carers may be nonbreeding helpers or individuals that also share reproduction. Feeding young, defending a territory and attending a nest are different possible contributions. Group membership alone does not establish that an individual provides care.^[ref]

An observational trap is that good territories may produce more surviving offspring and therefore more potential helpers. A positive association between helper number and later reproductive success need not measure the helpers' causal effect. In a Seychelles warbler study, removing additional carers and accounting for territory quality, breeder age and group composition gave a more direct comparison. One additional carer improved success in the tested circumstances; more additional adults on medium-quality territories could instead reduce it.^[ref]

An original feeding example shows another subtlety. Two parents each make ten visits per hour, for twenty total. A helper contributes eight visits, while each parent reduces its contribution to six: the total remains twenty. Unchanged total provisioning would not mean the helper contributed nothing; the parents' workload changed.

The benefit to a group, the cost to a particular carer and the evolutionary explanation for helping are separate questions. A useful experiment states which outcome it measures, rather than assuming that more adults always produce more offspring or that every helper has the same relationship to the brood.

Birdsong is patterned vocal behaviour used by birds in contexts such as territorial defence and mate attraction. Song production, auditory recognition and acquisition of a local song pattern are different processes. In many songbirds, hearing other individuals during development contributes to the song later produced.

In controlled chaffinch studies, hand-reared birds exposed to different auditory examples developed different song patterns. Artificial tutor sequences assembled from suitable notes could also be incorporated. This made auditory experience experimentally separable from the species-typical constraints on what could be learned. ^[ref] The comparison does not imply that every bird learns its vocalizations in the same way.

Changing the current value of an outcome after instrumental training, then testing the associated action. Testing without delivering that outcome helps distinguish immediate use of changed outcome value from new direct action reinforcement. Sensitivity and insensitivity provide operational tests for goal-directed and habitual control under stated conditions.^[ref]

A speeded test requiring responses associated with still-valued outcomes and withholding responses associated with devalued outcomes. Response errors can reveal deficient use of outcome knowledge, but also require controls for memory retention, comprehension and inhibitory control.^[ref]

Consumption of remains of dead organisms. It can remove, disarticulate and transport material before burial.

Movement of an organism through its environment, including walking, swimming and saltation.

locomotion by jumps. Skeletal geometry can constrain jumping capability but does not establish the exclusive behaviour of an extinct organism.

Behaviour whose expression depends on properties such as condition, age, reproductive state or current social position.

A mapping from perceived social information and internal state to action. A stable mapping can generate flexible actions when its inputs change.

A regulated behavioural and physiological state identified through converging criteria including reversibility, responsiveness and homeostatic regulation. Immobility alone does not establish sleep.

Slow-wave sleep is a deep stage of non-rapid eye movement sleep characterized by prominent slow, high-amplitude activity in an electroencephalography recording. It is a physiological classification based on measured brain activity and other sleep criteria, rather than simply an unusually motionless posture. In current human staging it broadly corresponds to N3. ^[ref]

Sleep-stage identification can guide interventions. In a human experiment, an odour associated with earlier learning was presented again during slow-wave sleep; subsequent declarative-memory performance was compared with control conditions. ^[ref] This links a defined physiological state to an experimentally timed cue; it does not establish that every kind of memory depends exclusively on that one stage.

The proposal that waking-associated strengthening is balanced by sleep-associated renormalization of overall strength of synapses. Selective weakening could improve capacity and relative signal contrast while preserving important differences. It is compatible with some strengthening or reactivation, and is not the claim that sleep indiscriminately erases learning.

Combined recording used to classify sleep and waking, commonly including electroencephalography, eye-movement signals and muscle activity. Classification of a sleep stage is not itself a measure of the content of a memory.

A sleep state identified by rapid eye movements and characteristic electrophysiological and muscle-tone patterns. Its contribution to a particular learned task must be tested rather than inferred from the existence of dreams.

A group of sleep stages distinct from rapid eye movement sleep. Deep stages feature prominent sleep slow oscillations; stage assignment uses polysomnography.

A transient thalamocortical oscillatory event, often around 10–16 Hz in humans, with frequency boundaries depending on the measurement convention. Its timing relative to other events can matter beyond its total count.

A slow alternation in cortical population activity, conventionally below about one cycle per second. Its phase can organize other sleep events; a scalp electrical rhythm alone does not reveal a particular represented episode.

Searching for, acquiring and handling food. Tests distinguish resource-specific actions from general movement or appetite.

Grooming is behavior directed at cleaning or maintaining the body surface, either of oneself or another individual. It can remove material and parasites while also functioning in social interactions. The physical action and its social consequences should be measured separately.

Coordinated changes in activity, appetite and social or other motivations associated with immune activation. Context dependence and experimental controls distinguish it from simple incapacity.

Use of substances or actions that reduce harm from parasites or disease. Tests require a health benefit, a relevant behavioural response and controls against ordinary nutritional preference.

Parental use or selection of substances that reduce disease in offspring. A benefit to offspring does not imply treatment or cure of the parent.

Sensory and behavioural processes reducing exposure to pathogens, such as avoidance of contamination. This broader defence usage should be distinguished from cellular and molecular immunity.

Removal or avoidance of contaminated material or infected individuals, potentially reducing exposure of an individual or group to pathogens. disease reduction must be measured rather than inferred from cleaning alone.

Transmission of information between animals through signals or cues, assessed through effects on receivers.

Discrimination among potential social or mating partners according to species-associated information. Recognition errors and discrimination costs affect its evolution.

Use of a signal by an unintended receiver, such as a predator detecting a mating call. This can impose selection opposed to intended-receiver attraction.

An informative feature not necessarily evolved for the receiver’s response. Distinguishing cues from signals requires an evolutionary argument, not just evidence of information transfer.

A feature or action shaped by selection for its effects on a receiver, with receiver responses also subject to selection. A merely informative incidental feature can instead be a biological cue.

A relation between a signal and a property relevant to its receiver. Honesty can follow aligned interests, physical constraints or differential costs and benefits; high production cost alone does not establish it.

Signalling in which costs contribute to the evolutionary incentives governing expression. Reliability depends on differences in net payoffs among signaller types, not merely on a signal being expensive.

A signal whose relation to the advertised property is constrained by its production mechanism. The receiver’s use and the relevant physical constraint should both be tested.

Communication using more than one sensory modality. Components can supply redundant information, different information or context-dependent detection benefits.

The sensory channel through which a signal acts, such as acoustic, visual, chemical, mechanical or electrical stimulation.

Use of a warning signal associated with defence. A conspicuous colour pattern alone does not establish either effective defence or predator avoidance.

Behaviour by a parent affecting offspring survival or development, including provisioning, defence and control of the developmental environment. Its costs and benefits need separate measurement.

Care supplied by both parents. It can create shared offspring benefits alongside conflict over which parent bears the costs.

Conditional adjustment of care in response to a partner, offspring or the environment. A matching response can reflect information about brood demand rather than absence of conflict.

An increase in care following a partner’s reduction that replaces less than the entire lost contribution. The predicted response depends on information, costs and the time available to adjust.

An allocation to current offspring that carries an opportunity cost for other reproductive opportunities. Time spent near offspring is not automatically an investment of the same value in both sexes.

Reproductive allocation is the distribution of resources and effort among reproductive activities, offspring and future opportunities. Investing more in one component can reduce resources for another. An allocation trade-off concerns limited resources rather than an assumption that every observed correlation is causal.

Information concerning the probability that offspring are genetically sired by a male. Its effect on care depends on detection, future mating opportunities and effects on offspring, not only average extra-pair paternity.

Biological material passed from one individual's body to another and capable of influencing the receiver. Simply carrying external food is not sufficient for the narrower framework; donor-processed or donor-produced components matter.

Maternal behaviour supplying offspring with resources and potentially bioactive transferred material. Resource quantity and material composition can affect offspring differently.

A provisioned dung mass surrounding the developmental site of an individual dung-beetle offspring. Its food quantity and maternal contributions can be manipulated separately.

Maternal faecal material associated with the egg chamber in certain dung beetles. Ingested material can provide an indirect route of maternal microbial transfer.

Use of mirror information in recognizing one's own body. A particular test outcome does not establish all forms of self-representation or theory of mind.

An assay asking whether an animal uses reflected information to investigate an otherwise unseen mark on its own body. Visible and sham marks, absence of the mirror and independent detection or action controls constrain the interpretation.

Monitoring or control of one's own cognition. Metacognitive monitoring estimates aspects of current or future performance, while metacognitive control uses such estimates to guide behaviour. Metamemory applies these functions to remembering; uncertainty responses provide a way to study some of them without verbal reports. Successful behaviour alone need not establish introspection or subjective awareness.

An individual's expressed certainty in a judgment. Its relation to objective correctness is measured separately from the certainty itself. Metacognitive calibration concerns numerical agreement with outcomes; metacognitive sensitivity concerns discrimination between better and worse performance.

A computation that evaluates an action using partially shared evidence about the action-producing system. First-order decisions and confidence need not have identical internal samples; calling the computation second order does not imply conscious verbal reasoning.^[ref]

Confidence inferred from agreement among sampled representations supporting a response. Internally consistent evidence can systematically favour a wrong answer, separating confidence from correctness.^[ref]

The ability of confidence to distinguish a person's correct from incorrect responses, or to rank items by later success. It is distinct from metacognitive calibration. A type 2 receiver operating characteristic measures discrimination in a binary-correctness task; other metamemory designs may need other measures.

The first-order sensitivity index that, in a fitted signal detection theory model, would generate the observed confidence discrimination. Model assumptions and sampling uncertainty matter; raw confidence–accuracy association alone does not estimate it.^[ref]

A model-based comparison of meta-d-prime with first-order discrimination, often their ratio. Interpretation becomes unstable when the denominator is close to zero. Efficiency, metacognitive calibration and overall confidence are different quantities.

If confidence $C$ is replaced by a strictly increasing function $g(C)$, the attainable threshold partitions have the same ordering and ties. Consequently the type 2 receiver operating characteristic and its area under the ROC curve are unchanged. Equivalently, $P(C_{\mathrm{correct}}>C_{\mathrm{error}})+\frac{1}{2}P(C_{\mathrm{correct}}=C_{\mathrm{error}})$ is invariant. Calibration need not be preserved: adding a positive constant to probabilities within their allowed range changes their absolute values while preserving ranks. A merely nondecreasing transformation can merge distinct reports and requires additional care.

A receiver operating characteristic treating high confidence as a prediction that a first-order answer was correct. Varying confidence threshold $t$ traces the false-positive and true-positive probabilities in the displayed expression. Raw discrimination can depend on first-order task performance; it is not automatically a performance-independent measure of monitoring quality.

For correctness indicator function $Y$ and stated probability $C$, ideal calibration means that answers assigned confidence $c$ are correct with conditional probability $c$. Overall overconfidence is $\mathbb{E}[C-Y]>0$; a zero overall difference does not establish calibration at every confidence level. Estimated calibration depends on sample size and how reports are grouped.

Confidence evaluated after making a response. It may use evidence, further processing and information about the action that were not all available at the moment of the response.

A normative Bayesian inference model of retrospective confidence. For two exhaustive alternatives with equal prior probabilities, let $L=\log [p(E\mid H_1)/p(E\mid H_0)]$ be the log-likelihood ratio. Bayes' theorem gives posterior odds $e^L$. Choosing the alternative with the larger posterior yields $P(\text{correct}\mid E)=(1+e^{-|L|}{)}^{-1}$. Unequal priors add prior log odds; choosing a different action changes which posterior probability is appropriate. This calculation does not establish that an observer knows the true likelihoods.

Metacognitive monitoring and metacognitive control applied to memory encoding, memory retention and memory retrieval. Different judgments predict different outcomes; a prediction of future recall need not agree with confidence in an answer currently retrieved.

A theory in which metamemory judgments infer likely performance from available cues, rather than necessarily inspecting a memory trace directly. Cue use and cue validity need not coincide.^[ref]

An aspect of the learner's current processing experience, such as retrieval ease, used in judging memory.

A property of the learning conditions, such as study time or repetitions, used in judging memory.

A property of the material, such as the difficulty or relatedness of paired words, used in judging memory.

An estimate that currently unrecalled information could be recognized or recovered later. It differs from confidence in an answer already produced and from a prediction made while initially studying it.

A prediction, made during or after study, of whether material will be remembered on a later test. Specify the delay, the test and whether the target remains visible; these determine what information the learner can use.

Improved itemwise predictive accuracy of judgments of learning made after a delay under suitable conditions. The available cue matters: delaying a cue-alone judgment can be useful even when delaying a cue–target judgment gives little benefit.^[ref]

Regulation of cognition using information about one's own processing: allocating further study, seeking information, answering or declining a test. A change in a monitoring report does not by itself demonstrate a beneficial change in control.

An action permitting a subject to decline a difficult test, seek information or select an explicit uncertainty option. Reward, delay and task contingencies influence the action; an uncertainty response is an operational behaviour rather than a direct observation of subjective awareness.

Under linear expected utility, let a correct answer yield $r$, an error yield $-w$, and declining yield $s$, with $r+w>0$. Answering has expected payoff $pr-(1-p)w$, so it is preferred exactly when $p>(s+w)/(r+w)$. Equality gives indifference. Different utility, time costs or beliefs change this criterion; behaviour across differently rewarded tasks cannot be compared as if the threshold were fixed.

Estimating aspects of one's own cognition, such as whether an answer is correct or whether material can later be recalled. Metacognitive monitoring should be distinguished from metacognitive control, which changes behaviour in response to such estimates.

Awareness of one's current mental state, such as noticing that attention has drifted. Off-task thought and awareness of that thought can be measured separately.^[ref]

Storage of food for future use, which creates opportunities for both memory and competition over stored resources.

Looking towards a predicted future action or event before it occurs. Gaze measures require prespecified regions, timing and exclusions.

An internal state or process used in representing information about the world, self or others. Behavioural tests should distinguish its proposed content from simpler predictive rules.

Perception-like experience without a corresponding current sensory input. Its emotional effects, voluntary control and relation to autobiographical memory are separable research questions.

Representation of aspects of one's own body, knowledge or actions. These components need not entail an autobiographical or linguistically expressed concept of self.

Awareness involving oneself as an object or subject of representation. Its proposed forms must be operationally distinguished; bodily self-recognition does not alone establish reflective or autobiographical awareness.

An action-related representation of bodily configuration and capacities. Its functional use need not entail a verbal or autobiographical concept of self.

Identification of one’s own body or body-related sensory information. It is narrower than reflective knowledge of one’s identity and does not alone demonstrate theory of mind.

Attribution or representation of mental states in oneself or others to explain or predict behaviour. Component abilities need not form an all-or-none package.

A task framework using video narratives to manipulate emotionality and mental-state reasoning separately. Affect and compassion ratings, reasoning accuracy and confidence can be analysed as distinct outcomes.

Interpreting another agent's action in relation to an outcome it is pursuing. It can be narrower than representing the agent's beliefs or intentions in a human-like way.

Representing an agent's belief as differing from reality and predicting action accordingly. False belief, ignorance and a correct remembered location require different controls.

An emergence account assigns a capacity such as language or executive control a role in constructing understanding of belief over development. An expression account assigns it a role in accessing and displaying already available understanding. The same observed correlation is compatible with both. Immediate task-demand manipulations primarily probe expression; lasting transfer from a controlled intervention can constrain emergence, provided strategy and practice explanations are addressed.

A task testing how an agent will act or respond when its representation differs from the actual state of affairs. Unexpected-transfer tasks contrast the location an agent witnessed with a subsequent displacement it did not witness; comprehension controls and response demands affect interpretation.^[ref]

Language interpretation, memory, inhibitory control, response generation and other demands affecting the observed answer. Lowering concurrent demands while preserving the mistaken-belief contrast can reveal successful performance earlier than a standard task. The modified procedure still needs controls against cueing and simplified nonmental strategies.^[ref]

Tracking another individual's informational access or knowledge. Success does not automatically imply understanding false belief.

Using differences in visual access between individuals. Knowing what another can see differs from representing exactly how an object appears to them.

A change in responses to ambiguous stimuli relative to trained reference stimuli. Motivation, general activity and perceptual discrimination require control.