Researchers have built tests to measure empathy, and the tests have limits. Here are two of them. Empathy is sensing and sharing what another person is feeling. Cognitive empathy means understanding another's perspective or mental state, and it is a different thing from affective empathy. The Interpersonal Reactivity Index is a 28-item self-report test by Mark Davis that measures empathy. The Interpersonal Reactivity Index has four parts: perspective taking, fantasy, empathic concern and personal distress. The Empathy Quotient, made by Baron-Cohen and Wheelwright, has 60 items (40 about empathy, 20 controls) and is scored from 0 to 80. Empathy measures often ask people to report on their own ability, which is a limit on what they can show.

In one task, an adult hung clothes on a line and dropped a clothes peg that he could not reach. Toddlers who were 18 months old helped a grown-up they had never met. In the 2006 toddler study, virtually all of the children helped at least once, and the adult never asked them for help. Young chimpanzees helped a person who was reaching for something. Nobody asked or rewarded the toddlers and chimpanzees for helping. As of November 2026, in 2006 the researcher Warneken said this was the first experiment to show helping toward goals in a non-human primate. Helping behavior is seen in humans from about two years old, when a toddler can understand subtle emotional cues.

Rats were also tested for helping, in a different kind of experiment. A free rat will open the door of a trap to let out a rat from its own cage that is stuck inside. The researchers did not train the rats to free their cagemates. A rat was as likely to free its trapped cagemate first as to open a container of chocolate. A free rat acted more upset when its cagemate was trapped, as if it picked up the cagemate's distress. Rats freed trapped cagemates even when the reward of social contact was prevented, so the contact was not the only reason. Researchers think that helping a cagemate may itself be rewarding for a rat.

Wilkinson first reported in 1984 that vampire bats may share regurgitated blood with non-relatives. Vampire bats form strong bonds with their colony mates, and some unrelated bats have traveled together for more than ten years. A vampire bat can share some of its blood meal with another bat in its colony. A vampire bat that misses a few nights of feeding in a row can starve. A vampire bat that found no food can beg another bat for blood, and a small amount lets it live and hunt another day. Among vampire bats, grooming is mostly associated with food sharing, and in some pairs grooming eventually gave way to it.

Next, two kinds of animals that respond to others who are upset. Asian elephants comfort other elephants that are upset, using touch and sounds. An elephant may put its trunk in another elephant's mouth to calm it, a gesture a researcher compared to a handshake or hug. As of November 2026, a 2014 study gave the first scientific evidence that elephants console one another. In a study, pet dogs went up to a person who was crying. Dogs in a study went to a crying stranger, not only to their owner. The dog study used a method first used with human infants, and most dogs responded to the crying person in a submissive way.

In studies, affective empathy goes with more activity in the insula, and cognitive empathy with activity in the mid cingulate cortex. Mirror neurons were first found in macaque monkeys: some cells fired when a monkey grasped food and when it watched a person do the same. Scientists still argue about what mirror neurons do in people, including how much they explain reading other minds. Oxytocin is a peptide hormone and neuropeptide normally made in the hypothalamus and released by the posterior pituitary. Oxytocin has a role in social behaviors in many species, so it likely also does in humans, and it may play a part in bonding between humans and dogs.

Doing something kind for another person can make you feel happier. A psychologist's study found that doing kind acts raised people's well-being. Doing a variety of kind acts seems to work better than repeating the same one. In a study of children under two, toddlers seemed happier giving a treat away than getting one. In a bonus study, how workers spent the money, not how big the bonus was, changed their happiness. In one experiment, people were handed a $5 or $20 bill and told to spend it by 5 p.m. that day; those who spent it on others reported more happiness. Regular volunteering has been linked to a lower risk of dying and better physical functioning into old age.

In the public goods game, each player secretly chooses how many private tokens to put into a shared pot, which is used to study altruism and cooperation. In repeated public goods games, giving typically declines over rounds but rarely drops to zero. In the dictator game one person decides how much to give to another, who has no say, so it measures how much a person is willing to give. In a brain-scan study, activity was larger when people gave money voluntarily than when they paid it as taxes. Tit for tat copies what the other player did last, starts with cooperation, and won two competitions against other strategies.

These studies asked how far a kind act can travel. In one experiment, one person's generosity spread to three people and then to nine. In experiments, generous behavior spread through a group, and so did stingy behavior. In a cooperation experiment, a small number of generous people changed how other people in the group behaved. Christakis compared a social network to a matching grant: one person's giving in the first round was ultimately tripled by others. Cooperative behavior in the experiments spread up to three degrees of separation. A science writer noted that Fowler and Christakis report that both cooperative and selfish acts spread, so it is unclear how the ripples affect real networks.

Words to know

• empathy: sharing feelings
• insula: brain region
• oxytocin: a hormone
• altruism: selfless concern
• submissive: yielding