Category: Consumer Behavior

— An artificially intelligent virtual gamer created by computer scientists at The University of Texas at Austin has won the BotPrize by convincing a panel of judges that it was more human-like than half the humans it competed against.

 

The competition was sponsored by 2K Games and was set inside the virtual world of "Unreal Tournament 2004," a first-person shooter video game. The winners were announced this month at the IEEE Conference on Computational Intelligence and Games.

"The idea is to evaluate how we can make game bots, which are nonplayer characters (NPCs) controlled by AI algorithms, appear as human as possible," said Risto Miikkulainen, professor of computer science in the College of Natural Sciences. Miikkulainen created the bot, called the UT^2 game bot, with doctoral students Jacob Schrum and Igor Karpov.

The bots face off in a tournament against one another and about an equal number of humans, with each player trying to score points by eliminating its opponents. Each player also has a "judging gun" in addition to its usual complement of weapons. That gun is used to tag opponents as human or bot.

The bot that is scored as most human-like by the human judges is named the winner. UT^2, which won a warm-up competition last month, shared the honors with MirrorBot, which was programmed by Romanian computer scientist Mihai Polceanu.

The winning bots both achieved a humanness rating of 52 percent. Human players received an average humanness rating of only 40 percent. The two winning teams will split the $7,000 first prize.

The victory comes 100 years after the birth of mathematician and computer scientist Alan Turing, whose "Turing test" stands as one of the foundational definitions of what constitutes true machine intelligence. Turing argued that we will never be able to see inside a machine's hypothetical consciousness, so the best measure of machine sentience is whether it can fool us into believing it is human.

"When this 'Turing test for game bots' competition was started, the goal was 50 percent humanness," said Miikkulainen. "It took us five years to get there, but that level was finally reached last week, and it's not a fluke."

The complex gameplay and 3-D environments of "Unreal Tournament 2004" require that bots mimic humans in a number of ways, including moving around in 3-D space, engaging in chaotic combat against multiple opponents and reasoning about the best strategy at any given point in the game. Even displays of distinctively human irrational behavior can, in some cases, be emulated.

"People tend to tenaciously pursue specific opponents without regard for optimality," said Schrum. "When humans have a grudge, they'll chase after an enemy even when it's not in their interests. We can mimic that behavior."

In order to most convincingly mimic as much of the range of human behavior as possible, the team takes a two-pronged approach. Some behavior is modeled directly on previously observed human behavior, while the central battle behaviors are developed through a process called neuroevolution, which runs artificially intelligent neural networks through a survival-of-the-fittest gauntlet that is modeled on the biological process of evolution.

Networks that thrive in a given environment are kept, and the less fit are thrown away. The holes in the population are filled by copies of the fit ones and by their "offspring," which are created by randomly modifying (mutating) the survivors. The simulation is run for as many generations as are necessary for networks to emerge that have evolved the desired behavior.

"In the case of the BotPrize," said Schrum, "a great deal of the challenge is in defining what 'human-like' is, and then setting constraints upon the neural networks so that they evolve toward that behavior.

"If we just set the goal as eliminating one's enemies, a bot will evolve toward having perfect aim, which is not very human-like. So we impose constraints on the bot's aim, such that rapid movements and long distances decrease accuracy. By evolving for good performance under such behavioral constraints, the bot's skill is optimized within human limitations, resulting in behavior that is good but still human-like."

Miikkulainen said that methods developed for the BotPrize competition should eventually be useful not just in developing games that are more entertaining, but also in creating virtual training environments that are more realistic, and even in building robots that interact with humans in more pleasant and effective ways.

"This persistent, growing national problem is frustrating for consumers and can lead to loss of business and, by extension, loss of tax revenue for the community," Kellaris said.

Kellaris, the James S. Womack/Gemini Chair of Signage and Visual Marketing in the UC Carl H. Lindner College of Business, will present this research during the October 10 -11 Fourth-Annual National Signage Research & Education Conference (NSREC) in Cincinnati.

Through a UC analysis of a market research survey of North American households, Kellaris found that inadequate signage could be construed as a communication failure.

"About half the population surveyed in 2011 has driven by and failed to find a business due to signage and communication failure," he said.

While communication failure affects all groups and ages, the study found that women experience signage communication failure more than men.

Shoppers, Kellaris noted, favor signs that are visible, legible and informative, but those preferences contradict current trends of smaller, more uniform signs, using non-verbal symbols/icons.

Kellaris said that shoppers are drawn to unfamiliar stores based on clear, attractive signs, and that often, these stores convey personality and character as perceived from signage quality.

The resolution, Kellaris says, "is to find the right balance between the interests of shoppers, businesses and the broader interests of the community."

The National Signage Research & Education Conference (NSREC) is a two-day event at the Kingsgate Marriott Conference Center focusing on the theme: "The Technology of Signage." Other UC presenters include Jeff Rexhausen, research associate in the Economics Center for Education & Research, and George Vredeveld, professor of economics. They will speak on the topic of "The Economic Value of Signs Research Study." The conference keynote address will be delivered by UC's Craig Vogel, an internationally known designer, educator and author who currently holds an interim appointment as the Terry Fruth/Gemini Chair of Signage Design and Community Planning.

 The price a consumer will pay for a product is often significantly less than the price they will accept to sell it. According to a new study in the Journal of Consumer Research, this occurs because ownership of a product enhances its value by creating an association between the product and consumer identity.

"Our studies support the idea that ownership enhances the attractiveness of a product because ownership creates an association between the item and the self," write authors Sara Loughran Dommer (Georgia Institute of Technology) and Vanitha Swaminathan (University of Pittsburgh).

In several studies, the authors found a link between possessions and consumer identity. They also discovered that men are more likely to consider a product's association with specific social groups when making a purchase. "Men strive to differentiate themselves and group distinctions are more significant for them. In contrast, women are focused on forming connections and less likely to classify themselves as separate from others. They are less likely to purchase products because of an association with a particular social group," the authors write.

Businesses can benefit from creating feelings of ownership through promotional strategies such as free trials, samples, and coupons. For instance, a consumer may be more willing to purchase a couch if they are offered a free trial, clothing stores increase sales by having customers try on items, and sporting goods stores could allow consumers to try out equipment in the store to boost sales.

However, companies wanting women to identify with and purchase their brands need to work harder to emphasize the identity differences across brands. A good example would be Apple's recent Mac versus PC advertising campaign that depicted the distinct identities of the two brands.

"If ownership increases the value consumers place on products, then companies could benefit from any action that creates feelings of ownership before actual purchase. Our findings regarding gender differences also suggest that in certain situations companies may benefit from prompting female consumers to make intergroup comparisons," the authors conclude.

Consumers enjoy products more in the long run if they don't overuse them when first purchased, according to a new study in the Journal of Consumer Research.

"Consumers are naturally prone to consume products they enjoy too rapidly for their own good, growing tired of them more quickly than they would if they slowed down," write authors Jeff Galak (Carnegie Mellon University), Justin Kruger (New York University), and George Loewenstein (Carnegie Mellon University).

We often face decisions about how rapidly to consume products we enjoy: how quickly to eat a favorite dessert; how often to listen to a favorite song; or how frequently to play a new video game. But do we make choices that maximize our enjoyment of such products? It turns out that there may a selfish reason to resist the temptation to overindulge.

The authors asked consumers to eat a well-liked food such as chocolate or play an exciting video game either at their own pace or at longer intervals. When consumers were given the ability to choose a rate of consumption and that decision was constrained to force them to consume slowly, they enjoyed the overall experience more than those who either chose their rate of consumption in an unconstrained manner or those whose rate of consumption was chosen for them.

Because consumers choose to consume too quickly, they don't appreciate that spacing out consumption decreases satiation and thereby increases enjoyment. Paradoxically, we tend to make choices that will bring us less pleasure overall.

"When you are lucky enough to be able to choose how often to consume the things you enjoy, space out your consumption. Not only will the experience last longer, but it will be more enjoyable as well," the authors conclude.

 Confident consumers pay more attention to advertisements and product information that focus on high-level features of a product, according to a new study in the Journal of Consumer Research. Less confident consumers, however, focus on the basics.

"When we feel confident, we think that abstract information is more relevant to us. But when we feel doubtful, we think that concrete information is more relevant. The more relevant we perceive information to be, the more we will focus on it," write authors Echo Wen Wan (University of Hong Kong) and Derek D. Rucker (Kellogg School of Management — Northwestern University).

The authors conducted a series of experiments to examine how psychological confidence affects consumers' attention and scrutiny of product information.

In one experiment, the authors induced participants to feel either confident or doubtful and then asked them to describe the action of "locking the door." Confident people tended to describe it in terms of its high-level meaning of "securing the house," whereas doubtful people tended to describe it in terms of its concrete action of "putting a key in the lock."

In another experiment, confident consumers paid more attention to an ad for a health club when the message was focused on the idea of long-term health, a high-level and abstract benefit. They paid less attention to the ad when the message was focused on the idea that they could enjoy daily workouts, a low- level and concrete benefit. When consumers felt doubtful, the opposite occurred and they paid more attention to an ad when it discussed concrete as opposed to abstract benefits.

"Feeling highly confident prompts consumers to consider things from a global perspective and focus on the high-level and essential aspects of products, whereas feeling less confident or uncertain makes people focus on low-level and contextual details," the authors conclude.

Growing communities can overcome conflict and fragmentation, and increase diversity, without losing their sense of collective belonging, according to a new study in the Journal of Consumer Research.

"Consumption communities are groups of people united by a shared passion. A persistent challenge to community is continued engagement, and collective enterprises can be destabilized by differences as they grow. Our research shows how community members leverage social and economic resources to overcome differences," write authors Tandy Chalmers Thomas (Queen's University), Linda L. Price, and Hope Jensen Schau (both University of Arizona).

The authors studied the distance running community, which has grown from a small, tight-knit, male-dominated group to a mainstream community with mass participation over the past three decades. Because runners have learned to value diversity, more companies can provide goods and services for the community. Additionally, different types of runners provide each other with social resources such as distinction, status, shared values, and a sense of belonging.

These findings are of interest to companies such as Facebook, Pinterest, and YouTube whose primary business model involves an online community, as well as major brands with associated consumption communities such as Apple, Nike, and Lululemon, or anyone interested in mobilizing social resources to incite broad participation in political campaigns and charities wishing to recruit and maintain donors and volunteers.

"Large, diverse communities experience tension because each different type of member seeks something different from the community and these desires often lead to conflict. However, this tension does not have to lead to conflict, fragmentation, and demise. On the contrary, communities are able to thrive when community members depend on each other for valued economic and social resources," the authors conclude.

The U.S. Census Bureau should reconsider income-based poverty measures in favor of a consumption-based method, according to a new study that strives to more accurately identify the neediest Americans.

The report found that the official poverty rate and the Census Bureau's new Supplemental Poverty Measure — both of which are income-based — do not gauge the extent of poverty as well as a method based on real purchasing ability. Bruce D. Meyer, the McCormick Foundation Professor at the University of Chicago's Harris School of Public Policy, co-authored the study with James X. Sullivan of the University of Notre Dame.

Correctly calculating deprivation helps identify the most disadvantaged individuals and track changes over time. Meyer and Sullivan were surprised to find remarkably little research on the accuracy of the current poverty indicators. Given the measures' importance, the authors decided to investigate how well the different poverty measures work by looking at 25 different characteristics of individuals, from car ownership and house size to education level and appliance ownership. The goal is to assess more of the advantages that individuals have, rather than their income alone.

"Few economic indicators are more closely watched or more important for policy than the official poverty rate," they write in their article, "Identifying the Disadvantaged: Official Poverty, Consumption Poverty and the New Supplemental Poverty Measure," published in the summer 2012 edition of the Journal of Economic Prospectives. "The poverty rate is often cited by policymakers, researchers and advocates who are evaluating social programs that account for more than half a trillion dollars in government spending."

The official poverty measure, the calculation of which has remained virtually unchanged since the 1960s, relies heavily on pre-tax income and long has been criticized for not reflecting the full resources available to families. For example, it does not subtract income tax liabilities, nor does it include tax credits like the Earned Income Tax Credit and the Child Tax Credit or non-cash benefits like food stamps, housing subsidies or public health insurance. In addition, it suffers from significant errors in income reporting by the most disadvantaged families, which is out of the Census Bureau's control.

Although it is based on the same imprecise data sources, the Supplemental Poverty Measure, which the government now publishes in tandem with the official poverty rate, attempts to more effectively estimate poverty by accounting for after-tax income and non-cash benefits. However, Meyer and Sullivan find that these additions have surprisingly counterproductive or mixed effects. In fact, the supplemental measure performs worse than the official poverty rate in capturing the poorest of the poor while including better-off people with higher consumption, more education, home and car ownership, and substantial assets.

"The new poverty measure that many people thought was going to be an improvement, in terms of figuring out who we should call 'poor,' doesn't seem to be an improvement," said Meyer. "It doesn't seem to capture people who are worse off in many dimensions than the official measure. Our consumption measure seems to do better in terms of capturing people who have many different types of disadvantage."

Meyer and Sullivan find their consumption-based measure better captures more of the most disadvantaged than those based on income, because it accounts for savings usage, ownership of durable goods, access to credit and the use of anti-poverty programs. The most disadvantaged families also appear to report their consumption more accurately than income.

"Based on the strength of our findings, we hope that if the Census Bureau wants to revise the official poverty measure, they will rethink what they are doing and go back to the drawing board, instead of proceeding with the Supplemental Poverty Measure," said Meyer.

 If experts thought they knew anything about individual investors, it was this: their emotions lead them to sell winning stocks too soon and hold on to losers too long.

But new research casts doubt on this widely held theory that individual investors' decisions are driven mainly by their feelings toward losses and gains. In an innovative study, researchers found evidence that individual investors' decisions are primarily motivated by their beliefs about a stock's future.

"The story is not about whether an investor hates losing or loves gains — it's not primarily a story about preferences," said Itzhak Ben-David, co-author of the study and assistant professor of finance at Ohio State University's Fisher College of Business.

"It is a story about information and speculation. The investor has a belief about where a stock is headed and that's what he acts on. Investors act more on their beliefs than their preferences."

Ben-David conducted the study with David Hirshleifer of the Paul Merage School of Business at the University of California, Irvine. Their results appear in the August 2012 issue of the journal Review of Financial Studies.

The researchers studied stock transactions from more than 77,000 accounts at a large discount broker from 1990 through 1996 and did a variety of analyses that had never been done before. They examined when investors bought individual stocks, when they sold them, and how much they earned or lost with each sale.

The result was a radical rethinking of why individual investors sell winning stocks and hold on to losers.

The findings don't mean that investors don't have an aversion to losses and a desire to sell winners, Ben-David said. But the trading data suggests that these feelings aren't dominating their decisions.

"People have a variety of reasons for trading stocks, which may include tax issues, margin calls, and an aversion to losses. These all may play a role, but what we show that beliefs are dominant for the trading of retail investors."

The tendency to sell winners too early and to keep losers too long has been called the "disposition effect" by economists.

"The disposition effect has been well-documented. The question is what we make of it. A lot of people look at the data and interpret it as meaning that the typical retail investor is irrational, simply reacting to their feelings about gains and losses," he said.

"But what we find is that, looking at the data, we can't really learn about their preferences. We don't learn about what they like or don't like. Surely, they don't like to lose money — but their reasons for selling stocks are more complex than that."

The simplest test was to see what investors do when a stock is trading just slightly higher or lower than the price they paid — in other words a small winner or a small loser.

If investors really did make stock trades based simply on their pleasure in making money and their aversion to realizing losses, a small winner should lead to more sales than a small loser.

But this study found that investors were not clearly more likely to sell when it was a small winner than when it was a small loser.

Another piece of evidence against the theory that investors' decisions are driven by their aversion to realizing losses was the fact that, the more a stock lost value, the more likely investors were to sell it.

"If investors had an aversion to realizing losses, larger losses should reduce the probability they would sell, but we found the opposite — larger losses were associated with a higher probability of selling," Ben-David said.

Interestingly, the stocks that investors sell the least are those that did not have a price change since purchase.

Another clue is the fact that men and frequent traders were more likely than others to sell winning stocks quickly to reap their profits and sell losers quickly to cut their losses.

"Past research has shown that overconfidence in investing is associated with men and frequent traders," Ben-David said. "They have a belief in their superior knowledge and so you would expect them to buy and sell more quickly than others as they speculate on stock prices. That's exactly what we found. They are engaged in belief-based trading."

The researchers also examined when investors were more likely to buy additional shares of a stock that they had previously purchased. They found that the probability of buying additional shares is greater for shares that lost value than it was for shares that gained value.

That shouldn't happen if investors are really acting on emotions rather than beliefs, Ben-David said.

"If you buy additional shares of a stock that has lost value, that suggests you are acting on your beliefs that the stock is really a winner and other people have just not realized it yet," he said.

"You wouldn't buy additional shares of a losing stock if your biggest motivation was to avoid realizing losses."

However, Ben-David noted that just because investors act on beliefs rather than feelings doesn't mean they are acting rationally.

"They may be overconfident in their own abilities. It is a different kind of irrationality from being averse to selling losers," Ben-David said.

This study's suggestion that investors act more on beliefs than preferences is likely to make waves in the economics profession, he said.

"In economics, these two stories are very different. Beliefs and preferences are very different concepts, and it is important to distinguish them and how they affect investors. Many economists had thought that an irrational aversion to selling losers was crucial for the trading decisions of retail investors."

People will reject an offer of water, even when they are severely thirsty, if they perceive the offer to be unfair, according to a new study funded by the Wellcome Trust. The findings have important implications for understanding how humans make decisions that must balance fairness and self-interest.

It's been known for some time that when humans bargain for money they have a tendency to reject unfair offers, preferring to let both parties walk away with nothing rather than accept a low offer in the knowledge that their counterpart is taking home more cash.

In contrast, when bargaining for food, our closes relatives chimpanzees will almost always accept an offer regardless of any subjective idea of 'fairness'.

Researchers at the Wellcome Trust Centre for Neuroimaging at UCL wanted to see whether humans would similarly accept unfair offers if they were bargaining for a basic physiological need, such as food, water or sex.

The team recruited 21 healthy participants and made 11 of them thirsty by drip-feeding them a salty solution, whilst the remainder received an isotonic solution that had a much smaller effect on their level of thirst. To obtain an objective measure of each individual's need for water, the team measured the salt concentration in their blood. The participants' subjective perception of how thirsty they were was assessed using a simple rating scale.

The participants then separately took part in an ultimatum game. They were given instructions that two of them had been randomly selected to play a game to decide the split of a 500ml bottle of water that could be consumed immediately. One of them would play the part of 'Proposer' and decide how the bottle should be split. The other would be a 'Responder' who could either accept the split and so drink the water offered to them, or reject the split so that both parties would get nothing. The participants knew that they would have to wait a full hour after the end of the game before they would have access to water.

In reality, all of the participants played the part of the Responder. They were presented with two glasses of water with a highly unequal offer that they were told was from the Proposer: the glass offered to them contained 62.5ml, an eighth of the original bottle of water, and the other contained the remaining seven eighths that the Proposer wanted to keep for themselves. They had fifteen seconds to decide whether to accept or reject the offer.

The team found that, unlike chimpanzees, the human participants tended to reject the highly unequal offer, and here that was the case even if they were severely thirsty. The participants' choices were not influenced by how thirsty they actually were, as measured objectively from the blood sample. However, they were more likely to accept the offer if they subjectively felt that they were thirsty.

Dr Nick Wright, who led the study, explains: "Whether or not fairness is a uniquely human motivation has been a source of controversy. These findings show that humans, unlike even our closest relatives chimpanzees, reject an unfair offer of a primary reward like food or water — and will do that even when severely thirsty. However, we also show this fairness motivation is traded-off against self-interest, and that this self-interest is not determined by how their objective need for water but instead by their subjective perception of thirst. These findings are interesting for understanding how subjective feelings of fairness and self-interested need impact on everyday decisions, for example in the labour market."