Summary: New research suggests that the gendered characteristics of service robots can influence customer decisions in the hospitality industry. Robots with masculine characteristics were more persuasive to women, who had a lower sense of personal power, while customers with more power felt less impressed.
However, adding “cute” design features, such as large eyes and round faces, neutralized these gender effects, leading to more equal responses across genders. The results offer insights into how companies can strategically design service robots to reduce or eliminate gender-related effects.
Important facts:
- Robot gender belief: Robots that were characterized as masculine were more convincing to women with a less sense of power.
- Beautiful design neutralizes gender: Features such as large eyes and round faces reduced the effect of gender.
- Commercial applications: Hospitality companies can strategically design robots to increase sales, persuade people, or avoid gender stereotypes.
Source: Penn State
New research from a team at Penn State’s School of Hospitality Management has found that the hotel industry can leverage the gender characteristics of service robots to influence customer decisions.
According to researchers, assistance robots with characteristics typically associated with men may be more persuasive when interacting with women who have a lower sense of power.
The team also found that “cute” features in a robot’s design (such as large eyes and bulging cheeks) can reduce the effect of the robot’s gender on persuasion. Male and female users responded equally to robots with these “cute” features.
Lavi Peng, a doctoral candidate; Anna Mattila, Marriott Professor of Lodging Management; and Amit Sharma, the Edward Friedman and Stuart Mann Professor of Hospitality Management, all at Penn State, led the research.
Their results were published in the Journal of Hospitality and Tourism Management.
“Robots can be designed or programmed to have human characteristics, such as names, voices and body shapes, that reflect gender,” said Mattila.
In addition to the robot’s gender, the user’s sense of power (how individuals perceive their ability to influence others or their environment) can also affect the success of a service robot in making recommendations.
Researchers conducted two studies to explore how the gender of service robots can influence consumer decisions.
The first study surveyed 239 people recruited through Amazon Mechanical Turk. They were asked to rate their sense of empowerment before imagining going to a new restaurant and getting a breakfast burrito menu recommendation from a service robot.
The service robots presented in the study were identical, except that they used gray or pink to represent men and women, respectively. After the menu recommendation, participants rated the robot’s persuasiveness.
“We found that women with a low sense of empowerment were more likely to accept male robot recommendations,” Peng said.
Among men with low power views, the difference was less clear. According to our results, high-power users generally make their own decisions, independent of social expectations. They are more confident and want to make decisions based on their own judgment.
According to the researchers, restaurants can use these findings to determine what types of service robots they want to deploy. For example, they could use “male” robots to recommend new menu items. The results suggest that robots with characteristics typically associated with men may have a greater influence on customer decisions.
According to the researchers, hotels could also use these findings to determine the gender characteristics they want to use in robots that convince guests to upgrade their rooms.
“Upselling and upgrades rely on persuasion, and our research suggests that robots with masculine characteristics can be effective,” Peng said. “If a company knows that its customer is female, it may consider deploying robots with different gender characteristics than for a male customer.”
The second study examined how companies could reduce gender stereotypes in robot design or reduce the impact of “male” robots on users who have a low sense of empowerment.
Because the results of the first study showed that the gender displayed in robots particularly affected low-power users, the researchers recruited 156 college students in the United States.
According to the researchers, previous research has shown that students typically hold subordinate positions or are dependent on teachers who have authority over their learning outcomes. This means they are a low-power group.
In the second study to change the gender representation of robots, researchers used an iPad screen displaying different gender-specific facial features projected onto a Bear Robotics Servi robot, which has no human features.
These facial features had beautiful designs, such as round faces and large eyes. After meeting and interacting with the robot, participants completed a computer scenario to evaluate the robot’s recommendation of avocado toast.
“Male and female users showed similar responses to both male and female robot designs,” Peng noted.
Companies that want to combat gender stereotypes might consider making their robots more attractive.
Funding: The Marriott Foundation supported this research.
Abstract
Gendered Robots and Persuasion: The Interaction Between Robot Gender, User Gender, and Their Power in Menu Recommendations
Despite extensive research on robot anthropomorphism, few studies have examined the role of robot gender in influencing persuasive tasks, such as recommendations.
This study examines the combined effects of robots’ gender, consumers’ gender, and their sense of power on accepting menu recommendations.
Research 1 shows that powerless female users find masculine-looking robots (compared to feminine-looking robots) more convincing and more accepting.
This difference between the sexes disappears in impotent men and power users of both sexes.
In Study 2, cleverness is tested as a precondition for the gender effect of robots. This shows that both male and female users accept recommendations that sound equally good.
In principle, this research advances our understanding of gender dynamics in human-robot interaction in the context of persuasive service delivery.
In practice, it offers insights into the strategic design of robots according to user profiles. It also sheds light on how cleverness can reduce gender bias.
