Gender Bias and Conversational Agents: an ethical perspective on Social Robotics

Fossa, Fabio; Sucameli, Irene

doi:10.1007/s11948-022-00376-3

Cited by 23 publications

(12 citation statements)

References 53 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Thus, given the results reported here, and in the light of the other studies cited above, it can be said that there is probably a need for a proactive approach, i.e. one that involves ethically acceptable design of artificial systems (Fossa and Sucameli 2022). It would then be possible to implement technologies that counteract discrimination towards people suffering from negative stereotypes.…”

Section: Discussionmentioning

confidence: 59%

Gender role stereotypes at work in humanoid robots

Parlangeli

Palmitesta

Bracci

et al. 2022

Behaviour & Information Technology

View full text Add to dashboard Cite

Research on gender role stereotypes activated in interactions with humanoid robots has yet to produce conclusive knowledge. To analyse how much, and in which way, gender role stereotypes used in interactions with humans are also called into play in interactions with humanoid robots, a study was conducted with 240 participants. The study was an online survey in which a scale was used for determining the appropriateness for performing four stereotypically masculine and four feminine by humans and robots. Overall, eight humanoid robotsfour judged feminine and four masculinewere considered. Results showed that gender role stereotypes are activated for both genders, but men most strongly activate those pertaining to male roles. These stereotypes are also adopted in reference to humanoid robots, though robots are generally considered less suitable for performing female roles. Furthermore, an increased degree of similarity of robots to humans has a positive effect in assessing the appropriateness to perform female roles only for female robots. The same does not happen with male robots. These results suggest that male and female robots are not categorised in the same way. Robots are essentially perceived as male entities, a particularly relevant hypothesis for the gender-sensitive design of humanoid robots.

show abstract

Section: Discussionmentioning

confidence: 59%

Gender role stereotypes at work in humanoid robots

Parlangeli

Palmitesta

Bracci

et al. 2022

Behaviour & Information Technology

View full text Add to dashboard Cite

show abstract

“…Participants were told that they would be participating in a survey about how people react to different voices, and following the completion of a consent form, they heard each of the 4 Siri voices reading the Rainbow Passage (Fairbanks, 1960), in randomized order. Listeners were initially asked if they heard the clip well, and then were permitted to play the clip as many times as they wanted.…”

Section: Methodsmentioning

confidence: 99%

“…Additionally, listeners in studies of synthesized voices also reproduced the types of social biases that researchers have observed in studies of natural voices, especially with respect to gender. For example, several studies have found that humans are more likely to be abusive to digital assistants with female names and voices than those with male names and voices (Penny, 2016;Fossa and Sucameli, 2022). Similarly, Jackson et al (2020) found that listeners judge "female-sounding" assistants more harshly than "malesounding" robots when they do not comply with user directions, indicating gendered expectations about robot compliance (Jackson et al, 2020).…”

Section: Perception Of Synthesized Voicesmentioning

confidence: 99%

Siri, you've changed! Acoustic properties and racialized judgments of voice assistants

Holliday

2023

Front. Commun.

View full text Add to dashboard Cite

As speech technology is increasingly integrated into modern American society, voice assistants are a more significant part of our everyday lives. According to Apple, Siri fulfills 25 billion requests each month. As part of a software update in April 2021, users in the U.S. were presented with a choice of 4 Siris. While in beta testing, users on Twitter began to comment that they felt that some of the voices had racial identities, noting in particular that Voice 2 and Voice 3 “sounded black.” This study tests whether listeners indeed hear the different Siri voices as sounding like speakers from different groups, as well as examines voice quality features that may trigger these judgments. In order to test evaluations of the four voices, 485 American English listeners heard each Siri voice reading the Rainbow Passage, via online survey conducted on Qualtrics. Following each clip, listeners responded to questions about the speaker's demographic characteristics and personal traits. An LMER model of normalized ratings assessed the interaction of voice and race judgment revealed that indeed, Voice 2 and Voice 3 were significantly more likely to be rated as belonging to a Black speaker than Voices 1 and 4 (p < 0.001). Per-trait logistic regression models and chi-square tests examining ratings revealed Voice 3, the male voice rated as Black, was judged less competent (X2 = 108.99, x < 0.001), less professional (X2 = 90.97, p < 0.001), and funniest (X2 = 123.39, x < 0.001). Following analysis of listener judgments of voices, I conducted post-hoc analysis comparing voice quality (VQ) features to examine which may trigger the listener judgments of race. Using PraatSauce, I employed scripts to extract VQ measures previously hypothesized to pattern differently in African American English vs. Mainstream American English. VQ measures that significantly affected listener ratings of the voices are mean F0 and H1–A3c, which correlate with perceptions of pitch and breathiness. These results reveal listeners attribute human-like demographic and personal characteristics to synthesized voices. A more comprehensive understanding of social judgments of digitized voices may help us to understand how listeners evaluate human voices, with implications for speech perception and discrimination as well as recognition and synthesis.

show abstract

“…Our efforts culminated in finding a wide array of areas for improvement, particularly concerning the role of psychological factors [91], gender and intersectionality considerations, and protective stops that play a crucial role in ensuring user safety [61,70]. Diversity, inclusion, and intersectional considerations have been identified in the literature as potential access and discriminatory problems for 'non-average' body sizes, genders, and pediatric populations [18,36,42,54,98]. Persons differ considerably in size and body form, so accommodating robots to each potential user is a complicated task, yet there seems to be no alternative to embracing complexity and Fig.…”

Section: Findings: Diversity Observations Concerning Lower-limb Exosk...mentioning

confidence: 99%

Accounting for Diversity in Robot Design, Testbeds, and Safety Standardization

Fosch‐Villaronga

Drukarch

2023

Int J of Soc Robotics

View full text Add to dashboard Cite

Science has started highlighting the importance of integrating diversity considerations in medicine and healthcare. However, there is little research into how these considerations apply, affect, and should be integrated into concrete healthcare innovations such as rehabilitation robotics. Robot policy ecosystems are also oblivious to the vast landscape of gender identity understanding, often ignoring these considerations and failing to guide developers in integrating them to ensure they meet user needs. While this ignorance may be for the traditional heteronormative configuration of the medical, technical, and legal world, the ending result is the failure of roboticists to consider them in robot development. However, missing diversity, equity, and inclusion considerations can result in robotic systems that can compromise user safety, be discriminatory, and not respect their fundamental rights. This paper explores the impact of overlooking gender and sex considerations in robot design on users. We focus on the safety standard for personal care robots ISO 13482:2014 and zoom in on lower-limb exoskeletons. Our findings signal that ISO 13482:2014 has significant gaps concerning intersectional aspects like sex, gender, age, or health conditions and, because of that, developers are creating robot systems that, despite adherence to the standard, can still cause harm to users. In short, our observations show that robotic exoskeletons operate intimately with users’ bodies, thus exemplifying how gender and medical conditions might introduce dissimilarities in human–robot interaction that, as long as they remain ignored in regulations, may compromise user safety. We conclude the article by putting forward particular recommendations to update ISO 13482:2014 to reflect better the broad diversity of users of personal care robots.

show abstract

Gender Bias and Conversational Agents: an ethical perspective on Social Robotics

Cited by 23 publications

References 53 publications

Gender role stereotypes at work in humanoid robots

Gender role stereotypes at work in humanoid robots

Siri, you've changed! Acoustic properties and racialized judgments of voice assistants

Accounting for Diversity in Robot Design, Testbeds, and Safety Standardization

Contact Info

Product

Resources

About