Searching for gaydar: Blind spots in the study of sexual orientation perception

Miller, Arianne E.

doi:10.1080/19419899.2018.1468353

Cited by 25 publications

(16 citation statements)

References 67 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…binary and static, that those who are openly gay on social media are representative of the whole gay population, and ignoring gender and racial variance-distort important difference makers in real-world populations (Miller [2018]). To make matters worse, existing scientific and social-scientific evidence either speaks against PHT theory, and against a dependency relation between facial features and sexual orientation or other personality traits (LeVay [1996(LeVay [ ], [2010; Magnet [2011]; Mustanski et al [2002]), or speaks against gender atypical traits being the driving factor (Valentova et al [2014]).…”

Section: Differences In Understanding; Differences In Link Uncertaintymentioning

confidence: 99%

Understanding from Machine Learning Models

Sullivan

2022

The British Journal for the Philosophy of Science

100

View full text Add to dashboard Cite

Abstract Simple idealized models seem to provide more understanding than opaque, complex, and hyper-realistic models. However, an increasing number of scientists are going in the opposite direction by utilizing opaque machine learning models to make predictions and draw inferences, suggesting that scientists are opting for models that have less potential for understanding. Are scientists trading understanding for some other epistemic or pragmatic good when they choose a machine learning model? Or are the assumptions behind why minimal models provide understanding misguided? In this article, using the case of deep neural networks, I argue that it is not the complexity or black box nature of a model that limits how much understanding the model provides. Instead, it is a lack of scientific and empirical evidence supporting the link that connects a model to the target phenomenon that primarily prohibits understanding. 1Understanding from Minimal and Complex Models2Algorithms, Explanatory Questions, and Understanding3Black Boxes3.1Implementation black boxes3.2Levels of implementation black boxes4The Black Boxes of Deep Neural Networks4.1Deep neural network structure4.2Deep neural network modelling process4.3Levels of deep neural network black boxes5Understanding, Explanation, and Link Uncertainty5.1Deep neural networks and how-possibility explanations5.2Deep neural networks and link uncertainty5.3Differences in understanding; differences in link uncertainty6Conclusion

show abstract

Section: Differences In Understanding; Differences In Link Uncertaintymentioning

confidence: 99%

Understanding from Machine Learning Models

Sullivan

2022

The British Journal for the Philosophy of Science

100

View full text Add to dashboard Cite

show abstract

“…There are a number of examples of inappropriate uses of machine learning. One example is the infamous attempt to use deep learning to predict a person's sexual orientation from a photograph, which was widely criticised for many reasons, including that such applications would have much potential for misuse in terms of supporting persecution and hate crime [22]. In the facial recognition context, there has been much debate regarding the acceptability of such technologies, leading some governments to institute a ban on their use by the public sector [6].…”

Section: Potential For Misusementioning

confidence: 99%

Monitoring Misuse for Accountable 'Artificial Intelligence as a Service'

Javadi

Cloete

Cobbe

et al. 2020

Proceedings of the AAAI/ACM Conference on AI, Ethics, and Society

View full text Add to dashboard Cite

AI is increasingly being offered 'as a service' (AIaaS). This entails service providers offering customers access to pre-built AI models and services, for tasks such as object recognition, text translation, text-to-voice conversion, and facial recognition, to name a few. The offerings enable customers to easily integrate a range of powerful AI-driven capabilities into their applications. Customers access these models through the provider's APIs, sending particular data to which models are applied, the results of which returned.However, there are many situations in which the use of AI can be problematic. AIaaS services typically represent generic functionality, available 'at a click'. Providers may therefore, for reasons of reputation or responsibility, seek to ensure that the AIaaS services they offer are being used by customers for 'appropriate' purposes.This paper introduces and explores the concept whereby AIaaS providers uncover situations of possible service misuse by their customers. Illustrated through topical examples, we consider the technical usage patterns that could signal situations warranting scrutiny, and raise some of the legal and technical challenges of monitoring for misuse. In all, by introducing this concept, we indicate a potential area for further inquiry from a range of perspectives.

show abstract

“…Most commonly, people’s ability to correctly perceive another person’s SO by using implicit signals is referred to as “gaydar” (Rule, 2017; Tskhay & Rule, 2013a). However, most “gaydar” studies examined whether raters are able to distinguish between lesbian/gay and straight targets with above-chance accuracy (Miller, 2018). There is growing controversy whether “gaydar” can be conceptualized in terms of accuracy or if it is merely stereotyping (Miller, 2018), that is, gender-role (non)conformity is the integral part when expressing and perceiving SO.…”

Section: Signal and Gender Differences In So Perceptionmentioning

confidence: 99%

“…However, most “gaydar” studies examined whether raters are able to distinguish between lesbian/gay and straight targets with above-chance accuracy (Miller, 2018). There is growing controversy whether “gaydar” can be conceptualized in terms of accuracy or if it is merely stereotyping (Miller, 2018), that is, gender-role (non)conformity is the integral part when expressing and perceiving SO. SO perception studies have shown that there is a correspondence between self-reported and attributed SO.…”

Section: Signal and Gender Differences In So Perceptionmentioning

confidence: 99%

“…However, for female targets, valid SO information contained in videos with sound did not exceed full appearances but both did exceed speech alone (Rieger et al, 2010). However, comparative research on different implicit signals is lacking (voices vs. faces vs. combined vocal and facial stimuli; Miller, 2018; Sulpizio et al, 2019). Based on these theoretical considerations, we tested the following hypotheses:…”

Section: Signal and Gender Differences In So Perceptionmentioning

confidence: 99%

See 1 more Smart Citation

Gender (Conformity) Matters: Cross-Dimensional and Cross-Modal Associations in Sexual Orientation Perception

Kachel

Steffens

Preuß

et al. 2019

Journal of Language and Social Psychology

View full text Add to dashboard Cite

Although sexual orientation (SO) is perceptually ambiguous, people are able to detect it with above-chance accuracy from faces and, sometimes, from voices. Despite a multitude of “gaydar” studies, it is unclear (1) whether vocal or facial signals carry more SO information, (2) whether raters refer to target’s SO instead of gender-role conformity when forming SO impressions, and (3) whether there are any differences for female and male targets. We collected face photographs, voice recordings, and self-reported gender-role conformity of 18 lesbian/gay and straight female and male target persons each. Study 1 (rating of SO) showed that faces led to higher accuracies than voices, which was especially true for female targets. Study 2 (rating of gender-role conformity) showed that the link between self-reported and attributed SO was mediated by self-reported and attributed gender-role conformity. Results support the centrality of gender-role conformity, more than that of SO, in impression formation.

show abstract

Searching for gaydar: Blind spots in the study of sexual orientation perception

Cited by 25 publications

References 67 publications

Understanding from Machine Learning Models

Understanding from Machine Learning Models

Monitoring Misuse for Accountable 'Artificial Intelligence as a Service'

Gender (Conformity) Matters: Cross-Dimensional and Cross-Modal Associations in Sexual Orientation Perception

Contact Info

Product

Resources

About