I really think this research is so well designed to address the problem of gender bias expressions in job contents given so many limitations. My question relates to how legitimate the algorithms in Textio and Unitive can be used as baseline algorithms, especially the given "gendered word list". As I look into the examples of male biased words and female biased words in figure 13, the measures to classify gendered biased words seem to be ambiguous for me. I don't quite get the point why words like "success", "build", "drive" are masculine while words like "support", "need" or "relationship" are feminine. I mean just by looking at these words, I cannot see they show any masculine or feminine characteristics. Therefore I highly doubt whether they are inherently masculine or feminine or it may be the case that people relate them with any masculine or feminine feature because of their stereotypes of these words. Then the problem becomes how to change people's stereotypes of these words and make these words become "gendered-neutral" in people's perception instead of replacing them with other "gendered-neutral" words. As from my perspective, I think words "success", "build", "drive" can totally be gendered-neural and it's the same with words like "support", "need", or "relationship". What's your opinion on this? Thank you!
I really think this research is so well designed to address the problem of gender bias expressions in job contents given so many limitations. My question relates to how legitimate the algorithms in Textio and Unitive can be used as baseline algorithms, especially the given "gendered word list". As I look into the examples of male biased words and female biased words in figure 13, the measures to classify gendered biased words seem to be ambiguous for me. I don't quite get the point why words like "success", "build", "drive" are masculine while words like "support", "need" or "relationship" are feminine. I mean just by looking at these words, I cannot see they show any masculine or feminine characteristics. Therefore I highly doubt whether they are inherently masculine or feminine or it may be the case that people relate them with any masculine or feminine feature because of their stereotypes of these words. Then the problem becomes how to change people's stereotypes of these words and make these words become "gendered-neutral" in people's perception instead of replacing them with other "gendered-neutral" words. As from my perspective, I think words "success", "build", "drive" can totally be gendered-neural and it's the same with words like "support", "need", or "relationship". What's your opinion on this? Thank you!