Saying 'a girl is the champion' lowered adults' view of women's ability in an invented activity

The invented activities are the core of the design. With "Gazzer", nobody arrives believing that men or women are better at it, so any gap between the two wordings has to come from the wording ^[3]. In some experiments the team swapped gender for a made-up group, the "Daxes", to ask whether naming any group causes trouble or only naming a familiar one ^[4]. The effect disappeared when the messages named the Daxes instead of a gender, and the authors take that to mean that people's existing beliefs about women and men play an important role ^[9].

The route from wording to belief is an inference. The researchers reason that mentioning gender may suggest girls or women are unusual or unexpected in the role ^[12]. Andrei Cimpian, the senior author and a psychology professor at New York University, said the work shows that "subtle communicative inferences can cause well-intentioned gendered messaging to undermine its own goals" ^[15]. April Bailey of the University of Edinburgh is the lead author, and Maya Kreitman of the Chicago School is the third ^[14].

Cimpian was explicit that the finding is no argument for silence. "What's counterproductive is highlighting women's gender as the reason for celebration," he said ^[10]. Bailey offered an alternative: "Rather than emphasizing that an achievement is by or for women, it may be more effective to simply show women succeeding in these fields without making gender the focus," she said ^[11].

In our view the practical lesson is to test the wording before a campaign ships, and the design to copy is in the paper: the same achievement described once with a gender and once with a name, then a read on what people believe and whether they want to take part ^[2]. That is a cheap experiment for a team to run on its own slogans.

The limits are real. The stimuli were sentences about a champion, not slogans. "Fight like a girl" and "The future is female" appear in the write-up as examples of campaign wording, not as tested items ^[13]. The measured outcomes were beliefs about how many women take part, adults' ratings of women's ability, and stated interest in the activity [c5, c6, c7]. The participants were in the United States ^[1].

The study cannot say how large the loss would be in a real campaign. Readers of a real campaign already hold views about the activity, and the invented-word design removed those views on purpose, so the effect there could be larger or smaller. The summary does not include effect sizes.