In the fall of 2017, actor Alyssa Milano responded to accusations of sexual harassment against Harvey Weinstein by tweeting.

“If you’ve been sexually harassed or assaulted write ‘me too’ as a reply to this tweet,” she wrote.

The #MeToo movement had been around for a decade at that point — it was originally started by civil rights activist Tarana Burke — but in 2017 it became part of a national conversation. What followed was a flood of stories and public accusations of sexual misconduct, and a sense that something significant had shifted.

At the same time, a sociology graduate student at Stanford University started a series of experiments.

Chloe Grace Hart asked study participants to read an employee file about a fictitious sales associate with satisfactory performance named “Sarah.” Their task was to decide whether or not to recommend her for promotion.

There were five different versions of Sarah’s employee file. The files were identical, except for their mention of harassment.

Here are the five versions:

- A file with no notes of harassment.

- A file with a note that Sarah experienced non-sexual harassment perpetrated by a co-worker. (The co-worker repeatedly shouted and swore at her.) In this file, another co-worker reported the harassment, not Sarah herself.

- A file with a note that Sarah reported non-sexual harassment. (The shouting and swearing mentioned above.)

- A note that Sarah had reported sexual harassment. (A co-worker repeatedly made sexual comments about Sarah’s body.)

- A note that a co-worker reported that Sarah had experienced sexual harassment. (The sexual comments mentioned above.)

In the fall of 2017, the study participants were a bit more likely than not to recommend her for promotion in four out of the five cases.

Which case was different?

The case in which Sarah reported the sexual harassment herself. It hurt her chances for promotion.

“In the case where she had self-reported sexual harassment … they were as likely to not want to promote her as promote her,” Hart said. “This study uses causal evidence … to show that it is specifically the self-reporting sexual harassment that does lead her to take a hit.”

But that changed after the #MeToo movement gained force.

Hart repeated the experiment several times over the next few months, using the same five employee files. By February 2018, as #MeToo was dominating the headlines, the negative effect of Sarah reporting the sexual harassment had vanished. At that point, there was no difference in how the participants reacted to the five different versions of Sarah’s file.

Hart isn’t ready to say that the #MeToo movement succeeded in eliminating bias when it comes to reporting sexual harassment.

“I think, at the very least, we can take this information to understand that the way we see sexual harassment targets is malleable,” Hart said. “We don't have to just see them one way in society because things like people coming out and sharing their experiences can actually shift our cultural understanding of who women reporting sexual harassment are.”

Another important takeaway: It helps to stand up for your co-workers.

“If there's a bystander, someone who's witnessed the harassment … my experiment suggests that that could be helpful in how [the targets] are viewed,” Hart said.