Extended Data Table 8.

Survey Response Rates

This table describes the details of the three pilots and main study. The first column details the date of the first e-mail, while the second column details the date of the reminder e-mail, typically two weeks after the first e-mail. The main study had two reminder e-mails because on March 1st, an error in the Qualtrics survey system caused an abnormally high number of e-mails to bounce (~11,000). On March 9th, after the error had been addressed, we re-sent reminder e-mails to those respondents that had been missed due to the error. Column three describes the sampling strategy; random sampling was used for the pilots, but once we learned that there were far fewer women in the population than there were men, we adjusted to a gender-stratified sampling strategy in order to gain enough power for two-sided t-tests comparing responses from men and women. Specifically, 10,000 (imputed) ORCID profiles belonging to men and 10,000 (imputed) ORCID profiles belonging to women were randomly selected to receive the survey in addition to 6,500 profiles that had gender ambiguous names. Column 4 indicates the number of e-mails sent, while columns 5-7 document the response rate. A large fraction of emails either bounced or received an automated response.