arcp://name,uts_public_data_repo/5a380da085c111ef88e8710d438700035a380da085c111ef88e8710d43870003http://schema.org/descriptiondescription"We extract data from two databases (web of science and Academic Analytics). For the web of science database we identify the outputs of each author. For example, an author might be listed as K Walsh, K D Walsh, Kathleen Walsh or Kathy Walsh. We investigate (using web searches of university websites and linkedIn) whether the publications of each author should be grouped under one author ID or several.
The web of science file doesn’t include gender, so this needed to be categorised manually. We augment standard gender methods such as those used in Faccio, Marchica and Murab (2016) and Adams et al (2018) and accessed databases of first names classified by gender. Authors were coded male or female if their names were unambiguously one gender (e.g John or Mary). For the approximately 7000 ambiguous names (e.g. unusual or foreign names or names such as Michelle which can be either male or female) we searched for the specific author using either names or paper titles and determine gender using photos or pronouns in their webpage, university profile, LinkedIn profile, ratemyprofessor.com or other online reference to the author or paper. "
.