The next step was to use the Journal Builder Files to begin to create the DA. In 1990, ASA began to use Journal Builder, a computerized process for articles submitted to all ASA journals. Journal Builder provides an accurate count of every unique manuscript submitted to each journal each year, its title, author(s), reviewers, transaction dates, reviewer decisions and final outcome, all connected through a unique Journal Builder manuscript number. A comma delimited file was developed from the Journal Builder files, so that multiple authors each have their own line in the database, connected by the manuscript number and the version number. All names of participants (authors and reviewers) taken from the Journal Builder Files have been uniformly formatted (Last Name, First Name, Middle Name, Suffix). A later section of this codebook will contain more information on the Journal Builder File.
Next, we disambiguated the files by removing all duplicate listing of authors and reviewers. We curated the names of all participants in the Journal Builder Files, removing redundancy through using Google search to determine if John Q. Public was the same person as John Public. We then coded the name with the most information (in this case John Q. Public as 1 and John Public as 0). This disambiguated process was used for the reviewer, author, and person files that were created.
The purpose of the next two steps was to add additional information to populate the appropriate file using the ASA Graduate Department Guide (Grad Guide) and the ASA membership data base. ASA publishes annually the Grad Guide of sociology departments that offer graduate degrees and that wish to be included in the Guide. This guide is organized by the name of university and the name of department. Those Guides from 1990-2010 (including Word, PDF, and SQRL files) were transformed into Excel files. They include institutional affiliation at the time of the article or review submission, PhD granting institutions, current institution, and rank (which will vary over time). These files can be used to see career trajectories (including changes in institutional affiliation and rank) over time. The ASA membership files, are filled out by ASA members on an annual basis at the beginning of each membership year. Not all sociologists are members of ASA. For those who were ASA members during this time-period, we extracted gender, race, and ethnicity, current institution, as well as current email addresses for the most recent year of membership. For those demographic characteristics and email addresses that could not be obtained through the membership files, we engaged in Google searches. The permissions survey that we sent out to authors and reviewers also included information of race, ethnicity, gender, and institution that could be used to populate the files when membership information was missing (more information on this survey see below).
Simultaneously, the process of scanning and digitizing the manuscripts, of varying age and conditions, into PDF files began. A scanning firm (Modern Image) was hired and the initial process of for transferring and scanning the materials was begun. The result was 10,551 manuscript files, including all versions of the manuscript, all reviews, and all letters and comments by the editors. We had hoped that, as part of the scanning process, Modern Image would delete non-relevant information from the files (e.g. pages from another manuscript). But, this additional process was not undertaken, so the CSSR needed to curate each manuscript file as part of the curation process.
Once these activities were complete we sent out an on-line survey to each author or reviewer in the file in order to obtain their permission to include the content of their manuscripts and/or reviews in the DA. We created a brief questionnaire in Qualtrics so that authors and reviewers could approve all of their manuscripts and /or reviews in the DA. Of the 10,551 manuscripts, we found contact information for 18,554 authors, and 26,693 reviewers to whom we sent the survey. We received permission for more than 5,000 manuscripts, almost half of the total.
The title of each manuscript or review submitted was listed separately in the survey format, making it easy to check yes or no. If they did not wish to give an across-the-board permission or they did not wish to be included in the DA, they were able to select specific articles or reviews that they wished to be included and those that they did not wish to be included. In the case of co-authors, all parties needed to agree or the manuscript was not included. Non-responses were considered a denial of permission. Authors also had the opportunity to decide if they would allow us to include their name with your manuscripts or not. Names of reviewers, were not be included in the DA. We send out five reminders. Then, to increase the number of individuals who gave permission, personal emails were sent by the co-PIs Roberta Spalter-Roth, Jean Shin, and James Witte. This strategy was successful and increased the number of survey respondents.