SURVEY METHODOLOGY

Our survey was mailed to the faculty in the departments of biology, biochemistry, chemistry, computer science, engineering, and physics at a large private Midwestern University in May of 1997. Emeritus, adjunct, and visiting faculty members were not contacted because they are normally peripheral to department activity. Data collection was completed by September 1997. The survey was diskette-based and computer-administered and took approximately 30 minutes to complete. The electronic survey guided respondents through the items (with appropriate branching) and created a data file on the diskette, which was returned to us. This procedure reduced errors of data entry by respondents and data entry assistants, and ensured that all respondents answered all items. All faculty members were sent a cover letter that briefly described the purpose of the study, assured confidentiality, gave references to our previous work in scientific journals of the hard sciences and at the National Science Foundation, and described the instructions for the use of the survey. Included along with the cover letter were a diskette and a pre-addressed envelope for returning the diskette. Table A1 describes the characteristics of our sample.

Faculty members who did not respond in three weeks or who used Macintosh computers were asked to set up a time for a telephone interview by a trained research assistant who entered responses in real time into a PC. Unfortunately, the biology department used Macintosh computers exclusively, which lowered the response rate of this department.