A two-stage stratified, clustered probability design was used to select the NSOPF:99 sample. The institution universe for NSOPF:99 was defined by the following criteria: Title IV participating, degree-granting institutions;1 public and private not-for-profit institutions;2 institutions that conferred associate, bachelors, or advanced degrees; and institutions that were located in the United States. This definition covered most colleges (including junior and community colleges), universities, and graduate and professional schools. It excluded institutions that either offered only less-than-2-year programs; were private for-profit; or were located outside the United States (e.g., in U.S. territories). In addition, it excluded institutions that offered instruction only to employees of the institutions, tribal colleges, and institutions that offered only correspondence courses. According to the NCES Integrated Postsecondary Education Data System, 3,396 institutions met these criteria and were eligible for the NSOPF:99 sample. The first-stage sampling frame consisted of this group of institutions, stratified based on the highest degrees offered and the amount of federal research dollars received. The strata distinguished public and private institutions, as well as institution type based on the Carnegie Foundations classification system.3
Each institution was asked to complete an Institution Questionnaire and to provide a list of all faculty and instructional staff at their institution. The faculty universe for NSOPF:99 included all those who were designated as faculty, whether or not their responsibilities included instruction, and other (nonfaculty) personnel with instructional responsibilities. Under this definition, researchers, administrators, and other institutional staff who hold faculty positions, but who do not teach, were included in the sample. Instructional staff without faculty status also were included. Teaching assistants were not included in NSOPF:99.4 Institution coordinators were asked to provide a list of all full- and part-time employees who had faculty status or instructional responsibilities during the 1998 fall term (i.e., the term that included November 1, 1998).
Of the 960 institutions in the sample, one was ineligible because it had merged with another institution. A total of 818 institutions provided lists of faculty and instructional staff, for a weighted list participation rate of 88.4 percent. A total of 865 institutions returned the institution questionnaire, for a weighted response rate of 92.8 percent. Initially, 28,576 faculty and instructional staff were selected from institutions that provided a list of their faculty and instructional staff. Subsequently, a subsample of 19,813 faculty and instructional staff was drawn for intensive followup. Approximately 18,000 faculty and instructional staff questionnaires were completed, for a weighted response rate of 83.0 percent. The overall weighted faculty response rate (institution list participation rate multiplied by the faculty questionnaire response rate) was 73.4 percent.
Faculty nonresponse bias analyses did not detect any bias. Item nonresponse occurred when a respondent did not answer one or more survey questions. The item nonresponse rates were generally low for the faculty questionnaire. For more information about NSOPF:99, including a full description of faculty and item nonresponse, see the 1999 National Study of Postsecondary Faculty: Methodology Report (NCES 2002154).