Survey Methodology

Survey Instrument

Our survey instrument was intended to collect data on faculty backgrounds and labor market links. Drawing on previous surveys by NCES and others, advice from the American Association of Community Colleges (AACC) and other experts, and a pilot test of a draft survey instrument with faculty at two sites in the Los Angeles Community College District, a final survey questionnaire was completed in September 1995. All questions pertained to any individual who had at least some instructional duties during the 1994-1995 academic year. Background items covered instructors' personal characteristics (e.g., age, sex, race, ethnicity), educational background (e.g., years of education, certification and degree status, colleges attended), work experience (e.g., years of labor market experience, type of positions held, current links to employers), and professional status (e.g., salary, full-time/part-time, tenure, subject specialty). Other questions concerned faculty's involvement in various college reform initiatives and use of innovative teaching practices, attitudes toward their job and institution, and the nature and extent of links to their institutions, teaching field, the labor market, and community.

In focusing on links, survey items cover the type and level of effort of the links, and some of the supports (e.g., professional development incentives) and barriers to constructing links. We asked about each of the four domains of linkages--faculty pedagogical/curricula activities, career assistance, professional and community activities, and institutional activities--noted above. The complete survey instrument is found in Appendix B.

Survey Sample

To obtain the survey sample, we first obtained mailing lists of community college faculty from slightly over one hundred randomly selected institutions nationwide (again with the assistance of AACC).[11] From these lists, we then randomly selected about thirty-five hundred names. We included academic and vocational instructors, tenure-track and non-tenure track, full-time and part-time, who had instructional duties in 1994-1995. The survey was administered by mail in late October 1995.[12] Data collection continued until April 1, 1996. During this time, we conducted three mailings and also placed follow-up phone calls; these calls indicated that many nonrespondents simply did not receive the survey due to bad addresses or job changes. The overall response rate was about 64% after excluding refusals, those who had changed schools, undeliverable surveys, and ineligible participants. The final sample consists of 1,725 faculty in 92 institutions.[13]

A profile of respondents is shown in Table 1, which contains selected mean characteristics for all respondents, and separately for academic and vocational faculty. Faculty were divided into four groups based on primary teaching field: academic, vocational, developmental, and other. "Academic" included faculty whose primary teaching field is English, mathematics, physical sciences, biological sciences, social sciences, humanities, and foreign languages. "Vocational" included faculty whose primary teaching field is in education-related subjects, social work, agricultural education, business and office education, health occupations, marketing/distributive education, occupational home economics, consumer and homemaker education, communications or computing, and technology education/industrial arts/trade. For ease of exposition we concentrate on academic and vocational instructors throughout this report. When we refer to "all" faculty, we include academic, vocational, developmental, and "other."

The table shows that community college faculty are overwhelmingly white, about half are male, and the average age is over forty-seven. Compared to academic faculty, vocational faculty tend to be older and less likely to be female or from minority backgrounds. Most community college instructors' highest degree is a Master's (or the equivalent), but almost one-quarter of all academic faculty have a doctorate. About one-third of all faculty have tenure, reflecting the fact that a large number of faculty hold instructor status, and about half are part-time.[14] Interestingly, in our sample, a higher proportion of vocational than academic faculty are part-time.

Our survey provides some institution-level data, including benefits of employment, professional development opportunities, and campus climate. Additional institution-level data from other sources were merged into our sample. Information on a college's region and size were obtained from the 1994-1995 Integrated Postsecondary Education Data System (IPEDS). This was further supplemented by the AACC Annual Survey on the urbanicity of a college and its governance structure (e.g., single-campus, branch campus of a state university, part of a multi-campus district).

	Academic	Vocational	All Faculty
Age	47.6 (9.8)	47.3 (9.0)	47.5 (9.5)
Female	.434	.484	.472
Years teaching in community colleges	12.1 (9.4)	12.1 (8.5)	11.9 (8.9)
Years teaching in current institution	10.7 (9.0)	11.1 (8.3)	10.7 (8.6)
Hispanic	.032	.020	.026
African American	.029	.033	.036
B.A.	.086	.284	.180
M.A.	.688	.565	.624
Ph.D.	.232	.070	.158
Full professor	.156	.150	.149
Associate professor	.089	.095	.094
Assistant professor	.079	.063	.072
Instructor	.277	.337	.307
Adjunct professor	.193	.120	.157
No rank	.136	.156	.141
Urban	.563	.593	.574
Rural	.109	.154	.133
Northeast	.167	.164	.160
North central	.152	.245	.189
West	.319	.253	.304
Single-campus college	.561	.590	.574
Multi-campus district	.233	.176	.201
Total enrollment	10501 (9563)	9408 (8557)	10275 (9380)
Tenured	.338	.346	.335
Faculty represented by union	.567	.569	.574
Part-time	.527	.439	.509
Vocational	---	---	.408
Number of observations	725	703	1725

Note: "Number of observations" refers to maximum number available; means may be based on a smaller sample due to missing observations.

One concern about our sample is whether it is representative of community college faculty nationwide. A point of comparison is the National Survey of Postsecondary Faculty (NSOPF), collected by the National Center for Education Statistics (NCES). This survey was conducted in 1987-1988 and again in 1992-1993 and was designed to produce nationally representative estimates of the characteristics of faculty in two- and four-year institutions using weights supplied by NCES to convert sample statistics. Using over eight thousand responses from public two-year college faculty in 1993-1994, we calculated selected faculty characteristics and compared them with our own sample. The results of this exercise are shown in Table 2.

The table shows that our sample is remarkably similar to NSOPF-93 in terms of faculty gender (53% male in our sample versus 54% in NSOPF-93) and race (88% white in our sample versus 87% white in NSOPF-93). Our respondents are slightly older, of higher rank, and more likely to have tenure than those in NSOPF-93. Overall, however, our final sample is broadly representative of community college faculty nationwide.[15]

Note: NSOPF-'93 is the National Survey of Postsecondary Faculty, 1993. Figures for NSOPF-'93 refer to public two-year college faculty only and are weighted (using NCES weights) to be nationally representative.

[12]In some cases, the survey was mailed to the home of the faculty member; in other cases, it was mailed to the school/departmental address.

[13]2,159 surveys were returned: 61.1% of the initial mailing. It was determined that 337 of these were refusals, people who had changed schools, were no longer teaching, had retired or died, were undeliverable, or were ineligible. We suspect that many of the surveys failed to reach faculty due to incorrect faculty lists and mailing addresses. 1,725 (49%) valid surveys were used from the 3,500 initially mailed.

[14]Our full- versus part-time classification is based on our best estimate of how many hours per week faculty say they work. We arbitrarily define those working more than thirty-five hours per week or more as full-time. Inspection of the data suggest that moving this cutoff (to say, thirty hours per week) does not affect the reported findings in this paper.

[15]It is not possible to compare our sample to NSOPF-93 on some important dimensions--for example, part-time status--due to differences in survey items. It should be noted, too, that conversations with NCES staff suggest they have considerable difficulties in calculating accurate sample weights for these items.