In winter of 2013, ACM conducted its first-ever survey of non-doctoral-granting academic departments in computing (NDC). The survey requests information on recent degrees, enrollments, faculty demographics and faculty salaries as well as gender and ethnic diversity characteristics of the faculty and students in the computing programs. It is designed to complement the Taulbee Survey of doctoral-granting departments in computing conducted by the Computing Research Association. This article reports the results of the first NDC Study, with comparisons and contrasts to data reported in the Taulbee Survey.
In the winter/spring of 2013, ACM conducted the first-ever survey of non-doctoral-granting departments in computing (NDC), intended to be an annual complement to the Computing Research Association (CRA) Taulbee Survey of Ph.D.-granting departments in computing . The ACM NDC Study (subsequently referred to as NDC Study) was conducted with support from ACM,1 Google,2 and CRA.3 The authors of this report comprised the NDC Steering Committee. Our survey was informed by and built upon the work of the two pilot TauRUs ("Taulbee for the Rest of Us") surveys [1,4] as well as the CRA Taulbee Survey. Undertaken annually, the NDC Study will help fill in gaps in data on non-Taulbee programs to present a more complete view of the academic landscape in computing and to expand pipeline information on programs that produce candidates for Ph.D. programs as well as the private and public labor markets.
The goals of the NDC Study are to document trends in student enrollment, degree production, faculty demographics and salaries at not-for-profit U.S. academic institutions that grant bachelor's and/or master's degrees (but not Ph.D.s) in the five major computing disciplines: computer science (CS), computer engineering (CE), information systems (IS), information technology (IT), and software engineering (SE). Diversity statistics and trends with respect to students and faculty are important features of this documentation.
The NDC survey was distributed in February 2013 to qualifying programs identified using data in the Integrated Post-secondary Education Data System (IPEDS) . These data are collected annually by the National Center for Education Statistics (NCES) from all U.S. institutions that participate in the federal financial aid programs . There were 926 surveys distributed to academic units (departments or institutions) identified through IPEDS as offering at least one program in computing. In some cases, a single institution received multiple surveys if programs are housed in different departments within the institution. Responses were received for 93 academic units and data were reported for 191 total programs (160 bachelor's and 31 master's level). We found that 83 out of the 93 responding academic units provided data on faculty and 81 of those provided faculty salary information. We expect to increase this response rate in future years as we continue to grow awareness and NDC Study experience. The following is a preliminary summary of some key NDC Study findings. Since this is the first year of the official survey, precluding longitudinal trend analysis, data was primarily used to establish baselines and make some basic comparisons with the Taulbee Survey. Furthermore, small response sizes to some parts of the survey make it difficult to draw hard conclusions from the data provided.
In reading this report, one should consider the following points.
- We use the term "department" to refer to the unit offering the program. We use the term "program" to refer to a course of study leading to a degree in one of the computing disciplines: computer science (CS), computer engineering (CE), information systems (IS), information technology (IT), or software engineering (SE).
- A given department may offer multiple programs.
- Degree production data (master's and bachelor's) refer to the previous academic year (20112012).
- Data for current faculty and new students in all categories refer to the current academic year (201213).
- Total enrollment data (the master's and bachelor's level) are reported for both 201112 and 201213.
2. Bachelor's Degree Production and Enrollments
Ninety institutions completed the bachelor's portion of the survey: 29 public institutions and 61 private. Table B1 shows 88 institutions offering one or more bachelor's degrees in a computing discipline. We also found that 84 gave sufficient information to discern whether or not they offered master's degrees in computing. There were 19 institutions offering master's degrees represented and 65 that only offer bachelor's degrees. The smaller size of the master's/non-master's group should be taken into consideration when data that categorizes by institution type is presented.
The survey results include information on 160 bachelor's degrees offered by the 88 institutions. Table B2 summarizes the number of programs by discipline and the ABET accreditation status of those programs. Not surprisingly, the percentage of engineering programs that are ABET accredited (CE at 91.7% and SE at 41.7%) is much higher than for the remaining disciplines. More programs tend to be ABET accredited in public institutions than private and in those institutions that grant master's degrees.
The anticipated production of bachelor's degrees in 20122013 shows a double-digit increase over the number produced in the 20112012 academic year (Table B3). This increase is predicted to be 13.9% over all disciplines and types of institutions. By comparison, the Taulbee survey reported a 15.7% increase in degree production over the period 20102011 to 20112012 but only anticipates a 6.9% increase between 201112 and 201213. The anticipated increase, based on the 88 institutions represented in the NDC Study, is much more pronounced at public institutions (18.1%) and those that grant master's degrees (18.9%) than at their respective counterparts (8.1% and 10.2%).
Table B4 shows anticipated degree production growth by computing discipline. Computer science degree production is expected to closely align with the overall growth rate in CS (13.8%), while both IT and SE anticipate much more marked increases (50.8% and 23.2% respectively). The number of CE degrees is projected to be 9.3% lower while anticipated degree production in IS programs is flat. The validity of the numbers in these last four types of programs is questionable, however, due to the small number of programs reporting.
Among the 20112012 graduates, 16.2% were female (Table B5). This compares favorably to gender data from Taulbee institutions where 13.3% of graduates were female. IS, IT, and SE programs had significantly higher rates of female graduates (21.8%, 20.5%, and 18.2%), while CS institutions were slightly lower (15.4%) and CE programs were significantly lower (12.0%). The percentage of female graduates was higher at private and non-master's degree granting institutions for CS and IS programs but lower for the other three disciplines.
A comparison of ethnicity data between NDC and Taulbee institutions (Table B6) shows that NDC institutions have a significantly higher percentage of white/U.S. resident graduates (69.3% vs. 62.6%), more black/African American graduates (7.3% vs. 5.3%), and fewer graduates that are Asian/U.S. residents (8.1% vs. 16.7%) or multi-racial/non-Hispanic (0.4% vs. 1.4%).
Enrollment over all types of institutions and programs shows an 11.0% increase between 201112 and 201213 (Table B7). This is comparable to that reported by Taulbee institutions for the period 201011 through 201112 (10.6% for those departments that reported both years) and suggests that bachelor's level computing enrollment has continued its upward trend for another year. Enrollment growth at public and private institutions is comparable (11.1% vs. 10.6%) while that at master's degree-granting institutions is significantly higher than non-master's degree-granting institutions (14.0% vs. 7.0%).
As shown in Table B8, IT showed the highest enrollment growth (23.9%), with CS, CE, and SE being more modest and similar (11.0%, 14.6%, and 15.4% respectively). IS appears to be much more flat (1.6%) and, coupled with the flat degree production in the discipline, raises a question as to why IS does not appear to be experiencing the same type of growth as other computing disciplines at NDC institutions.
When enrollments are broken down by discipline and program type (Table B9), the number of new majors in CS and SE is a large portion of the overall CS and SE enrollment, indicating that potential growth in degree production for these programs in future years may be much higher than that already seen, particularly at public and master's degree-granting institutions. In CS, this situation is similar to that reported at Taulbee institutions.
The average number of majors per program is much lower at NDC institutions than Taulbee institutions over all disciplines as well as in CS programs. Within NDC institutions, the number of majors per program in all disciplines is much higher at public institutions than private for all disciplines. Those institutions granting master's degrees see higher numbers of majors per program for all disciplines except SE, where the averages are comparable.
3. Master's Degree Production and Enrollments
In this year's survey, 22 institutions provided data on 31 master's level programs in computing. Of the 22 institutions, 15 were public and seven private (Tables M1M2). The small number of participating institutions, students and programs should be taken into account when drawing conclusions from the data presented here.
For 20122013, these institutions anticipate an overall 26.2% increase in the production of master's degrees over the previous year (Table M3). Notably, IT programs reported a 56.5% increase and SE programs reported a 31.1% increase, both higher than the 19.4% increase for CS. In contrast, Taulbee institutions reported an anticipated decrease in overall master's degree production for 201213 of 9.8% and a decrease in CS specifically of 10.3%.
Among the 20112012 master's degree graduates, 29.1% were female (Table M4), tracking closely with Taulbee numbers, where 28.7% of the master's degree graduates were female. CS, the discipline with the largest response size, reported 36.2% female graduates, significantly higher than the 22.6% reported by Taulbee CS master's level programs. (One possible explanation for higher numbers of women in NDC programs compared to those in Taulbee programs is that NDC departments have a greater percentage of female faculty members than do the Taulbee departments, as will be discussed in the next section.)
A comparison of ethnicity data between NDC and Taulbee institutions (Table M5) shows that NDC institutions have a higher percentage of Asian U.S. resident graduates (13.4% vs. 8%) and black/African-American resident graduates (7.9% vs. 2.7%). There were fewer non-resident (49.9% vs. 53.8%), white (27.1% vs. 32.2%), and Hispanic (1.1% vs. 2.5%) graduates.
Overall enrollment at NDC master's level programs increased 14.9% from 20112012 to 20122013 (Table M6). The largest increases were in CS (17.2%) and IT (24.8%), while SE saw 7.1% growth and 11.5% in IS. In CE there was a 9.5% decline.
4. Faculty Demographics
Table F1 shows the average size of faculty in the responding departments, broken down by type of faculty. On average, there are 8.1 faculty members (accounting for an average of 6.5 FTE) per department, of whom 5.0 (4.9 FTE) are tenure-track, 0.2 (0.2 FTE) visiting, 0.7 (0.7 FTE) full-time non-tenure-track, and 2.2 (0.8 FTE) part-time/adjunct. Private universities and departments having no master's level programs tend to have a somewhat greater percentage of tenure-track faculty and visiting faculty, but a smaller percentage of full-time non-tenure-track faculty members than do public universities and departments having master's level programs.
Full professors and associate professors account for about the same percentage of the total tenure-track faculty members (a little more than 33% each), while assistant professors make up about 25% of the total faculty (Table F2). These percentages do not vary much between public and private universities, nor between departments having master's level programs and those not having master's level programs.
There are about three times as many men as there are women among tenure-track faculty (Table F3) in the reporting NDC institutions. However, the more junior the faculty rank, the greater the percentage who are women. At each rank, the percentage of tenure-track faculty members who are women exceeds its counterpart among doctoral granting computing departments as reported in the Taulbee Survey. The percentage of women among assistant professors (which comprise most of the newly hired faculty) at the reporting NDC departments exceeds both the percentage of female students graduating from bachelor's programs in these departments and the percentage of female graduates from their master's level programs.
Table F4 shows the breakdown of tenure-track faculty members by ethnicity. White and Asian ethnicities account for more than 80% of the full and associate professor ranks and over 75% of the assistant professor rank, and more than 80% overall. Collectively, the underrepresented minority categories of African-American, American Indian/Alaska Native, Native Hawaiian/Pacific Islander, Hispanic/Latino, and Multiracial account for only 6.7% of the total. However, this is better than the 5.0% reported by the Taulbee Survey for doctoral granting computing departments, though the doctoral granting departments do somewhat better than the NDC departments at the assistant professor level. The percentage of assistant professors from these underrepresented minority categories is somewhat lower than the percentage of graduates from master's level programs at the NDC departments, and much lower than the percentage of graduates from their bachelor's programs.
Table F5 summarizes faculty recruiting at the 83 responding departments. On average, there were 0.48 tenure-track openings per department, or one tenure-track opening for about every two institutions during the 201112 recruiting year (though some institutions had more than one opening). Of these, 83% were filled. The vast majority of the new tenure-track positions were at the assistant professor level, as is to be expected. Close to one third of these new positions (30.3%) were filled by women, but only 3% from underrepresented minority ethnicities (Table F6). It appears that the departments in the NDC Study did better at recruiting women but not as well at recruiting underrepresented minorities than did their counterpart doctoral-granting departments reporting to the Taulbee Survey. However, the small number (33) of faculty recruited by the reporting NDC departments makes it inappropriate to draw strong conclusions.
The NDC survey asked the departments about the degrees required for hiring new faculty members at each rank, and for promotion and tenure decisions. The results are in Table F7. Doctoral degrees are almost universally required for hiring at the associate professor or full professor level. Even at the assistant professor level, over 80% of the 83 responding departments required the doctoral degree; this did not vary much between public and private institutions, nor between departments that offered both undergraduate and master's degrees and those that only offered undergraduate degrees. For full-time non-tenure-track positions, the master's degree is the predominant requirement. The doctorate was also almost universally required for promotion and tenure, though there were instances where a institution had a slightly weaker requirement for promotion to the rank than they had for hiring into that rank. At these institutions, once a person was at the university (and presumably met the degree requirements for hiring), and performed at a level expected for promotion, the candidate's highest degree apparently was not deemed an essential ingredient in evaluating their qualifications for advancement.
The 83 responding departments reported a total of 29 tenure-track faculty departures during the past academic year (Table F8). That represents an attrition rate of approximately 7% of their total tenure-track faculty. The top three reasons for departures were retirement, leaving for a non-academic position, and leaving for another academic position. The Taulbee Survey also had the same three top reasons. However, the NDC faculty members were more likely to leave for a non-academic position, while Taulbee faculty members were more likely to leave for another academic position.
5. Faculty Salaries
Fifty-nine of the responding departments reported individual salaries for their tenure-track and full-time non-tenure-track faculty members. Another 18 departments reported only aggregate salaries for their faculty at the different faculty ranks.
Table F9 shows the median salaries by faculty rank among the faculty in those departments that reported individual salaries. Median salaries at the public universities exceed those at the private universities at each tenure-track rank, but not for full-time non-tenure-track faculty. Median salaries at departments that have master's level programs exceed those at departments that do not have master's level programs for all tenure-track ranks and for full-time non-tenure-track faculty.
Table F10 shows the corresponding information for departments that reported aggregate salaries. The salary entries are the averages of the various median salaries at each rank as reported by these departments, so these are neither true median salaries nor true average salaries. Among these programs, these median averages were higher at private departments than at public departments, the opposite of the data reported above for the departments that provided individual salaries. Median averages were higher for those departments that had master's level programs than they were for departments that did not have master's level programs. This comparison is the same as that for the departments that provided individual salaries. It may be tempting to assert that the tenure-track salaries at those departments that reported aggregate salaries typically were lower than those at departments that reported individual salaries, as almost all of the tenure-track salary entries in Table F9 are less than their corresponding values in Table F8. Such an assertion would be invalid, however, based on the manner in which the aggregate salary entries are computed.
Although it is difficult to perform trend analysis and draw out hard conclusions from this first study, the data gathered generally confirms that positive trends in enrollment and in degree production extends beyond Taulbee institutions to the many institutions responding to the NDC survey. In addition to valuable pipeline data, the results of the NDC Study also give the computing education community a previously unavailable snapshot of the students and faculty at these institutions, which themselves annually produce thousands of graduates in the computing disciplines.
If your program participated in the 20122013 ACM NDC Study, thank you for your help. The 20132014 survey will go out to qualifying programs in the fall of 2013. We would love to hear from you about how the survey can be improved, and look forward to your continued annual participation. If you are at a qualifying program but were not able to participate, or were never contacted, we want to hear from you as well. Please send all comments and queries to Yan Timanovsky, ACM Education Manager at email@example.com.
 NCES 2012, IPEDS. https://surveys.nces.ed.gov/ipeds. Accessed 2013 July 13.
 NSF 2012, NCSES. http://www.nsf.gov/statistics/degrees. Accessed 2013 July 13.
 Zweben, S. and Bizot, B. "2012 Taulbee Survey." Computing Research News, Vol. 25, No. 5, May 2013, 1160. http://cra.org/uploads/documents/resources/crndocs/2012_taulbee_survey.pdf. Accessed 2013 July 16.
Jane Chu Prey
University of Colorado, Boulder, Colorado USA
Program Director, Division of Undergraduate Education
National Science Foundation
4201 Wilson Blvd, Arlington, Virginia 22230 USA
ACM Education Manager, ACM Headquarters
Two Penn Plaza, Suite 701
New York, New York 10121-0701 USA
Jodi L. Tims
Professor and Chair, Department of Mathematics and Computer Science
Baldwin Wallace University, 275 Eastland Road
Berea, Ohio 44138 USA
Professor Emeritus, Computer Science and Engineering
The Ohio State University
2015 Neil Avenue, Columbus, Ohio 43210 USA
© Association for Computing Machinery, Inc. Permission to copy, redistribute, and republish results of this study is granted provided credit is given to ACM and/or the ACM Education Board.
The Digital Library is published by the Association for Computing Machinery. Copyright © 2013 ACM, Inc.Contents available in PDF
View Full Citation and Bibliometrics in the ACM DL.
To comment you must create or log in with your ACM account.