In fall 2019 and winter 2020, ACM collected data about enrollments, degree completions, and faculty demographics and salary in Non-Doctoral-Granting Departments in Computing. Referred to as the ACM NDC Study, the data provides timely information about the state of academic computing in the departments of use to the computing community, academic administrators, and the media. This year's enrollment and degree completions data comes from the National Student Clearinghouse Research Center (NSC), and is quite comprehensive in its coverage of the relevant departments. The NSC data is disaggregated by gender, ethnicity, and class rank to allow further depth of analysis in each of the six computing program areas of computer engineering, computer science, cybersecurity, information systems, information technology, and software engineering.
Since 2011, ACM has conducted an annual study of enrollment, graduation, and faculty in Non-Doctoral-Granting Departments of Computing (NDC). This ACM NDC Study complements the annual Taulbee Survey of doctoral-granting programs performed by the Computing Research Association (CRA) . Together, the two surveys afford the computing community a comprehensive look at the status of key elements of computing programs of study within academia. Of particular interest to the NDC Study is the data about bachelor's programs in each of the areas of computing in which ACM provides curricular guidelines  and in which ABET program accreditation criteria exists . Prior to 2018, there were five such areas: computer science (CS), computer engineering (CE), information systems (IS), information technology (IT), and software engineering (SE). In 2018, ACM approved curricular guidelines in the area of cybersecurity (CY) and ABET accredited its first programs in this area.
ACM modified the manner in which the study was conducted this year. Previously, we surveyed academic departments about the various enrollment and graduation statistics from the most recent academic year, and the faculty statistics from the current academic year. However, the response rate was low in each year, and we were concerned about the representativeness of the data we were gathering.
For this year's report, we continued to gather faculty statistics directly from departments, since that is the only means we have to obtain such data. We had 151 departments responding with data on faculty demographics and 89 responding with data on faculty salaries. Both of these numbers are a bit higher than last year, but of course represent only a fairly small fraction of all of the possible NDC departments.
With support from the ACM Education Board, enrollment and graduation statistics were obtained from the National Student Clearinghouse Research Center (NSC), a non-profit organization to which nearly all academic institutions in the United States provide data annually . The data provided to NSC typically is reported by an institution-level data office rather than an academic department-level office. Data is reported at the individual student level and includes the student's current program of study, using the Classification of Instructional Program (CIP) Code .
We included data from those institutions not classified as either Doctoral Very-High Research (aka R1) or Doctoral High Research (aka R2) . These "non-R1 or R2" institutions do not provide data to the CRA Taulbee Survey; therefore they would have been candidates for the annual ACM NDC Study. There also are several R2 institutions that do not grant doctoral degrees in computing and therefore do not report to the CRA Taulbee Survey. However, the level of granularity of the data obtained from the NSC did not allow us to further break down the R2 institutions into those that grant computing doctoral degrees and those that do not. From past experience, we estimate that about one-third to one-half of the R2 institutions grant doctoral degrees in computing. The CRA Taulbee Survey provides good coverage of the doctoral programs from both R1 and appropriate R2 institutions.
A CIP code was assigned to the six computing areas according to the mapping in Table 1. The selection of CIP codes for CS is consistent with the codes used in a recent study by the Computing Research Association of the CS enrollment surge . Identifying codes for CE and SE are relatively straightforward from the engineering set of CIP codes. For the other areas, we consulted persons who have been involved on behalf of ACM and CSAB in computing curriculum and accreditation activities in order to get recommendations for appropriate codes. The resulting code mapping is similar to that used in an earlier study of the representation of women in academic computing programs .
To illustrate the increase in coverage using data from NSC when compared to data from the former NDC departmental survey, Figure 1 shows the number of participating institutions using each approach for the academic year 2017-18. Enrollment and graduation data from academic year 2017-18 was published previously in the 2018-19 ACM NDC Study in all areas except CY .
In the remainder of this report, we provide the NSC enrollment and graduation data from non-R1 or R2 institutions for both 2017-18 and 2018-19. This allows us to provide the community with vastly improved 2017-18 data over what was reported last year, and also allows us to compare results over a two-year period from this comprehensive set of institutions offering programs in non-doctoral-granting departments of computing. We also report the faculty results obtained from the much smaller set of departments who responded to this year's survey. These faculty data cover the previous, 2019-20, academic year.
Enrollment and Graduation Results
The NDC report historically included analysis of student data (demographics, degrees earned, enrollment) in both the bachelor's and master's programs of survey respondents. As the report shifts to the NSC data source, some changes to NDC reporting are necessary. Most notable are the following:
- As described above, the number of institutions for which data is available is greatly increased, resulting in a more reliable understanding of the state of enrollment and graduation.
- A profile of degree offerings at institutions for which data is included is provided. This profile includes minority serving institutions, a categorization not previously available in NDC. However, enrollment and graduation data for the non-R1 or R2 institutions could not be broken down by institutional control (public vs. private), by highest degree offered (bachelor's vs. master's) or for MSIs.
- The set of institutions that report data to NSC includes private, for-profit institutions, a group not previously included in the NDC survey.
- NSC enrollment and graduation data is for bachelor's programs only.
- Enrollment data supports the inclusion of gender, ethnicity, and class rank statistics not previously available in NDC reports.
- Although there is some data about freshmen, NSC data does not specifically identify first-year majors, eliminating the reporting of this leading indicator of enrollment trends.
The goal of this section is to present a foundation of NSC data for the two-year period spanning academic years 2017-2018 and 2018-2019. Reference to prior NDC reporting is only made when a substantial difference in previously reported trends occurs.
For all institutions reporting data in 2017-2018 and 2018-2019, Table 2 and accompanying Figure 2 summarize program offerings within the six curricular areas of computing broken down by institutional control. Notable trends over all institution types, with the exception of for-profit, private institutions, include computer science as the most common degree offering and the growth of cybersecurity programs.
The number and percentage change of degrees granted in each program area are presented in Table 3. Over all disciplines, there was a 4.7% increase in degree production, with largest increases occurring in software engineering (9.0%), computer science (7.5%), and information systems (5.3%). Computer engineering was the only discipline to experience a decline in degree production (-0.3%).
Tables 4a and 4b present the gender representation in degrees awarded broken out by discipline for 2017-2018 and 2018-2019 respectively. There were increases in the percentage of females aggregated over all program types (+0.1%), in computer science (+0.3%), and in information systems (+0.5%). Information technology remained steady in its representation of females among graduates while decreases were seen in computer engineering (-0.5%), software engineering (-0.6%), and cybersecurity (-0.6%).
Figure 3 compares the representation of women over all NDC programs, in NDC CS programs, and in CS programs reporting to the CRA Taulbee Survey (hereinafter Taulbee CS programs). Previously, NDC consistently reported a higher representation of women in NDC CS programs than in Taublee CS programs. The expanded data set from NSC reveals a contrary situation, however, with Taulbee CS programs outpacing NDC CS programs in female representation over the past two years by 2.9% and 2.7%, respectively. However, the trend is upward in both NDC overall and NDC CS programs.
The breakdown of degrees awarded by ethnicity for 2017-2018 and 2018-2019 is presented in Tables 5a and 5b. In comparison to Taulbee CS programs and consistent with prior NDC surveys, a higher percentage of degrees is awarded to Black/African-Americans, Hispanic/Latino, and White students. The percentages of degrees awarded to U.S. residents that are considered underrepresented (i.e., non-White, non-Asian) are presented in Figure 4. Over all NDC programs, the percentage is higher in 2018-2019 than 2017-2018 (27.0% vs. 25.4%). NDC CS programs have higher percentages than those reported by Taulbee CS programs in both 2017-2018 (+8.5%) and 2018-2019 (+9.3%). Note that, in the IT area, more than 1/3 of the graduates' ethnicity was unreported or unknown. Comparisons may be affected by this unknown data and should be interpreted with this understanding.
Enrollment change from 2017-2018 to 2018-2019 is shown in Table 6. Over all programs, a 2.2% increase in enrollment was seen at NDC institutions. Increases occurred in software engineering (5.1%), cybersecurity (4.9%), computer science (3.2%), and information technology (1.6%). Computer engineering and information systems enrollments experienced declines of 5.4% and 0.6%, respectively.
In NDC programs overall, representation of females is 0.6% higher in 2018-2019 than in 2017-2018 (Tables 7a and 7b). Largest increases were in information technology (+1.8%), computer science (+0.6%), and software engineering (+0.6%). Only cyber-security experienced a decline (-0.5%) in female representation in the two-year period. Taulbee CS programs had higher representation of females among enrolled students than NDC CS programs in both 2017-2018 (+0.7%) and 2018-2019 (+1.9%).
The breakdown of enrollment with respect to ethnicity is presented in Tables 8a and 8b. Over all NDC programs, representation of all ethnic groups outside of Asian and White representation has increased between 2017-2018 and 2018-2019. As shown in Figure 5, NDC CS programs had higher representation of underrepresented minorities among US residents (i.e., non-white, non-Asian) than Taulbee CS programs in both 2017-2018 (+12.0%) and 2018-2019 (+13.9%). As was the case for the ethcnity breakdown of graduates, the enrollment break-down.also has disciplines (IT and CY) with more than 1/3 of unreported or unknown ethnicities. Interpretation of ethnicity data should bear this in mind.
Tables 9a and 9b summarize the breakdown of bachelor's enrollment by computing discipline and class rank for both the 2017-2018 and 2018-2019 academic years. Figure 6 depicts the change in percentage of enrollment over this two-year period at each class rank, broken out by computing discipline. Care must be taken in drawing conclusions based on this data; class rank was unreported for about 50% of the IT students and, in 2018-2019, for more than 1/3 of the CY students.
As noted in the introduction, we conducted our usual survey of non-doctoral-granting departments for the purpose of obtaining data about faculty demographics and salaries. Responses were received from 151 departments, four more than responded last year. Nearly two-thirds of the respondents were from private institutions, and nearly two-thirds were from departments that do not grant master's degrees in computing. The public-private split is similar to that last year, but this year there is a larger percentage of master's granting institutions among the respondents (Table 10).
The average faculty size this year was 13.9 in headcount, and 12.1 FTE, both slightly higher than the respective 13.1 and 11.3 reported last year. Increases were seen in both tenure-track and full-time non-tenure-track faculty, while the average number of part-time/adjunct faculty and visiting faculty remained the same. As was the case last year, departments at public institutions tended to rely more heavily on tenure-track and fulltime non-tenure-track faculty than did departments at private institutions, while private institutions had a greater fraction of part-time/adjunct faculty members. Also similar to last year, non-master's-granting departments had a greater percentage of their faculty as tenure-track and smaller percentages as fulltime non-tenure-track and part-time/adjunct (Table 11). Within each category of institutions, and overall, there is a fairly even distribution of tenure-track faculty members at each of the three faculty ranks (Table 11).
Gender diversity was slightly lower this year than last year, with 24.5% female of those whose gender was reported. Last year's percentage was 26.6. Reduction in percentage versus last year was present at the full professor and assistant professor levels (Table 12). There also was a higher percentage of Asian and lower percentage of Hispanic faculty reported this year compared with last year. Only 5.9% of the entire tenure-track faculty was Black, Hispanic, Native American, Native Hawaiian/Pacific Islander, or two or more races. This is lower than last year's 6.8% (Table 13).
There was less recruiting among this year's respondents although fewer departments reported about recruiting than did so last year. Departments that sought new tenure-track faculty were successful about 3/4 of the time, similar to last year. Though the numbers are very small, there were more new tenure-track faculty members hired at senior ranks this year than last year. Full-time non-tenure-track faculty recruiting was more successful this year than last year, with only one slot going unfilled. As was the case last year, part-time/adjunct faculty members were also recruited with very high success, in fact 100% success this year (Table 14). Among the newly hired tenure-track faculty, a smaller percentage were women this year (20.0% vs 27.3%), while there was a bit more diversity in this year's recruiting class (6.8% aggregate across Black, Hispanic, Native American, Native Hawaiian/Pacific Islander, and two or more races, vs 3.9% last year). However, with such small numbers in these categories, a difference of one can change the year-to-year comparison in a meaningful way (Table 15).
Table 16 contains information about the extent to which different types of responding departments require a specific level of degree in order to hire, tenure, or promote tenure-track faculty. These fractions do not change much from year to year. Except for hiring full-time non-tenure-track faculty, the vast majority of departments require the doctoral degree for any of these types of actions; however, hiring of assistant professors without doctoral degrees appears to be more common at private institutions than at public institutions. This also was the case last year.
More faculty departures were reported among this year's respondents, with forty departments reporting a total of 56 departures, compared with 33 departments reporting a total of 41 departures last year. There were faculty this year who left for non-academic positions, while there were none reported last year. However, retirement and departure for other academic positions continue to be the primary reasons that faculty members leave (Table 17).
Tenure-track faculty salary data is reported in Tables 18 and 19. The former table reports those departments that provided data about individual faculty members. The median values reported in these tables are true medians of the collection of individual faculty from these units. The latter table reports all departments that provided salary information, both those providing individual salaries and those that reported only aggregated averages by faculty rank. For Table 19, the medians are medians of these department averages, so they are not in general true medians, nor true averages. Unfortunately, many fewer departments reported individual salaries than reported aggregated salaries (30 vs 59). This is similar to last year. However, this year we have 89 units in total for whom we have some salary data, versus 81 last year.
Among this year's respondents, median individual salaries were higher at public and at master's-granting departments than, respectively, at private and non-master's-granting. While medians of the aggregated salaries showed the same direction of difference between master's and non-master's-granting respondents, there was a lot of similarity between public and private departments. Last year's salary data showed a similar comparison between master's and non-master's departments, while there was more similarity between public and private departments for individual salaries, and higher salaries at private departments for aggregated salaries.
The data from the National Student Clearinghouse used in this report affords a comprehensive view of enrollment and degree production in bachelor's programs at non-doctoral-granting computing programs. The results, with data from approximately 300,000 students in each of 2017-18 and 2018-19, demonstrate overall growth in both enrollment and degree production, but at a slower rate than has been reported in past years using less comprehensive data. The results also demonstrate clear differences across the different computing disciplines, with computer engineering showing declines over this two year period while computer science, software engineering, and cybersecurity show increases. The data also illustrates the growth in the number of institutions with cybersecurity programs.
Data on gender and ethnicity also illustrate differences across the various computing disciplines, with NDC enrollment data being disaggregated by gender and ethnicity for the first time. The representation of women in the information systems and information technology areas is much higher than it is in areas like computer engineering and software engineering, and representation by underrepresented ethnicities is highest among Blacks in IS but highest among Hispanics in CE.
Data on faculty demographics and salaries revealed no major changes from the previous year. There was some growth in faculty size, in both tenure-track and full-time non-tenure-track faculty.
It is our intent to use the NSC data in future NDC reporting of enrollment and graduates data. As reported in the section on Enrollment and Graduation Results, there is some data that used to be reported but that was not obtained from NSC. But the data we can and do report provides a much more complete picture of non-doctoral-granting programs than we were able to do previously.
1. Accreditation Board for Engineering and Technology. Abet.org/accreditation/accreditation-criteria. Accessed 2020 March 30.
2. Association for Computing Machinery Curriculum Recommendations. Acm.org/education/curricula-recommendations. Accessed 2020 March 30.
3. Carnegie Classifications. Carnegieclassifications.iu.edu. Accessed 2020 March 30.
4. Computing Research Association (2017). Generation CS: CS Undergraduate Enrollments Surge Since 2006. https://cra.org/data/generation-cs/. Accessed 2020 April 24.
5. National Center for Education Statistics (2019). CIP Codes.nces.ed.gov/ipeds/cipcode/browse.aspx?y=56. Accessed 2020 March 30.
6. National Student Clearinghouse. Studentclearinghouse.org. Accessed 2020 March 30.
9. Zweben, S. and Bizot, B. 2018 CRA Taulbee Survey. Computing Research News, 31, 5 (2019), 3–74; http://www.cra.org/resources/taulbee/. Accessed 2020 April 24.
Professor Emeritus, Computer Science and Engineering
The Ohio State University
2015 Neil Avenue, Columbus, Ohio 43210 USA
Director of Computer Science Programs—
San Francisco/Silicon Valley and Professor of the Practice
360 Huntington Ave., Boston, Massachusetts 02115 USA
Education and Professional Development Manager
Association for Computing Machinery
1601 Broadway, 10th fl., New York, NY 10019 USA
©2020 ACM 2153-2184/20/09 $15.00
Permission to make digital or hard copies of part or all of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and full citation on the first page. Copyright for components of this work owned by others than ACM must be honored. Abstracting with credit is permitted. To copy otherwise, to republish, to post on servers, or to redistribute to lists, requires prior specific permission and/or fee.
The Digital Library is published by the Association for Computing Machinery. Copyright © 2020 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.