Across North America, universities and colleges are facing a significant increase in enrollment in both undergraduate computer science (CS) courses and programs. The current enrollment surge has exceeded previous CS booms, and there is a general sense that the current growth in enrollment is substantially different from that of the mid-1980s and late 1990s. For example, since the late 1990s, the U.S. Bureau of Labor data shows that the number of jobs where computing skills are needed is on an upward slope [1], illustrating the increased reliance our society has on computing. We also know that more disciplines are becoming increasingly reliant on large amounts of data, and that handling this data effectively depends on having good computational skills. This makes computer science courses at all levels of greater interest to students from other majors.

The enrollment growth in the mid-1980s is sometimes referred to as the "PC boom" and the enrollment growth in the late 1990s is sometimes referred to as the "dot-com boom." CRA Snowbird Conference attendees suggest that we are currently in "Generation CS", where CS enrollment across the nation is surging due to the pervasiveness of computing today.

In early 2015, the Computing Research Association (CRA) created a committee to investigate several questions related to increasing enrollments. The CRA Enrollment Committee's Institution Subgroup (who are the authors of this article) worked to answer high-level questions such as "How are units1 handling the current growth in computer science?" Specifically, we worked to answer questions that concern computer science units, such as:

  1. Are all units seeing a similar degree of growth?
  2. Does the growth exist at all levels of the curriculum?
  3. Are nonmajors and minors having a significant impact on enrollment?
  4. How is the current growth impacting diversity in our student population?
  5. What are units doing to respond to the growth?

To answer these types of questions, we created a CRA Enrollment Survey. The CRA Enrollment Survey was administered in parallel with CRA's annual Taulbee Survey of doctoral-granting units [3] and ACM's annual NDC Study of non-doctoral granting units in computing [5]. Responses were sought only from units that have a computer science undergraduate degree program. The goal was to measure, assess, and understand enrollment trends and their impact on computer science units, diversity, and more [2].

One section of the CRA Enrollment Survey asked respondents to provide detailed demographic data on students enrolled in four representative CS courses. While annual data on degrees awarded and enrollment in majors is available from other sources, we are unaware of any other data regarding student demographics in specific courses over the last decade. Questions have been added to the CRA Taulbee survey to continue collecting this type of data.

In this article, we provide a portion of what we learned from the CRA Enrollment Survey. Specifically, we first document the phenomenal growth of computer science majors in North America since 2006, at both doctoral2-granting and non-doctoral granting units; furthermore, the data indicates that continued growth is likely. We then consider degree completions in computer science from the Integrated Postsecondary Education Data System (IPEDS) data. This section helps advance our understanding of the data collected in the CRA Enrollment Survey, and provides more information about the current surge in computer science at non-doctoral granting units (where data from the CRA survey is limited). Lastly, we consider the widespread increase of nonmajors taking computer science courses and discuss the data that units reported on the increase in computer science minors. Follow-up articles will appear in ACM Inroads (in the September and December issues) to provide what we learned from the CRA Enrollment Survey regarding diversity and institutional impacts/responses.

We are grateful to the 134 doctoral-granting units and 93 non-doctoral granting units that responded to the CRA Enrollment Survey, which produced a response rate of ∼70% (for doctoral institutions via Taulbee) and ∼13% (for non-doctoral institutions via NDC). The data collected from the CRA Enrollment Survey is extremely rich, and allows us to consider a unit's context (e.g., size, public or private) and resources available when considering the impact from enrollment growth.

The Phenomenal Growth of CS Majors Since 2006

The average number of undergraduate CS majors is larger today than at any previous time, and greatly exceeds the peak enrollment of the dot-com boom period. For example, the average number of CS majors at doctoral-granting academic units has more than tripled since 2006 and more than doubled since 2011 [3].

Academic units are undertaking a range of actions to handle the increased enrollment and the demand on resources. Without question, the demands are putting enormous stress on academic units and their faculty. Institutions will need to respond with actions that recognize the reasons for the increased student interest from both majors and nonmajors, and the role that computing plays in a wide range of disciplines and jobs.

This section provides details on the magnitude of the growth of CS majors since 2006. We provide data on the growth of the number of majors as well as the cumulative growth of majors, compare the cumulative growth of majors with the growth in tenure-track and teaching faculty, and illustrate the enrollment increase for courses at three different levels of the curriculum. Most of this section focuses on doctoral-granting units, for which more abundant data is available. Data on non-doctoral granting units is included when available.

Enrollment Growth in Numbers and Percentages

The current growth period began a decade ago. Figure 1 shows how the average undergraduate enrollment has increased each year during this period3. While this growth is impressive, it is natural to ask just how pervasive enrollment increases have been. To illustrate this, we first show enrollments when partitioning academic units by tenure-track faculty size (as done by the CRA Taulbee Survey). Figure 2 shows the average number of CS majors in "small" and "large" units, where "large" is defined as units having at least 25 tenure-track faculty. In 2015–16, 70 large and 75 small units completed the survey. Both groups experienced significant enrollment increases, with large units having roughly twice as many majors as small units. In percentages, however, the increases are very similar. We also examined the growth at public versus private institutions and found the increase in enrollment similar at both types of institutions.

Reporting means on enrollments could allow a few units to excessively skew the overall growth patterns. Figure 3 shows the cumulative percentage of units experiencing different levels of increase from 2009–2014. Only 18% of the units experienced growth under 50% and only 37% experienced growth under 100%. To express this differently, over 60% of the units more than doubled their enrollment since 2009.

• Teaching Capacity

The increase in the number of tenure-track faculty and teaching faculty in no way matches the growth in the number of undergraduate CS majors, as is illustrated in Figure 4. As a result, faculty are teaching larger classes and more classes are taught by visitors, adjuncts, postdocs, and graduate students. According to the CRA Enrollment Survey results, many units are trying to hire teaching faculty (e.g., professors of practice or lecturers). While the growth in teaching faculty since 2006 is over 50%, the average number of teaching faculty an academic unit had in 2015 was only six. By comparison, the average number of tenure-track faculty in 2015 was 28.

• Enrollment in CS Courses

Course enrollment increases are being experienced in all stages of the curriculum. Increases are not only due to the increase in the number of CS majors, but also due to a significant increase in the number of nonmajors enrolled in CS courses. Nonmajor enrollment is discussed later in this article.

Figure 5 illustrates growth of CS majors in representative courses at the introductory, mid-level, and upper-level, at five-year intervals beginning with 2005. As discussed previously, the CRA Enrollment Survey asked responders to provide detailed demographic data on students enrolled in four representative CS courses: an introductory-level course that is mainly for nonmajors (discussed later in this article), an introductory-level course that is mainly for CS majors, a mid-level course, and an upper-level course. Data was requested on these four representative courses across three different time periods: 2005, 2010, and 2015. In Figure 5, the number of units from which such data was obtained is given in parenthesis next to the course name on the horizontal axis.

• Non-Doctoral Granting Academic Units

There is limited historical data on CS enrollments and CS faculty from non-doctoral granting institutions from any sources including the CRA Enrollment Survey. As a result, we cannot produce analogs to Figures 1,2,3,4 for non-doctoral granting academic units. Data collected by the CRA Enrollment Survey suggests that non-doctoral granting units have also seen significant increases in all course levels, though not of the magnitude seen in the doctoral-graning programs. This is illustrated in Figure 6, which shows enrollment by course level for non-doctoral granting units. Several interesting questions regarding non-doctoral granting units deserve attention in the future. For example, our community needs a better understanding of whether the surge in CS majors at non-doctoral granting units is not keeping pace with the surge in CS majors at doctoral-granting units, whether non-doctoral granting units have had the resources to allow enrollment increases in the recent past, and whether there is less student interest in CS at non-doctoral granting units. We can, however, approximate differential growth in doctoral-granting and non-doctoral granting units using graduation data available from IPEDS. We provide this comparisonon the next page.

• Enrollment by Institution Size and Type

The data obtained from the CRA Enrollment Survey is extremely comprehensive, and allows us to consider the growth in computer science by a unit's context (e.g., size, public or private). Figure 7 provides collected data that shows all types of institutions are seeing significant growth in a representative mid-level computer science course. We note that large public institutions have, on average, doubled their mid-level course enrollment from 2010 to 2015. Other types of institutions have seen either a slightly larger increase or a slightly smaller increase in enrollments for their mid-level course. In five short years, the number of students in a representative mid-level course has, on average, more than doubled.

IPEDS Degree Completion Data

Comprehensive data from the Integrated Postsecondary Education Data System (IPEDS) on degree completions at CS and CIS programs can improve our understanding of the CRA Enrollment Survey data, especially in relation to non-doctoral granting institutions. The data provided in this section is from the IPEDS Data Center on degree completions [4] for two Classification of Instructional Programs (CIP) codes—Computer and Information Sciences, General (11.0101) and Computer Science (11.0701). The data is for all Carnegie-classified, not-for-profit, 4-year public and private institutions in the United States, Guam, Puerto Rico, and the U.S. Virgin Islands. We looked at completions under both 11.0101 and 11.0701 since some institutions with computer science degree programs report completions under CIP code 11.0101, while others report completions under 11.0701. In this section, we use "CS" to represent reported completions under both 11.0101 (CIS, general) and 11.0701 (CS).

• Comparing IPEDS Data to the CRA Enrollment Survey Data

The CRA Enrollment Survey collected data from 134 doctoral-granting units and 93 non-doctoral granting units. To compare the CRA Enrollment Survey data collected with IPEDS degree completions, we consider two groupings of Carnegie-classified institutions: (1) Highest, Higher and Moderate Research Doctoral Institutions (Doctoral in Figure 8) and (2) Large, Medium and Small Masters and Liberal Arts and Diverse Baccalaureate Institutions (Non-Doctoral in Figure 8).

Figure 8 shows remarkable growth in CS completions (94%) at the 313 Carnegie-classified doctoral-granting units from 2009 (which reflects growth in majors that began three years earlier) to 2015. The IPEDS degree completion data is consistent with the substantial growth reported previously in this article. Figure 8 also shows noteworthy growth in IPEDS CS completions (48%) at the 1,185 Carnegie-classified non-doctoral granting units. Since the increase in the number of degree completions at non-doctoral granting units is less than the increase in the number of degree completions at doctoral-granting units, it is not surprising that non-doctoral granting units are less likely than doctoral-granting units to report they are experiencing a significant impact from increasing enrollments [2]. We note, however, that CS degree production at non-doctoral granting institutions increased by almost 50% during the period 2009–2015 (see Figure 8), and Figure 6 suggests that this growth in CS degree production at non-doctoral granting institutions will continue to increase for the near future.

The CRA Enrollment Survey collected data from 134 doctoral-granting units, which is a substantial portion of the 313 Carnegie-classified doctoral-granting units in IPEDS. However, the CRA Enrollment Survey collected data from only 93 non-doctoral granting units, which is a small fraction of the 1,185 non-doctoral granting units in the IPEDS data. The IPEDS data represents a much wider variety of institutions than the CRA Enrollment Survey data. The IPEDS institutions include small, medium, and large master degree programs; liberal arts bachelor degree programs; more specialized bachelor degree programs; and other kinds of non-doctoral institutions.

• Degree Completions at Public vs. Private Institutions

Figure 9 shows the total number of IPEDS CS degree completions at public and private institutions. The growth in total completions from 2009 to 2015 was much larger for public institutions (79%) than private institutions (58%).

The Widespread Increase in Nonmajor Enrollment

In addition to the phenomenal increase in computer science majors discussed previously, there is a large increase in the number of nonmajors taking computing courses. Increases in the number of nonmajors are occurring throughout the curriculum (i.e., at the introductory course level, in mid-level courses, and in upper-level courses). Any analysis that only considers the growth of computer science majors therefore underrepresents the increased demand that units are trying to meet. To fully understand the demand that exists, we need to also consider the large increase of nonmajors taking computing courses.

An overview of the nonmajor growth in computing courses, based on courses surveyed from both doctoral- and non-doctoral granting units, can be found in Figure 10. Between 2005 and 2015, in representative courses primarily intended for majors, the number of nonmajors in computing courses increased at a rate equal to or greater than the increase in majors. For the intro majors course, majors increased by 152% and nonmajors by 177%; for the mid-level course, majors increased by 152% and nonmajors by 251%; and for the upper-level course, majors increased by 165% and nonmajors by 143%.

In the following, we consider these increases separately for doctoral and non-doctoral granting units. Specifically, Figure 11(a) summarizes the mean enrollments of nonmajors in each course category for doctoral-granting units, and Figure 11(b) summarizes the mean enrollments in each course category for non-doctoral granting units.

• Introductory Courses

As discussed previously, data was collected from institutions for two types of introductory courses: an intro-level course mainly for nonmajors and an intro-level course mainly for majors. At doctoral-granting units, mean enrollment by nonmajors in the representative intro-level course for nonmajors had an increase of 55% from 2005 to 2015 (38 respondents). Enrollment by nonmajors in the representative intro-level course for majors had a much larger increase of 184% (47 respondents).

Non-doctoral granting units have also seen growth (from 2005 to 2015) in the number of nonmajors taking both types of introductory courses. The growth, however, is somewhat less dramatic than the growth seen at doctoral-granting units—a 25% increase in the intro-level course for nonmajors (13 respondents) and a 92% increase in the intro-level course for majors (19 respondents). We note, however, that the sample size is small, especially when one considers the large number of non-doctoral granting units that exist; in other words, as mentioned previously, more study of non-doctoral granting units is needed to fully understand the situation at the diverse set of non-doctoral granting units.

• Mid-Upper Level Courses

The growth in mid-level and upper-level courses from 2005 to 2015 due to nonmajors was also phenomenal at doctoral-granting units. Specifically, the number of nonmajors in mid-level courses grew by 265% (45 respondents) and the number of non-majors in upper-level courses grew by 146% (44 respondents).

The growth in the number of nonmajors in mid-level and upper-level courses from 2005 and 2015 at non-doctoral granting units was also quite dramatic. Specifically, the number of nonmajors in mid-level courses grew by 133% (21 respondents) and the number of nonmajors in upper-level courses grew by 102% (22 respondents). It is clear that there is an upward trend in the number of nonmajors in mid-level and upper-level courses at non-doctoral granting units, yet we note that the mean numbers of students in both these courses are extremely small. In other words, this data should be interpreted cautiously. Figure 12, however, does show that most units that responded to the survey have seen growth in both their major and nonmajor enrollment for a representative mid-level course.

• Other Enrollment Observations

An important category of nonmajors is minors. Unfortunately, course enrollment changes due to minors are difficult for units to track. Thus, the CRA Enrollment Survey asked for qualitative impact from minors. Of the doctoral-granting units surveyed, none said the number of minors has decreased in recent years, 22% said the number of minors is unchanged, 50% said the number of minors has increased, and 28% said the number of minors has increased significantly. We compared the reported change in the number of minors to the unit's perception regarding the overall impact on CS enrollment increases. We found that units with a greater increase in minors also reported a greater overall impact. In fact, the impact was rated at the highest level (Having big impact with significant challenges to unit) by 46% of the units that stated the number of minors is unchanged, 76% of the units that stated the number of minors has increased, and 96% of the units that stated the number of minors has significantly increased.

It is important to remember that our data about course enrollment is only a sample (i.e., four representative courses from those units who responded). Furthermore, 45% of the doctoral-granting units stated that they restrict their upper-level courses to only majors and minors and, therefore, the data provided in this section may under-represent the actual demand by nonmajors.

Finally, the number of nonmajors may be slightly inflated, especially in the introductory and mid-level courses. That is, some of the reported nonmajors may later become computer science majors. Nonetheless, it is clear from the data that nonmajors represent a significant aspect of the current surge in CS enrollments. Thus, units must develop strategies for managing the increased demand by both nonmajors and minors within the context of their institutions. These strategies should include increasing the unit's understanding regarding both the motivations and needs of nonmajors for enrolling in computing courses. Some of the enrollment demand is driven by the growth of other types of degrees with significant computational components (e.g., "X+CS" degrees that include course requirements from computer science and another discipline X). Thus, units should work across their institution to develop institutional strategies and support for handling the significant enrollment demand from nonmajors.

Summary

The current surge of CS majors is pervasive. Large and small academic units in public and private institutions have been affected similarly. Doctoral-granting and non-doctoral granting units are affected, though doctoral-granting units to date have experienced larger increases. That is, an analysis of IPEDS CS degree production data shows that the growth in computer science is lower at non-doctoral granting institutions than at doctoral-granting institutions, which is consistent with the limited enrollment data available from the CRA Enrollment Survey.

Increases in the number of nonmajors are occurring in courses at all levels—intro-level, mid-level, and upper-level. It appears that the impact from nonmajors is greater at doctoral-granting units than non-doctoral granting units. However, our data indicates that non-doctoral granting units are also seeing significant increases in enrollments from nonmajors. Students pursuing a minor in computer science (who are counted as nonmajors) are an important category of nonmajors.

While academic units are taking a range of actions to handle the increased enrollment, the percentage increases in tenure-track faculty are about 1/10th the increases in the number of majors. This discrepancy in students versus faculty has impacted the operation of programs [2]. Many units face increased faculty retention problems, are unable to hire teaching faculty into newly created teaching positions, and realize that there are not enough new PhDs to fill open faculty slots in the targeted areas.

The fundamental role that computing plays in society and in preparing students of all majors for a career in a competitive workforce suggests that course demand will remain high. Units need to work within their institution to develop a sustainable model that meets the needs of students (both majors and nonmajors), maintains the quality of instruction, and fulfills their role in educating students for a successful career in the 21st century.

Next Steps

Many members of the computer science community are very concerned about the impact of the current student enrollment surge on diversity, as we learned several hard lessons regarding diversity in previous enrollment booms. While more data is needed, there appears to be some good news regarding both the numbers and percentages of women and underrepresented minority students involved in computer science as majors and as students in CS courses; unfortunately, this good news does not apply for all units that responded to the survey. The September issue of ACM Inroads will include an article that concerns data on diversity from the CRA Enrollment Survey.


Institutions will need to respond with actions that recognize the reasons for the increased student interest from both majors and nonmajors, and the role that computing plays in a wide range of disciplines and jobs.


An article in the December issue of ACM Inroads will consider the impact of the current enrollment surge on the unit (e.g., challenges with space and instructional staff), as well as how units are responding to the current surge (e.g., increasing section sizes or number of sections taught).

Readers who are interested in the survey's methodology should see [2]. We also encourage those interested in more details and analysis about the current enrollment surge in computer science to obtain an upcoming report from the National Academies of Sciences, Engineering, and Medicine's ad hoc Committee on Growth of Computer Science Undergraduate Enrollments. The report is expected to be published later this year.

• Acknowledgements

We acknowledge everyone who has assisted with the survey, data, analysis, or report in [2].

References

1. Bureau of Labor Statistics, Employment Projections; www.bls.gov/emp/ep_table_102.htm. Accessed 2017 March 24.

2. Computing Research Association (2017). Generation CS: Computer Science Undergraduate Enrollments Surge Since 2006; http://cra.org/data/Generation-CS/. Accessed 2017 March 24.

3. Computing Research Association, The Taulbee Survey. http://cra.org/resources/taulbee-survey/. Accessed 2017 March 24.

4. National Center for Education Statistics; https://nces.ed.gov/ipeds/. Accessed 2017 March 24.

5. Tims, J., Zweben, S., Timanovsky, Y., and Prey, J., ACM-NDC Study 2015–2016: Fourth Annual Study of Non-Doctoral-Granting Departments in Computing, ACM Inroads, 7, 3 (2016), 50–63.

Authors

Tracy Camp
Department of Computer Science
Colorado School of Mines
1500 Illinois Street, Golden, CO 80401 USA
tcamp@mines.edu

W. Richards Adrion
College of Information and Computer Sciences
University of Massachusetts Amherst
140 Governors Drive, Amherst, MA 01003 USA
adrion@cs.umass.edu

Betsy Bizot
Computing Research Association
1828 L Street NW, Suite 800, Washington DC 20036 USA
bizot@cra.org

Susan Davidson
Department of Computer and Information Science
University of Pennsylvania
3330 Walnut Street, Philadelphia, PA 19104 USA
susan@cis.upenn.edu

Mary Hall
School of Computing
University of Utah
50 S. Central Campus Drive, Salt Lake City, UT 84112 USA
mhall@cs.utah.edu

Susanne Hambrusch
Department of Computer Science
Purdue University
305 N. University Street, West Lafayette, IN 47907 USA
seh@purdue.edu

Ellen Walker
Computer Science Department
Hiram College
11730 Garfield Road, Hiram, OH 44234 USA
walkerel@hiram.edu

Stuart Zweben
Department of Computer Science and Engineering
The Ohio State University
2015 Neil Ave., Columbus, OH 43210 USA
zweben.1@osu.edu

Footnotes

1. We use the term "academic unit" or "unit" to denote the administrative division responsible for the CS bachelor's program. Often, but not always, this is an academic department.

2. Our report mainly provides data on doctoral-granting units, as more data is available on doctoral-granting units than non-doctoral granting units. We strongly encourage non-doctoral granting units to complete the annual ACM NDC!

3. The years shown in the figures indicate the start of an academic year. For example, 2006 denotes academic year 2006–07 with data typically available in early 2008. Enrollment for 2015–16 is based on preliminary analysis of data from the 2016 Taulbee Survey.

Figures

F1Figure 1. Average number of CS majors per unit since 2006.

F2Figure 2. Average enrollment by CS majors at large and small academic units (based on number of tenure track faculty). The percentages denote cumulative changes since 2006.

F3Figure 3. Cumulative percent of units with the indicated level of growth in CS majors from 2009 to 2014.

F4Figure 4. Cumulative percent growth of CS majors and instructional faculty since 2006.

F5Figure 5. Average enrollment by CS majors in three representative computing courses at doctoral-granting units from 2005 to 2015. The number in parentheses in each category indicates sample size.

F6Figure 6. Average enrollment by CS majors in computing courses at non-doctoral granting units from 2005 to 2015. The number in parentheses in each category indicates sample size.

F7Figure 7. Course enrollment increases in representative mid-level courses from 2010 to 2015, separated by type of institution. Each dot represents one institution's data, and the solid lines represent a linear fit on the data points. The number in parentheses in each category indicates sample size.

F8Figure 8. PEDS data (CIP 11.0101 and 11.0701) on CS bachelor's degree completions (2009–2015) for non-doctoral granting units (blue) and doctoral-granting units (red).

F9Figure 9. IPEDS data (CIP 11.0101 and 11.0701) on CS bachelor's degree completions (2009–2015) for private institutions (red) and public institutions (blue).

F10Figure 10. Cumulative nonmajor enrollment (red) and major enrollment (blue) in computing courses at doctoral- and non-doctoral granting units from 2005 to 2015. The number in parentheses in each category indicates sample size.

F11Figure 11. Average enrollment by nonmajors in computing courses at doctoral- and non-doctoral granting units from 2005 to 2015. The number in parentheses in each category indicates sample size.

F12Figure 12. Course enrollment increases in representative mid-level courses from 2010 to 2015, separated by majors and nonmajors. Each dot represents one institution's data for either majors or nonmajors. Dots above the 2x dotted line represent units that have more than doubled their enrollment.

Copyright held by Authors/Owners.

The Digital Library is published by the Association for Computing Machinery. Copyright © 2017 ACM, Inc.

Contents available in PDF
View Full Citation and Bibliometrics in the ACM DL.

Comments

There are no comments at this time.

 

To comment you must create or log in with your ACM account.