University-based engineering research in the United States
This paper reports results from an NSF-supported survey ofa national probability sample of U.S. engineering faculty on the nature and organization of their research activities. We analyze how engineering faculty spend their time, the types of research faculty conduct, characteristics ofthat research, how that research has changed over time, and the involvement of students in engineering research, based upon the engineering discipline, institutional research intensiveness and age ofthe respondents. These data and results represent a perspective on university-based engineering research activities from the point of view of active researchers across most of the engineering schools in the United States.
During the last 50 years, the United States has experienced rapid growth in its engineering capacity. Concomitant with this growth has been an expansion in the number and types of academic engineering programs. Today, many engineering institutions are reevaluating their academic programs based upon changing circumstances and resource constraints. Included in these considerations is a sense that the nature and characteristics of engineering research may be changing in response to technological advancement, new collaborative opportunities, and the call for more public accountability for the outcomes of research activity. However, as university administrators, academic researchers and policy-makers broach these issues, there has been relatively little attention paid to the views of academic engineering researchers.
In 1993, the Center for Technology Assessment and Policy (CTAP) at Washington University in St. Louis initiated a study on engineering research in U.S. universities, in response to the dearth of information available about the nature of academic engineering research in this country. In particular, we did not know enough about the kinds of research going on in engineering schools, or how this research may have changed over time. We also did not know enough about what role this research plays in broader issues of industrial productivity and competitiveness, or about its role in education. The study was conducted with three primary goals: 1) to provide a clearer picture about the nature of engineering research being performed in U.S. universities today; 2) to analyze and describe the major trends and factors influencing academic engineering research; 3) and to build a better description of how academic engineering research is serving the educational and economic needs of society.
The data presented in this paper focus on the nature and organization of university-based U. S. engineering research with particular attention being given to information on how these factors are affected by engineering discipline, by the research intensiveness of the institution, and by the age of the faculty respondents. Overall results have been published previously from the CTAP engineering faculty survey and from a companion survey of directors of engineering research units.1,2,3 This paper contains previously unpublished data and analyzes. The data presented in this paper were collected in 1993, therefore, some perceptions reported by the respondents may have changed since that time.
II. METHODOLOGY AND KEY VARIABLES
In the Fall of 1993, CTAP administered a mail survey to a national probability sample of engineering faculty, with support from the National Science Foundation. The survey included questions that covered a wide range of topics concerning university-based engineering research: its nature and organization; federal and industrial involvement; the role of students and the extent of their involvement, and changes in research over time. The survey sample was drawn from the 200 engineering institutions that reported the highest annual engineering research expenditures to the American Society for Engineering Education (ASEE).4 The sampling frame was constructed from university catalog listings of engineering faculty at these 200 institutions. Faculty who met the following criteria were included in the sampling frame: 1) full-time, tenured or tenure-track faculty; 2) those whose principal appointments were in engineering, and 3) those who were currently or had been engaged in university-based engineering research during their careers. A probability sample of 3,534 faculty, stratified by institutional research intensiveness (as measured by annual research expenditures) and governance (public or private) was selected to receive the survey. Some 2,829 met the criteria for inclusion in the survey sample.
Of these faculty, some 1,727 returned usable questionnaires, yielding a response rate of 61.0%. Key variables to be examined in this paper include the following: engineering discipline, institutional research intensiveness and faculty age*.
University-based engineering activity has traditionally been organized by discipline. In order to examine how engineering research is affected by discipline, responses to the faculty survey were analyzed by the principal disciplinary affiliation of the respondents. Table 1 shows a breakdown of the respondent population by disci– pline – this division will be used to describe selected results throughout this paper.
B. Research Intensiveness
The engineering schools represented in the survey vary in size and mission, from those offering at least one master’s program and as little as $358,000 in annual engineering research expenditures, to those with many master’s and doctoral programs and tens of millions of dollars of annual engineering research expenditures.7 The nature and organization of engineering research activities at these institutions may vary, based on the size and scope of the research effort. We define research intensiveness as a measure of the magnitude or extent of research activities. A proxy for this measure, namely, the institution’s annual engineering research expenditures, was used to stratify the survey sample.8 Using the research intensiveness variable, we could then, for example, examine how large, research-intensive engineering institutions might differ from small ones.
The 200 institutions used to draw the survey sample were grouped into four strata based on their relative rank as determined by annual engineering research expenditures. The strata were defined as follows: the top 20 institutions with the highest research expenditures are referred to as Stratum One; followed by the next 30 (Stratum Two); the next 50 (Stratum Three); and the next 100 (Stratum Four). Stratum One faculty comprise 34.0% of the respondent population, and their research accounts for over half of all engineering research expenditures; 20% of the faculty are from Stratum Two institutions; 23% are from Stratum Three; and 22% are from Stratum Four, which accounts for less than 10% of all engineering research expenditures. Stratum One was overrepresented in the overall sample to reflect the greater research activity in Stratum One schools*.
A third factor of interest is faculty age. It may very well be that older faculty perform research that differs from that of their younger colleagues. In order to determine whether the nature and organization of research may vary with the age of the faculty, the survey results were divided into three groups, based on the age of the respondents. Table 2 shows the distribution by age that was used for purposes of analysis.
III. FINDINGS A. Faculty Characteristics
In table 3, some characteristics of the engineering faculty are shown by disciplinary group, as well as for the overall population of faculty. The EE/CSci faculty (Electrical and Electronics Engineering, Computer Science and Computer Engineering)’ are, on average, the youngest group of faculty and they have the least years of research experience as faculty members. Over their careers, the Chem/Bio/MatSci faculty have spent the greatest percentage of time on research activities and they most often report having been a consultant to government or industry.
Demographically, the EE/CSci group is made of up a greater percentage of minorities than the other disciplinary groups. The Chem/Bio/MatSci faculty report the smallest percentage of minorities, but the highest percentage of female faculty. The ME/Aero group reports the smallest percentage of female faculty.
B. Nature of Academic Engineering Research
During the 1992-93 academic year, engineering faculty report spending their time on the following activities: research – 40.0%; teaching – 39.7%; administration – 15.7%; other activities (not including outside consulting) – 4.5%. Examining these results by discipline, CE/Env faculty spend less time on research than their counterparts in other disciplines (37% vs. 40-44% for the other groups). The Chem/Bio/MatSci and Other faculty spend slightly less time on teaching than the remaining faculty (36-37% vs. 4042% for the other groups).
Faculty at the 100 least research-intensive institutions, Stratum Four, spend significantly less time on research than their counterparts at institutions of greater research intensiveness (32% vs. 4144% in the other strata). Some 95.4% of engineering faculty were teaching courses at the time of the survey; these faculty report teaching an average of 3.2 courses. Faculty in Stratum Four tend to teach more courses per academic year (-4 courses) than those in the other strata (~3 courses). Engineering faculty in all the strata report that they would like to spend somewhat more time on research, slightly more time on teaching, and less time on administrative and other duties.
For the 1992-93 academic year, younger faculty (
Overall, faculty described the distribution of their research activities as 32.6% basic research, 49.6% applied research, and 17.8% development. When discipline is considered, the Chem/Bio/MatSci and CE/Env faculty were found to differ from their counterparts in other disciplines in their responses (see table 4). CE/Env faculty report conducting the smallest percentage of basic research, but the greatest percentage of applied research. The opposite is true for Chem/Bio/MatSci faculty. The Chem/Bio/Mat Sci faculty conduct the least percentage of research (R&D) that is classified as development.
The types of research faculty conduct also varies with research intensiveness (see table 5). Faculty in the more research-intensive strata report that, when compared to faculty in the less research-intensive strata, a greater percentage of the research they conduct can be classifled as basic research and a smaller percentage as development.
To aid in the classification of categories of the nature of engineering research, a series of bipolar dimensions were developed. Faculty were asked to place their research activities on a continuum for each of these dimensions. Figure 1 shows the mean responses, by discipline. The disciplines differ statistically on most of the continua; however, there are only a few cases in which the disciplines differ substantively. For example, the Chem/Bio/MatSci faculty report that their research tends to be more “big leap forward,” while the other disciplines tend to favor “incremental gains.” The CE/Env faculty are driven more by funding agency interests than faculty in the other disciplines who are driven by principal investigator interests to a greater extent. Finally, the “Other” group of faculty, which includes a large percentage of industrial, manufacturing and operations research faculty, reports that their research is less publication-output driven than the remaining four groups.
Since the time that the respondents were first involved as faculty members in university-based engineering research, they report that there have been some changes in their research activities. University-based engineering research is described as being less theoretical and more practical and applications-oriented today, as well as contributing more to industrial needs than it was at the beginning of the respondents’ faculty careers (see table 6). There is more crossdisciplinary focus and team-orientation today. Students are more involved now and research activities contribute more to students’ education. Some other qualities have also become more a part of academic engineering research: faculty report that research has become more short-term, that there is more difficulty in obtaining funding, and that the time required to get a paper published has increased. The only attribute that has decreased over time is the theoretical nature of the research. There were no statistically significant differences between faculty members of different disciplines on these changes, with two exceptions: The EE/CSci faculty reported a greater extent of increase in applications orientation to their research than did the other groups, and both the EE/CSci and CE/Env reported a lesser extent of increase in the time required to get a paper published.
Table 7 indicates the mean rating for all faculty, and by discipline, for the extent of involvement of faculty in some categories of research activity. The ratings range from 1 to 5, where 1 represents “not at all,” 3 represents “to some extent” and 5 represents “to a very great extent.” Design and simulation activity, adding to theoretical knowledge and experimentally verifying theory are ranked at or near the top for nearly all the disciplines. The Chem/Bio/MatSci faculty included developing empirical information of use in generating new technologies as third in their ranking; the “Other” group of faculty ranked developing applications for specific industries and technologies as second, probably due to the high percentage of involvement of industrial and manufacturing engineers in this group.
C. Organization of Academic Engineering Research Almost half of the engineering faculty (45.2%) are involved in organized, university-based engineering research units. This value varies by institutional research intensiveness: Stratum One – 52%; Stratum Two – 48%; Stratum Three – 45%; Stratum Four – 31%. The lower percentage of faculty in the less research intensive institutions who participate in research units is likely due to the smaller number of units that exist at those institutions.
When asked to identify organizational categories into which their research fell, 40% of all faculty engaged in research activities report working as individual principal investigators with zero to two students or staff; 56% work as individual principal investigators with more than two students or staff; 38% work a, collaborating investigators within a small research group (6 or fewer people); and 20% as collaborating investigators within a large research group (7 or more people). Of these classifications, the category that faculty most often identify as their principal research group type is as an individual principal investigator with more than two persons. The tendency of faculty in the less research intensive institutions to work in smaller groups is evident: Stratum Four faculty are less likely to report working as collaborating investigators within large research groups (6% vs. 17-30% in the other strata), and more likely to report working as individual principal investigators with zero to two students or staff (53% vs. 32-43% in the other strata).
The organizational nexus in which faculty participate also varies by age. The oldest group of faculty is least often involved in a research unit (Younger – 48.4%; Middle – 47.5%:, Older – 36.9%), and they more often work in smaller groups as individual principal investigators than do younger faculty (see Table 8)
D. Student Involvement in Research Activities
There is significant indication that students are integrally involved in the research activities of faculty. On a scale of 1 to 5, where 1 represents “no involvement at all,” 3 represents “marginal involvement” and 5 represents “great extent of involvement,” faculty rate graduate students at 4.71 and undergraduate students at 3.11. Undergraduate students are involved in the research activities of faculty to approximately the same extent in all the research intensiveness strata; Stratum Four faculty, however, report somewhat less involvement of graduate students (Stratum One – 4.89; Stratum Two – 4.78; Stratum Three – 4.71; Stratum Four – 4.34). Because the institutions represented in Stratum Four have fewer graduate programs, this result may be a reflection of fewer numbers of students available to participate in faculty research activities* and/or of the heavier faculty involvement in teaching versus research. Among the disciplinary categories, Chem/Bio/MatSci faculty reported the greatest extent of both graduate and undergraduate student involvement. Compared to both groups of their younger colleagues, older faculty (56 or more years) tend to have a slightly lesser extent of graduate student involvement in their research (Younger 4.77; Middle – 4.74; Older- 4.58).
When asked to report the percentage of students they were supervising in 1992-93 who were working on research that constituted their degree research requirements, all faculty responded that this value was, on average, 90.3%. Chem/Bio/MatSci and younger faculty reported the greatest percentage of students working on such research (by discipline: EE/CSci – 90.4%; ME/Aero 92.8%; CE/Env- 86.5%; Chem/Bio/MatSci – 95.3%; Other Engineering – 86.9%), (by age: Younger – 93.0%; Middle – 90.6%; Older- 85.1%).
The most frequent roles graduate students play in research activities are as associate researchers and independent researchers; undergraduate students are most often assistants to others in research and general “go-fors.” Stratum Four faculty more often reported using their graduate and undergraduate students as “go-fors” (see table 9). Both Stratum Three and Four faculty more often reported that their graduate students were used as technicians. In general, as research intensiveness increases, graduate students more often play the roles of associate researchers and independent researchers and less often as assistants to others in research. Examining the results by discipline, it was found that graduate students are most often used as independent researchers and least often as technicians by Chem/Bio/MatSci faculty.
On average, each engineering faculty member engaged in research supervises 2.8 doctoral students and 2.4 master’s students. Faculty in Stratum Four tend to supervise more master’s students (3.0 vs. 2.3 in the other strata), and fewer doctoral students (1.2 vs. 2.4-3.8 in the other strata) than faculty in the more research-intensive strata. Again, this is not a surprising result, given the smaller number of doctoral programs at these institutions. Chem/Bio/MatSci faculty supervise the most doctoral students (EE/CSci – 3.2; ME/Aero – 2.5; CE/Env – 2.3; Chem/Bio/MatSci – 3.6; Other Engineering – 2.6), but the least master’s students (EE/CSci – 2.4; ME/Aero – 2.4; CE/Env – 3.1; Chem/Bio/MatSci – 1.6; Other Engineering – 2.5). CE/Env faculty report the opposite result with the least doctoral students and the most master’s students. Older faculty supervise fewer doctoral students than their younger colleagues (Younger – 2.9 doctoral students; Middle – 3.0; Older – 2.3).
Faculty report that the most important contributions to graduate students who are involved in their research activities are knowledge in a specific area of interest, fundamental research skills and problem-solving skills (see table 10). Interaction with industrial and governmental researchers was mentioned as occurring to a much smaller extent. Faculty in the two most research-intensive strata report that acquisition of fundamental research skills is their most highly ranked contribution to graduate education; Stratum Three and Four faculty gave acquisition of knowledge in a specific area of interest the highest rank.
Examination of these items by discipline shows several differences. Students working with Chem/Bio/MatSci faculty are reported to receive the greatest extent of acquisition of fundamental research skills and the most cross-disciplinary research experience. CE/Env faculty supervise students who are reported to receive the least extent of interaction with industry researchers. Students supervised by ME/Aero faculty are reported to receive the greatest extent of interaction with government researchers. Finally, `Other Engineering’ faculty report the greatest extent of understanding of real-world industrial problems obtained by their students compared to students working with faculty in the remaining disciplinary groups.
Faculty members were asked to rate the prospects for their doctoral students to obtain research-oriented positions in a variety of sectors. On a continuum where 1 = excellent, 2 = good, 3 = fair, and 4= poor, the faculty rated the mean prospects in all job sectors between “good” and “fair.” Faculty report that the best prospects for their doctoral students in research-oriented positions are in universities in non-faculty research positions, and in big companies (Universities, non-faculty research position – 2.20; Big companies 2.27; Government all levels – 2.48; Universities, faculty research position- 2.52; Small companies – 2.60; Consulting Firms – 2.67). When the results were examined by discipline, there were significant differences in all the categories, as shown in figure 2. With one exception, EE/CSci, all of the disciplines reported that the best prospects for their students were in universities in non-faculty research positions. The EE/CSci faculty said the best prospects for their doctoral students were in big companies. It should be kept in mind that these results were obtained in 1993, and the faculty perception of doctoral job prospects may have changed since that time.
E. Changes in Research over Time
From a list of attributes of research, faculty reported that the biggest change in academic engineering research since the time they first became involved in university-based engineering research as a faculty member has been that they must now devote more time to seeking and sustaining research funding (see figure 3). At the same time, there is less freedom to pursue new ideas. However, they also reported that there is a slight increase in the relevance of academic engineering research to the needs of society; its quality has increased and it is more innovative. Faculty also reported that consulting has become a slightly more important activity and that the need for non-research oriented advanced graduate programs to complement research-oriented doctoral programs in engineering has increased to a small degree. When we examined these results by discipline, research intensiveness, and faculty age, we found very few statistically significant differences.
A. Team versus Individual Orientation in Research Our survey results indicate that almost half of the engineering faculty population is involved to some extent in organized, university-based research units. Many of the faculty also report being involved in larger, collaborative research teams in addition to or instead of as principal investigators who head small research teams. However, when asked whether their research was team or individual, faculty tended toward the individual side. These results give some indication that the team-orientation of engineering research is growing, but that the individual principal investigator model of research is still very important. This trend is highlighted in the more team-oriented organizations that younger faculty tend to participate in when compared to older faculty, who may be more accustomed to a research group centered on an individual. The more research-intensive institutions appear to have more widely adopted the concept of working in teams or in research units: the faculty at these institutions more often report being affiliated with units, and they more often work as collaborating investigators within larger research groups than do faculty at less research-intensive institutions. There were no substantive differences among faculty in the different disciplines regarding the extent of team-orientation; all the disciplines show an increase. As faculty continues to build new types of research relationships (by working within units or across disciplines, for example), some balance will have to be achieved between the individual and the team. Just what the best balance is may not be easy to determine.
B.Applied Orientation of Engineering Research Although some differences exist among disciplines and among institutions of varying research intensiveness, the majority of the engineering research conducted by engineering faculty in all disciplines and institutions is reported to be either applied research or development. Furthermore, overall, faculty reports an increase of orientation towards applications over time, and that their research now contributes more to industrial needs, pointing to a rise in the importance of applied research. Research has also become more short-term. The nature of the data collected does not allow determination of when these changes may have started, but there are indications that these trends will continue. As university researchers encounter stiffer competition for limited governmental research funds, they may try to sustain their research programs by developing closer ties to industry, which tends to support more applied research.9
C. Research and Teaching
On average, engineering faculty spend approximately the same amount of their time on teaching and research. However, faculty at the 100 least research-intensive institutions (Stratum Four) spend significantly more time on teaching than on research, teaching more courses per faculty member than do their counterparts at the more research-intensive institutions. Younger faculty tend to spend a greater percentage of their time conducting research, and a smaller percentage of their time teaching than older faculty. However, younger faculty are more likely than older faculty to be teaching at least one course. This would suggest that although younger engineering faculty in general spend a greater proportion of their time on research activities versus teaching activities, at least some teaching duties are more often a part of a younger faculty member’s professional life when compared to older faculty.
D.Student Involvement and the Focus of Graduate Education
Students are heavily involved in university-based engineering research activities. The roles they play are important, and particularly in the case of graduate students, are quite substantial. The extent of involvement is affected by the educational level of the student (more involvement of graduate than undergraduate students); by institutional research intensiveness (more involvement of students at more research-intensive institutions); by the age of the faculty supervisor (greater extent of student involvement with younger faculty); and there are variations among the disciplines. The Chem/Bio/MatSi faculty, who spend the greatest percentage of their time on research, also report the greatest extent of involvement of students; the CE/Env faculty, who spend the least percentage of their time on research, report the least amount of student involvement.
Our survey results indicate that the education of graduate students is still focused on the acquisition of knowledge in a particular area, fundamental research skills, and problem-solving skills. Opportunities to gain real-world experience or to interact with nonacademic researchers are less often provided to students. The engineering faculty in our sample were certainly modest about the prospects for their doctoral students in 1993; job prospects in all sectors were reported to only be between “fair” and “good.” Additionally, the faculty report that the need for non-research oriented advanced graduate programs to complement research-oriented doctoral programs in engineering has increased sLightly, although a relatively small percentage (29%) share this view.
E. Attitude and Prospects
Although engineering faculty report that there are some less than desirable aspects about changes that have occurred in engineering research over their careers, such as an increased difficulty in obtaining funding, greater time required to get a paper published, and less freedom to pursue research they really want to do, they are still very positive about their research participation. The less desirable changes are tempered by the opinion that the relevance, quality, and innovativeness of engineering research has increased, and they want more involvement of industry. As changes continue to occur in the form and substance of engineering research, further data collection activities, such as the survey we report here, if performed periodically could serve to ensure that the views of university-based engineering researchers receive adequate consideration.lo
*The latest expenditure data available at the time from ASEE was for 1991. The sampling and stratification procedure used for the survey reflect institutional engineering research expenditures at that time.
*Another variable that we used to stratify the sample was university governance, i.e. public or private. Our results indicate that very few distinctions can be made between research conducted by faculty at public versus private institutions. The differences that do exist are modest; some may be attributed to differences in the sources of support that researchers at these institutions rely on. For further information, see the report by Kannankutty”
“The following formula was used to calculate the number of desired respondents from public and private institutions in each research-intensivencess stratum:
P = [(NiRi)o.(NiRo
where N is the number of faculty in each stratum governance cell (i.e. Stratum One – public institutions, Stratum One – private institutions, Stratum Two – public institutions, Stratum Two – private institutions, etc.), R is the research expenditure value associated with that cell and P is a proportionality factor, which when multiplied by the desired number of survey respondents, gives the desired number of respondents in each stratum governance cell (Fi). Using values of Fi, the expected response rate and the estimated percentage of faculty engaged in research, the number of participants to be included in the sample could then be calculated
*See table 1 for a complete explanation of the disciplinary abbreviations.
*Of the 100 institutions in Stratum Four, only 68 have at least one doctoral program in engineering all of them have at least one master’s program in engineering.
REFERENCES AND NOTES
1. Morgan, RP., D.E. Strickland, N. Kannankutty, and J. Grillon, “Research on Academic Engineering Research: Part II,” ASEE Prism (November 1994), vol. 4., no. 3, pp. 30-35.
2. Morgan, R.P. and D.E. Strickland, “Research on Academic Engineering Research,”ASEE Pri.sm, (April 1994), vol. 3., no. 8, pp. 21-26.
3. Morgan, RP., N. Kannankutty, and D.E. Strickland, “Future Directions for University-Based Engineering Research,”ASEE Prism (March 1997), vol. 6, no. 7, pp. 30-36.
4. The primary source of data on research expenditures was the American Society for Engineering Education, Engineering Education, vol. 81, no. 2, 1991.
5. Kannankutty, Nirmala, Academic Engineering Research in the United States: Characteristics and Trends, Center for Technology Assessment and Policy, Report 96-1, Washington University, St. Louis, October, 1996.
7. Kannankutty, Nirmala, Engineering Research in US. Higher Education: Survey Sample Development and Implementation, Center for Technology Assessment and Policy, Report 95-1, Washington University, St. Louis, March, 1995.
9. Strickland, D.E., N. Kannankutty, and R.P. Morgan, “Forging Links Between Engineering Education and Industry: The Research Connection,” Proceedings, 1996ASEE Annual Conference, ASEE, 1996.
10. The research reported in this article was supported in part by the National Science Foundation under Grant No. SED-9154555 to Washington University. Any opinions, findings, conclusions or recommendations expressed in this publication are those of the authors and do not necessarily reflect the views of the National Science Foundation.
NIRMALA KANNANKUTTY Division of Science Resources Studies National Science Foundation
ROBERT P. MORGAN Department of Engineering and Policy Washington University
DONALD E. STRICKLAND Department of Management Southern Illinois University at Edwardsville
Copyright American Society for Engineering Education Apr 1999
Provided by ProQuest Information and Learning Company. All rights Reserved