Funding, Status, and Usage of Academic Research Instruments
Data Considerations for this Chapter
This chapter discusses the status of approximately 61,500 movable science and engineering research instruments with an original cost of at least $20,000 that were in research use in S&E departments and facilities at the 318 universities and medical schools represented in the Instrumentation Survey. The aggregate cost of this stock was $5.1 billion. All discussion of research instruments in Chapter III refers to these instruments, which are depicted in detail in Table 6. This information was collected via the survey's Instrument Data Sheets and does not include data for any of the approximately 150 supersystems, aggregating $1.2 billion in cost, for which this evaluative information was not collected.
Where to find the data: Table 6 presents the total aggregate cost of all the instruments discussed in Chapter III, by detailed type of instrument and by field of science and engineering.
Funding for the Acquisition and Upkeep of the Research Instrumentation
Source of Funds for the Acquisition of Academic Research Instruments
Source of the Data
The Principal Investigator responsible for each sampled instrument was asked to estimate the "source(s) of funds for acquisition of this equipment, including dedicated accessories" and was given the following sources: Federal (National Science Foundation; National Institutes of Health; Department of Defense; Department of Energy; and other Federal sources) and non-Federal (institution or department funds; state grant or appropriation; industry; and other sources including private, nonprofit foundations, gifts/donations, and bonds).
Where to find the data: Table 7 presents the purchase price of the aggregate stock of instrumentation in 1988-89 and 1993 as well as the source of funding for that stock.
The Federal Government accounted for approximately 50 percent of the funding for the aggregate stock of instruments in research usage in both 1988-89 and 1993. The National Science Foundation was the largest Federal source of the funds that had been used to acquire the aggregate stock of research instrumentation in both survey years. The NSF had contributed the funds to purchase approximately 16 percent of the aggregate cost of the stock of equipment in research usage in 1993.
The Department of Defense (DOD) contributed approximately 10 percent of the aggregate cost of instruments extant in both years. The aggregate contribution, however, increased from approximately $300 million in 1988-89 to approximately $600 million in 1993; this made DOD the second largest Federal source of funds for academic instrumentation in research use in 1993. The National Institutes of Health (NIH ) had been the second largest Federal source of funds for research instruments in usage in 1988-89 (approximately $400 million). Although the aggregate contribution increased to approximately $500 million in 1993, proportionately NIH dropped to third place among Federal sources.
The Department of Energy was the fourth largest Federal contributor to the cost of the aggregate stock of research instrumentation in both survey years; in 1993, it was also the source of approximately $400 million in funds. The share of the total cost of the aggregate stock of instruments contributed by all other Federal agencies was approximately $200 million in 1993, or about 4 percent of the total aggregate cost of the instruments.
The contribution from all non-Federal sources also totaled about 50 percent in both survey years. Institutional funds were the largest single non-Federal source of academic research instrumentation, accounting for 29 percent of the total aggregate cost in 1993, approximately the same proportion as in 1988-89. Grants and appropriations from state governments totaled approximately 8 percent of total aggregate cost in 1993. Industry as a source of funds accounted for approximately 5 percent of the total aggregate cost of instruments in research usage in 1993. Other non-Federal sources accounted for 7 percent of total funds in 1993. Other sources include foundations, gifts/donations, and bonds. (Bonds, of course, are a "source" of funds only in a limited time sense: in future years, the institutions must pay back the borrowed money and accompanying interest payments from other sources of income.)
1993 Expenditures for the Maintenance/Repair of Academic Research Instruments
Source of the Data
Respondents were asked to "estimate the expenditures for maintenance/repair (NOT for operation) of this equipment and its accessories in FY 1993. (For multiyear service contracts, warranties, etc., prorate to indicate cost of coverage in FY 1993.)"
Where to find the data: Table 8 presents the total expenditures and the mean and median per-unit expenditures for maintenance/repair on the stock of instrumentation in research usage in 1993, by instrument type.
The total expenditures for maintenance/repair (M/R) of the existing stock of instruments was approximately $200 million in 1993. Taken as a percent of the aggregate purchase price of the instruments, these expenditures amounted to 4.1 percent of the aggregate cost. The median M/R cost was $1,200 per instrument.
In terms of total expenditure, molecular/electron/ion beam systems were the most costly to maintain, both in total M/R expenditures ($40 million) and in mean expenditure (an average cost of $108,500 per instrument in 1993). However, those average costs were heavily weighted by very expensive outlays not reflective of all systems; the median cost for M/R for these systems was $6,000.
Adequacy, Research Status, Age, and Working Condition of Academic Research Instruments
Perceived Adequacy of Maintenance/Repair of Instruments
Source of the Data
Respondents were given a five-point scale ranging from 1 (excellent) to 5 (inadequate) and were asked to rate "The adequacy of the maintenance/repair this equipment received in FY 1993."
Where to find the data: For a depiction of the adequacy of maintenance/repair by type of instrument, see Figure 5; for a depiction by field of science and engineering, see Figure 6.
By type of instrument, respondents for bioanalytical instruments reported the highest satisfaction with the adequacy of maintenance/repair: 56 percent were rated as receiving excellent M/R, and only 3 percent received care that was inadequate or poor (see Figure 5).
Overall, the respondents reported that 38 percent of the instruments received excellent maintenance/repair in 1993. Only 8 percent of instruments received maintenance/repair that was less than adequate. Instruments inventoried to chemistry received the lowest percentage of excellent ratings for the quality of the M/R they received in 1993 (22 percent), and computer science units reported the next lowest satisfaction rate (only 28 percent of the respondents rated the M/R as excellent).
On the other end of the satisfaction ratings, 13 percent of physics/astronomy instruments received inadequate to poor M/R during 1993, and 10 percent of the instruments in agriculture and in engineering were judged to have received inadequate or poor care.
Interestingly, instruments inventoried to chemistry, which received the lowest percentage of excellent ratings, also received the lowest percentage of inadequate or poor ratings for the quality of M/R received during 1993 (6 percent) (see Figure 6).
Respondents were also given the opportunity to report that no servicing at all was needed on their instruments during 1993. Less than 1 percent of respondents in every discipline reported that the surveyed research instruments needed no servicing during 1993.
Rated Research Status of Instruments
Source of the Data
Adequate instrumentation is indispensable for performing excellent research. With the rapid improvements in instrumentation, many state-of-the-art instruments quickly lose their cutting-edge status, although they often continue to provide the capabilities necessary to satisfy current research users' needs. In an attempt to discern the extent of state-of-the-art status among instruments in the national inventory as well as the ability of that stock of instruments to satisfy researchers' needs, respondents were asked to rate the technical quality of each instrument according to three research criteria: (1) State-of-the-art: the most highly developed and scientifically sophisticated equipment of its kind; (2) Not state-of-the-art, but adequate to meet the needs of researchers in this department/facility; and (3) Not state-of-the-art; inadequate to meet the needs of researchers in this department/facility.
Where to find the data: Table 9 presents the three rating categories for all instruments, with an emphasis on the primary field of usage of these instruments.
Overall, 63 percent of the instruments were rated as not state-of-the-art but adequate for the researchers' usage. This was the modal response for all types of instruments: between 60 and 70 percent of instruments in each major category were deemed adequate to meet current researchers' needs. (see Table 9).
Twenty-seven percent of the instruments were given a state-of-the-art rating. The instruments receiving the highest proportion of state-of-the-art ratings were the "other" instruments, in which 34 percent were judged to be state-of-the-art; this category includes major prototype systems, robots, and lasers. Microscopy instruments followed closely, with 31 percent rated state-of-the-art.
Although only 9 percent of instruments overall were rated as inadequate for the researchers' needs, 23 percent of computers were given this low rating. Computers also had the lowest percentage deemed state-of-the-art, with only 14 percent given the high rating. Not surprisingly, then, when analyzed by field of primary usage, respondents in the computer science field reported the highest percentage of instruments rated as inadequate, with 37 percent of all instruments receiving this rating. Respondents from computer facilities, in fact, reported that 62 percent of the instruments were inadequate for researchers' usage.
Chemistry had the next highest percentage of instruments deemed inadequate, at 18 percent. Sixteen percent of instruments in agriculture were judged to be inadequate, as were 11 percent of instruments in physics/astronomy. For all other major fields of science and engineering, less than 10 percent of the instruments were rated inadequate for researchers.
Average Age of Instruments
Where to find the data: Table 10 presents the percentage of in-scope instruments falling into each of five age distributions, by detailed type of instrument. Table A-5 presents the same age distributions for all instruments combined, by field of science and engineering and by cost range of the instrument.
Average Age by Type of Instrument
Overall, four of every 10 research instrument systems in use in 1993 had been acquired within the previous four years. At the other end of the age spectrum, almost one quarter (23 percent) of the instruments in active research use in 1993 were over eight years old. The average age of all research instruments was 5.8 years.
Seventeen percent of all instruments were less than two years old in 1993. This was a smaller percentage than those between two and four years old or four to six years old (23 and 21 percent, respectively), suggesting a decrease in the rate of acquisition of instruments costing at least $20,000 in recent years.
Classified by cost of instrument, 17 percent of all instruments costing less than $1 million were less than two years old in 1993, but only 7 percent of instruments over $1 million were that new (see Table A-5). Thirty-six percent of those high-cost instruments were purchased between 1987 and 1989, making their modal age between four and six years in 1993.
By broad type of instrument, computers were the newest, with an average age of 4.0 years. Fifty-five percent of computers costing between $20,000 and $50,000 were less than four years old, but only 24 percent of computers costing over $1 million were acquired within the previous four years. In fact, 70 percent of these most-costly instruments were between four and six years old in 1993. Bioanalytical instruments had the highest average age: 8.2 years. Thirty-one percent of all bioanalytical instruments costing at least $20,000 were eight years or older in 1993.
By individual type of instrument, robots and manufacturing machines had the lowest average age: 3.4 years, reflecting the relative newness of these kinds of instruments and their growing importance. Almost 70 percent of these instruments were less than four years old, and 47 percent were two years old or less in 1993. At the other age extreme, 45 percent of the NMR/EHR spectrometers, 43 percent of the scintillation/gamma radiation/counters/detectors, and 41 percent of electron microscopes in research use in 1993 were more than eight years old.
The molecular/electron/ion beam systems in the database reflected a large variance in age: 43 percent were less than two years old, while 34 percent were over eight years old in 1993.
Average Age by Science and Engineering Field
On average, certain disciplines employed newer instruments than others. The instruments inventoried to computer science units were generally the newest, with an average age of 3.9 years. Instruments inventoried within computer science departments were the newest of all: the average age was 3.2 years, and 30 percent were less than two years old. Reflecting the shift away from a reliance on large mainframe computers, instruments inventoried to computer science facilities (centers) were considerably older, with an average age of 5.4 years (nearly twice the average age of computers in computer science departments). The majority (53 percent) of instruments in computer science facilities were six to eight years old, and only 14 percent were less than two years old (see Table A-5).
On the other hand, the average age of instruments was over six years in three fields: chemistry (6.4 years), biology, and agriculture (both 6.7 years). Further, agriculture had the largest proportion of instruments that were eight years or older: 34 percent.
Average Age of Research Instruments Rated State-of-the-Art
Where to find the data: See Figure 7 for a depiction of the age distribution of the instruments that were deemed state-of-the-art. Table 11 presents the same data on the age distribution of the instruments that were rated state-of-the-art, by detailed type of instrument. (Table 9 presents the status ratings of all instruments.)
Twenty-seven percent of all instruments were rated state-of-the-art. This high rating was strongly associated with instrument age: as instruments became older, they were less likely to be rated as state-of-the-art. For example, 51 percent of all instruments less than two years old in 1993 were rated state-of-the-art, but only 10 percent of instruments over eight years old were given this high rating (see Figure 7).
Only 44 percent of computers less than two years old were given the high state-of-the-art rating; this was the lowest percentage of any major instrument category. (By comparison, 64 percent of bioanalytical instruments under two years old were given this high rating.) The low proportion of these relatively new computers earning the highest rating is not entirely surprising; given the extremely rapid rate of technological changes in computers, state-of-the-art status is a fleeting honor.
On the other hand, certain other individual types of new instruments in the academic research setting were very likely to be rated state-of-the-art: 86 percent of the electron microscopes and 82 percent of robots less than two years old were given this high rating.
Certain types of relatively old research instruments hold their state-of-the-art rating. For instance, 27 percent of the DNA/protein synthesizers/sequencers/analyzers and 23 percent of the temperature/pressure control/measurement instruments that were older than eight years were still given a state-of-the-art rating.
Interestingly, the newest temperature/pressure/measurement instruments did not follow the same pattern, in which newer instruments were more likely to be rated state-of-the-art. Only 22 percent of the instruments less than two years old were rated state-of-the-art, which is about the same percentage as those that were over eight years old (23 percent) and less than half the percentage of those four to six years old (60 percent were rated state-of-the-art) (see Table 11).
General Working Condition of Research Instruments
Source of the Data
To further ascertain the research capabilities of the instruments, respondents were asked to describe the "equipment's general working condition in FY 1993" on a rating scale ranging from excellent (1) to inadequate (5).
Where to find the data: Figure 8 depicts the general working condition of research instruments in usage in 1993, by field of science and engineering.
In general, respondents were satisfied with the working condition of the instruments in research usage in academic settings. Seventy-four percent of all instruments were judged to be in either above-average or excellent working condition during the year. In fact, chemistry and agriculture were the only two fields in which respondents reported that less than 40 percent of their instruments were in excellent working condition during 1993. Further, agriculture was the only field in which as many as 10 percent of the research instruments were in inadequate or poor working condition during 1993 (see Figure 8).
Technical Capabilities of Research Instruments
Source of the Data
Respondents were asked to rate the "equipment's technical capabilities (resolution, speed, etc.) to meet the needs of the research users" on a scale ranging from excellent (1) to inadequate (5).
Where to find the data: For a graphic depiction of the capabilities of the stock of instrumentation by field of science and engineering, see Figure 9; by type of instrument, see Figure 10.
Overall, respondents were satisfied with the technical capabilities of the instruments in meeting the researchers' needs. Thirty-two percent of the instruments were rated excellent in this area: the modal response for this question. An additional 29 percent were rated above adequate, and 26 percent were adequate. Only 13 percent of instruments were rated as having technical capabilities that were below adequate or poor (see Figure 9).
By field, more respondents from biology (37 percent) and engineering (36 percent) gave an excellent rating to their instruments' technical capabilities than respondents in the other fields. Respondents from computer science were the least satisfied with the technical capabilities of their equipment: perhaps they were reacting to the rapid technological change in computers. Thirty-nine percent rated their instruments either less than adequate or poor.
By type of instrument, respondents were most satisfied with the capabilities of bioanalytical instruments, in which 43 percent were rated excellent, and an additional 29 percent were rated above adequate. Microscopy instruments were a close second in technical capability, with 40 percent of the instruments rated excellent and an additional 31 percent rated above adequate. Only 22 percent of computers were rated as having excellent technical capabilities in terms of meeting the researchers'needs, and 21 percent were rated above adequate. Twenty-nine percent of all computers were judged to have less than adequate to poor technical capabilities (see Figure 10).
Usage and Users of Instrument Systems
Data Considerations for this Section
The survey encompassed only those instruments that were used partially or totally for research. Therefore, instruments used exclusively for instruction or for administrative purposes were excluded from the survey.
Usage of Instruments
Source of the Data
Each respondent was asked to check whether the sampled research instrument was used entirely for research, used predominantly for research with some instructional use, or used predominantly for instruction with some research use.
Where to find the data: Table 12 presents the three patterns of usage by detailed type of instruments.
Of the estimated 61,500 instruments in research usage in 1993, 64 percent were used exclusively for research. Most of the remaining instruments (32 percent) were utilized predominantly for research with some instructional use, and only 4 percent of the total were used primarily for instruction with some research usage.
The usage pattern of computers ran counter to the utilization of the rest of the research instruments. Computers were the only type of instrument in which "exclusively for research" constituted less than half of their usage (only 48 percent of all computers covered by the survey were used exclusively for research) (see Table 12). However, while computers were often used for instructional purposes in 1993, very few were used predominantly for instruction. None of the higher-priced computers costing over $500,000 were used primarily for instruction, and less than 10 percent of each of the price categories of research computers under $500,000 was used primarily for instruction.
The usage pattern of instruments that were used exclusively for research varied widely by type of instrument. At the high end of research use, 85 percent of electron/auger/ion scattering instruments, 84 percent of lasers and optical instruments, and 81 percent of all the bioanalytical instruments covered by the survey were used exclusively for research. At the other end, less than 35 percent of three kinds of instruments were used exclusively for research: computers over $1 million (14 percent), electron microscopes (28 percent), and telescopes and astronomical instruments (34 percent ).
Research vessels/planes/helicopters had the highest percentage of instructional use of any type of research instrument covered by the survey (23 percent were used primarily for instruction). These instruments were followed by NMR/EPR spectrometers (15 percent) and telescopes/astronomical instruments (13 percent).
Users of Research Instruments
Source of the Data
Respondents were asked to list the instrument users and were given the following categories for classification: faculty of the home department/facility, graduate students and post-doctorates from the home department/facility, researchers from other departments/facilities of the home institution, researchers outside the home institution, and other. (Respondents checking the "other" category reported that it comprised primarily undergraduates and staff.)
Where to find the data: Table 13 depicts the mean number of research users of each detailed type of instrument, regardless of the rated adequacy of the instrument. Table 14 presents the mean number of users per instrument by the rated adequacy of the detailed instruments. Table A-6 depicts the mean number of users of the 27 percent of all instruments that were deemed state-of-the-art. Figure 11 is a depiction of the usage pattern of state-of-the-art instruments and of all instruments combined.
Overall, there was an average of 24.2 users per research instrument system in 1993. Repeating the pattern of usage found in previous instrumentation surveys, the largest single category of user was comprised of the graduate students and post-doctorates assigned to the same unit (host unit) where the instrument was inventoried: In total, there was an average of 8.5 of these users per instrument. On average, there were also 3.5 faculty users from the host unit, 6.0 researchers from other units in the host institution, 4.5 researchers from outside the host institution, and 1.8 other users, primarily staff and undergraduates (see Table 13).
Although graduate students constituted the largest user-component for every major category of instrument, some individual types of instruments show a different user pattern. For instance, the in-scope computers costing over $1 million had an average of almost 2,100 users per instrument, of which the majority were researchers from outside the host unit but within the host institution. For computers costing between $500,000 and $999,999, researchers from outside the host institution were the single largest group of users on average. This illustrates the importance of the networking of these large, extremely powerful mainframe computers and supercomputers.
For certain major other instruments, such as nuclear reactors/nuclear science instrument systems and research vessels/planes/helicopters, researchers from outside the host institution were also the single largest group of users on average. Table 13 shows how important these costly instruments are as tools for the wide research community.
Mean Number of Users by Research Status of Instruments
In general, the higher the ranking of an instrument in terms of start-of-the-art status, the higher the number of researchers who used it. For instance, an average of 25.7 researchers used state-of-the-art instruments, and an average of 24.2 researchers used instruments that were not state-of-the-art but were adequate for their research. The usage dropped to 20.5 researchers working on instruments that were rated inadequate for the needs of the researchers in the instrument's department/facility. Computers most exemplify this usage pattern. There was an overall average of 58.5 users of research computers costing at least $20,000. An average of 88.5 researchers utilized those that were deemed state-of-the-art (14 percent of computers), while a much lower average of 26.5 researchers used computers that were rated inadequate (23 percent of all computers). (See Table 14 for average number of users by research status and Table 9 for the rated research status of each major category of instrument.)
There were certain exceptions to the general pattern of more researchers using state-of-the-art instruments. Particularly within the category of chromatographs and spectrometers, a higher average number of researchers used inadequate instruments than used state-of the art versions. For example, Table 9 illustrates that only 9 percent of all chromatographs and spectrometers were rated inadequate; for those inadequate instruments, there was a higher average number of research users per system (21.0) than for those instruments rated state-of-the-art (14.0 users).
As stated previously, 27 percent of all instruments received state-of-the-art ratings from the respondents. An additional tabulation was made to ascertain if there were any differences in the types of users who performed research with state-of-the-art instruments compared to the users of all instruments combined. Figure 11 illustrates that the pattern of users of state-of-the-art instruments was remarkably similar to the usage reported for all instruments.
 For the reader's convenience, the text discussion in this chapter generally rounds dollars to the nearest million and numbers to the nearest hundred. The accompanying tables, however, contain the unrounded data.
 This is in contrast to Chapters I and II, which present data on the overall inventory of instrumentation at the nation's major R&D-performing universities and medical schools. Data for these two chapters came from both survey questionnaires, the Instrument Data Sheets and the Supersystem Data Sheets. See Tables 4 and 5 for explanatory details on the make-up of supersystems and the Introduction and the Technical Notes for a more complete discussion of these instruments.
 Findings from the Instrumentation Survey's companion Department/Facility Questionnaire sent to department chairs and heads of facilities indicated that the Federal Government provided 52 percent of the funds that were used to purchase new instrumentation in 1993. See Academic Research Instruments: Expenditures 1993, Needs 1994 (NSF 96-324).
 Institutional funds generally come from one of four sources: indirect recovery from awards from the Federal Government and other sources; state operating appropriations from general revenues; student tuition and fees; and unrestricted gifts and income (e.g., endowments).
 The unrounded figure$208 millionwas the estimated amount spent on maintenance/repair for the research instruments surveyed with the Instrument Data Sheet questionnaire (i.e., those instrument systems with an original purchase price of $20,000 or more that were not supersystems). Asking the question on a per-instrument basis allowed the tabulation of the percentage of M/R to purchase price by type of instrument (see Table 8). A similar question was asked on the survey's companion Department/Facility Questionnaire sent to department chairs and heads of facilities. The question was designed to capture the total maintenance/repair expenditures for the respondents' entire units, which in concept also covers the less-expensive instruments under $20,000 in each unit as well as the units that were supersystems. Department/Facility Questionnaire respondents reported a total expenditure for maintenance/repair of $234 million in 1993. It must be noted that the respondents to both questionnaires had the option to estimate their maintenance/repair costs, and therefore, the reliability between the two national estimates is not known.
 In other words, the point at which half the M/R expenditures were above this figure and half were below.
 Except for age, the tabulations in this survey classify an instrument by the information given for the original instrument plus any dedicated accessories. For age, however, the instruments are classified by the age of the original instrument only.
 When a very large computer or any other instrument system was upgraded in place, the cost of the upgrade was included in the tabulations as an increased purchase price for the system, and the respondents rated the whole system based on its upgrade. However, its age classification remained that of the original system, without any adjustment for the dates that any new accessories might have been added.
 Again, the reader is reminded that computers are classified according to the age of the original system, regardless of whether or not there had been a major upgrade in intervening years.
 To be considered in-scope for the survey, an instrument must have been in operable condition at some time in 1993. Data were not collected for obsolete, unused instruments. Therefore, this report includes only those instruments which were operable and otherwise available for research usage during 1993.
 All discussion of instruments in this section refers to the instruments detailed in Table 6. Their total aggregate cost was $5.1 billion.
 In this section, which deals with instrument users by fine detail, the reader should be cautioned that the numbers in particular cells may be very small, and therefore, there may be a large standard error. See the Technical Notes for details.
 Although these usage figures exclude data from the supersystems included in Chapters I and II, it can be assumed that supersystems also have a very profound effect on the wide research community.