Section C.
An Assessment of Needs for Academic Research Instrumentation

Adequacy of Current Equipment to Meet Researchers' Needs
Top Priority Items of Need
Need by Type of Instrument
Top Priority Items of Need by Detailed Type of Instrument
Reasons for Needing the Top Priority Items
Special Analysis: Measuring Needs
Price-Range of Optimal Federal Funding
Perceived Limitations of Current Instrumentation


Data Considerations
Respondents were asked about expenditures for research instrumentation purchases on the basis of their actual 1993 expenditures, and the analysis and tables in the previous sections reflect those expenditures. Respondents were also asked for their perception of the needs of their units at the time they were filling out the survey form in 1994. For this reason, tables and analysis in this report concerned with the respondents' perceptions of the adequacy their current instruments, and the need for new instrumentation, reflect the collection date of late 1994.

The Question
Respondents were given a five-point rating scale ranging from "substantially increased" (1) to "substantially decreased" (5), and were asked to assess, "over the past two years, the needs for research equipment in my unit."

Findings
Overall, 69 percent of respondents reported that their needs had increased in the two-year period: Forty-five percent said that their needs had increased, and an additional 24 percent said that their needs had increased substantially. Thirty percent said that the needs had remained about the same, while only 2 percent reported that the needs had decreased (table A-8). Thus, for many disciplines, increased needs since 1992 were set against a leveling of expenditures since the 1988-89 survey (table 1).

Adequacy of Current Equipment to Meet Researchers' Needs

The Question
To set the context for specific questions about needs, respondents were asked about the capability of their current instrumentation. Given a five-point scale ranging from "excellent" (1) to "poor" (5), they were asked to rate "the overall capability of the research equipment in (my) unit to enable existing faculty investigators[12] to pursue their major research interests."

Findings
Overall, 42 percent of the respondents from the larger institutions gave ratings of less than adequate to the capability of their research instruments to enable the faculty to pursue their major research interests. Respondents with the highest percentage reporting inadequacy were from chemistry (61 percent) and from computer science facilities (51 percent). However, only 26 percent of computer science departments rated their instrumentation as inadequate. A large minority in several other fields reported inadequacy: Forty-nine percent of respondents in engineering and in physics/astronomy, and 46 percent of environmental respondents. Agriculture (35 percent) and biology (32 percent) were the fields with the lowest percentage of respondents reporting inadequate equipment (table A-9). As the modal response, 39 percent of the respondents rated their instrumentation as inadequate to support the research interests of their faculty—a rating of 4 on the five-point scale (table A-9).

The Question
If the respondents answered that the instrumentation was "inadequate" or "poor," they were asked to estimate "the cost to acquire sufficient research equipment that would fully support your existing faculty."

Findings
The 42 percent of respondents who rated their instruments as being less than adequate estimated that it would cost a total of $1,438 million to bring their research instrumentation to a point that would fully enable existing faculty to pursue their major research interests. Engineering respondents recorded the highest amount necessary, $435 million. On a median cost-per-unit basis however, engineering needs at $500,000 were in the mid-range of costs: The median per unit cost to acquire this equipment ranged from $950,000 in chemistry to $250,000 in the agricultural and multidisciplinary sciences (table 6). (See also the special analysis of the comparison of needs of these respondents with those who rated their instrumentation as adequate, beginning on page 33.)

Top-Priority Items of Need

Presentation of the Data

As would be expected from so large a cohort of respondents, responses to this question yielded a great amount of data. To assist in user analysis, the data are presented in several tabular ways, allowing the researcher to review the data by type of instrument needed, as well as by discipline. It is not appropriate in a report of this type to describe each table in great detail. Instead, for the convenience of the reader, the text will highlight the different types of analyses that can be derived from the data, and indicate the tables where the information can be found.

The sampled instruments were tabulated into five major categories: computers; chromatographs and spectrometers; microscopy instruments; bioanalytical instruments; and "other" instruments.

Readers wishing to ascertain the cost of acquiring the most needed instruments will find that table 7 presents data on the total cost of the top priority item and the top three priority items needed, disaggregated by field of science or engineering.

Appendix table A-10 presents the cost of acquiring the needed instruments on a per-unit basis, by displaying the median cost, disaggregated by major instrument category and by field. Table A-10 also shows how popular the five major categories of instrument were, by presenting the frequency of requests by respondents for them.

Appendix table A-11 gives greater detail on the top priority items shown in table 7; table A-11 presents total cost in dollars for each major category of instrument. It also depicts the concentration of need within fields. For instance, the cost of the top priority items requested by respondents in two fields represented 51 percent of the total cost: The requests of respondents in engineering and in physics/ astronomy were 27 percent and 24 percent of the total cost, respectively. (Table A-12 presents identical data to table A-11 for the top three priority instruments combined.)

Finally, for researchers who want to know which types of instruments are needed by individual disciplines, table 8 presents the greatest amount of detail (i.e., the total cost of the top priority instruments by detailed type of instrument). (Table A-13 presents identical data to table 8 for the top three priority instruments combined.)

Context
The rating-scale questions on capability discussed above were asked in terms of the overall capability of the stock of existing instrumentation to enable current faculty to pursue their major research interests. In a slight variation of that question, respondents were requested to name three pieces of equipment that were most needed to bring the unit's research equipment up to the faculty's full capabilities. While 58 percent of the respondents were satisfied with the overall capability of the instrumentation (see discussion above), 90 percent had some need for additional items.

The Question
Respondents were asked to indicate "The three pieces of equipment, costing $20,000 or more (including the cost of accessories), that are most needed to bring your unit's research equipment up to your faculty's full capabilities." They were asked to list these three top items in priority order and to estimate the purchase price of each item.

(Realizing that not all needed instruments were in this high-cost category, particularly since many very high-powered personal computers and workstations can be purchased for less than $20,000, the questionnaire also allowed respondents to respond that no additional equipment costing more than $20,000 was needed, which 10 percent of respondents did. The findings below are based on the 90 percent of respondents who indicate that their topmost priority was for instruments over $20,000.)

Findings
If all three of the top priority items needed by each respondent could be purchased, the total estimated cost would be $2,048 million. Of this amount, $942 million, or 46 percent of the total, was the estimated cost of acquiring only the topmost priority item (table 7).

Need by Type of Instrument

Computers.—Twenty-eight percent of all respondents listed a computer as their topmost priority, a graphic illustration of the importance of computers to research in all fields. The range of top priority needs for computers went from highs of 100 percent of respondents in computer science departments and 35 percent in engineering and the environmental sciences, to a low of 14 percent of respondents in chemistry and biology (table A-10).

The cost to acquire only computer top-priority items was $202 million, or 21 percent of the total cost of all first-priority items. The leading fields by cost were computer sciences ($99 million) and engineering ($48 million) (table A-11).

The median cost for all computers listed as top priority was $75,000 (table A-10). As would be expected, the median cost cited by computer science facilities was much higher than that reported in other fields, $200,000. This reflects the high cost of mainframes and supercomputers needed by the central facilities still operating for research, many of which perform computations for off-site researchers.

Chromatographs and spectrometers.—The total cost for all chromatographs and spectrometers listed as top priority need was $213 million, and 21 percent of respondents reported them as their top priority need (table A-10). The majority of need was in biology ($62 million), chemistry ($57 million), and engineering ($53 million). Proportionately, however, chemistry respondents had the highest demand for these instruments ($57 million out of $62 million). This category includes NMR/EPR spectrometers as well as instruments such as x-ray diffraction systems. In terms of cost, most of the need cited by chemistry respondents was allocated to NMR/EPR spectrometers, which comprised 83 percent of the cost of the top priority instruments mentioned by the chemistry respondents.

Bioanalytical instruments.—The total cost for all bioanalytical instruments listed by respondents as the top priority need was $76 million, and 20 percent of all respondents mentioned this category as their top priority item (table A-10). As would be expected, respondents from biology fields expressed the highest need-$36 million, or 47 percent of the total cost. Respondents in physics/astronomy also needed a sizable amount of bioanalytical instruments, totaling $19 million, or 25 percent. The need in engineering fields comprised $11 million, or 15 percent (table A-11).

Microscopy instruments.—The total cost of the top priority items in microscopy was $76 million, the same total as for bioanalytical instruments. However, fewer respondents cited a need for this category (11 percent) than for bioanalytical instruments (20 percent). Respondents from the biological sciences cited a need for 47 percent of the total cost of top priority microscopy instruments, $36 million. Thirty-one percent of the total cost, or $24 million, was needed by respondents from the engineering fields (table A-11). Engineering respondents reported that the bulk of their need was for electron microscopes (table 8).

"Other" instruments.—"Other" instruments was used to describe miscellaneous instruments, none of which was large enough to constitute a category of its own. Taken together, however, "other" comprised the largest category of need, both in terms of total cost as well as the percentage of respondents mentioning the item. Twenty-nine percent of respondents cited "other" instruments as their top priority needs (table A-10). The $375 million cost comprised 40 percent of the total cost of all top priority items (table A-11).

"Other" instruments includes lasers, robots, temperature/pressure control devices, and "major instruments," the most costly single type of instrument requested. ("Major instruments" are often unique, individually fabricated items-such as nuclear reactors, research vessels, wind tunnels, and telescopes). Because each major instrument is very costly, it is not surprising that the $375 million total cost is the highest of all the categories.

The need for "other" instruments as a top priority was not uniform throughout the sciences, but was heavily concentrated in physics/astronomy ($190 million, with 59 percent of the respondents mentioning items in this category), engineering ($119 million, 32 percent), and environmental sciences ($48 million, and 19 percent.) (Table A-10 shows the percentage of respondents requesting a particular category of instrument. Table A-11 shows the dollar cost for each category of instrument.)

Top Priority Items of Need by Detailed Type of Instrument

While table 7 depicts the total cost of the top priority item and the top three priority items by each field of science, table 8 presents a detailed picture of the exact types of instruments needed, by field of science, and the total cost of each top priority item. The reader can find information on the types of instruments needed, the magnitude of the need, and the concentration of instruments within the fields of science. For example, only computers were listed by respondents in every field as the top priority item needed in their units. (Table A-13 presents identical data to table 8 for the top three priority instruments combined.)

Reasons for Needing the Top Priority Items

The Question
When listing the three topmost priority items needed by the unit, the respondents were asked to "state the primary purpose for acquiring the instrument-whether to (1) replace an existing item; (2) expand capacity-i.e., more copies of existing equipment; or (3) upgrade capabilities-i.e., perform experiments that you cannot do now." It must be understood that in many cases these reasons are not mutually exclusive-for example, few researchers replace an existing item with its exact replica; today's instruments are often better, faster, and more complex than their predecessors.

Findings
Overall, 47 percent of respondents in all fields reported that the primary reason they needed the top priority research instrument was to "upgrade capabilities" for the unit, i.e., to perform experiments that they "cannot do now." This was the modal response for all fields except chemistry, where only 32 percent of respondents mentioned this need. (At the same time, 48 percent of respondents in chemistry cited "replace an existing item" as their primary reason for the need (figure 5). Only 25 percent of respondents overall reported that they needed the item to replace existing equipment.)

Twenty-eight percent of all the respondents cited expanding capacity, or the need for more copies of a particular item, as the primary reason for need.

Special Analysis: Measuring Needs

Issue 1—Specific items on the questionnaires

Context
It is very difficult to devise a questionnaire that will produce national estimates of the pressing needs for research instrumentation in academia. Questions must be devised so that the national estimates reflect the mix of need; that is, it is recognized that some units are always highly successful in obtaining the bulk of their most needed instrumentation, while others seem always to be in dire need. The survey's national estimates must reflect the needs of both of these types of units. Therefore, the instrumentation survey questionnaire asked about need in two ways:

Issue 2—Comparisons of costs of top priority needs with costs to correct overall inadequate instrumentation

Context
Short of analyzing the internal records of every respondent, there is no definitive way to determine whether the reported top priority needs reflected the respondents' true needs, or were a "wish list" of instruments that would be helpful. It has been assumed that respondents reply to the survey in good faith, and that an enumeration of their three most pressing top priority needs represents some portion of their total needs. However, the addition of the new question added to the 1993 survey makes it possible to analyze the relationship between the top three priority needs of the group whose instrumentation is inadequate, and the amount to bring the instrumentation up to support their faculty's research needs. Presumably, the top three priority needs would indicate the most pressing needs, and not equal the entire amount of expenditures necessary to support the existing faculty.

Findings
In all fields, this was exactly the case: The cost of the top priority items needed by the "inadequate" group ($939 million for all fields) was less than the estimated cost to "fix" their entire instrumentation needs ($1,438 million). Closest in total were the needs of the computer science respondents, where the $68 million in three "top priority" needs was 89 percent of the cost to bring instrumentation up to a level that would satisfy the faculty (figure 6). This is not surprising, as the addition of three state-of-the art computers in any one "inadequate" unit would make a major impact on the ability of the faculty to conduct research effectively.

The largest percentage disparity was in agriculture. The $20 million cited as the total cost of their three top priority needs was only 29 percent of the cost of bringing the inadequate group's instruments up to a level that would completely satisfy the faculty's needs. The largest cost disparity was in engineering. Respondents reported a total need of $435 million, or $134 million more than the cost of acquiring their three top priority needs. The $301 million in top priority needs was 69 percent of what would be needed to bring the instrumentation up to the needs of the faculty.

The needs of the "inadequate" group were also compared with their level of spending in 1993. In all fields the cost cited to acquire sufficient instrumentation during the survey collection period in 1994 was several times the level of expenditures that they were able to achieve in 1993 (figure 6). The largest disparity was in engineering. At $79 million, the "inadequate" group's 1993 expenditures for new purchases were the second highest of all fields. Nevertheless, they estimated that it would take an additional $435 million to bring their instrumentation up to a level that would satisfy the needs of the current faculty.

Issue 3—Comparisons of the data reported by respondents reporting adequate and inadequate instrumentation

Context
An analysis was made to determine the extent of related differences, for both expenditures and needs, in the data reported by the 42 percent of respondents who rated their instruments as less than adequate for their faculty's research needs, compared with the 58 percent who reported their instruments as adequate.

Findings
The "inadequate" group as a whole had lower total purchases of new instrumentation in 1993, and lower median per-unit purchases, than did the "adequate" group. Conversely, perhaps reflecting the lower outlays, the total cost of the top priority need cited by the "inadequate" group ($507 million) was higher than total cost of the top priority need reported by the "adequate" group ($436 million). The total cost of the top three items needed by respondents, however, was higher for the "adequate" group ($1,111 million) than for the "inadequate" group ($939 million).

1993 Expenditures.—Lower yearly expenditures perhaps point to much of the cause of dissatisfaction in the "inadequate" group. Figure 7 illustrates that this group spent less than their proportionate share in new research instrumentation in 1993. The 42 percent reporting inadequate instrumentation spent $334 million on new purchases during 1993, only 28 percent of the total. The 58 percent reporting adequate instrumentation spent $870 million, or 72 percent. (On a per-unit basis, the median expenditure by the "inadequate" group was $100,000, versus $235,800 by the "adequate" group.)

In some fields the "inadequate" group reported particularly low proportions of total expenditures. In chemistry, the 61 percent reporting inadequate instrumentation made just 27 percent of the total purchases in chemistry in 1993. In computer science, the 38 percent reporting inadequate instrumentation made 17 percent of the total purchases in 1993.

Price-Range of Optimal Federal Funding

The Question
Respondents were asked to state the price-range of instruments that would be most beneficial to faculty investigators in the unit, if increased Federal funding were possible.

Findings
The modal response of all the respondents was in the $20,000 to $49,999 range (30 percent stated that this price range was potentially the most beneficial.) Close behind, 29 percent of respondents mentioned the high-price range of $100,000 to $499,999 (table 9).

The majority of respondents in two fields replied that the most beneficial price-range area was under $50,000: agriculture (59 percent) and biology (54 percent). (See related discussion of median cost of top priority needs, above.) Their needed instruments tend to be less expensive than those needed in several other fields. It is therefore reasonable to expect that the price-range for most needed Federal funding, if available, would also be lower than for those other fields.

In contrast, only 13 percent of the chemistry respondents replied that the most beneficial price-range area would be under $50,000. The modal response for chemistry (51 percent) cited items between $100,000 and $500,000.

At the very upper end of the price range for most beneficial Federal funding, only 6 percent of all respondents cited a preference for the very expensive instrumentation over $500,000. The few fields whose respondents expressed a substantial need in this area were physics/astronomy (17 percent), chemistry (18 percent), and computer science facilities (51 percent).

The respondents from computer science facilities indicated a strong preference for the highest-priced items: Forty-two percent reported that the optimal range for Federal funding would be for instrumentation over $1 million. This finding is consistent with the previous discussion-many campuses no longer utilize their central computer facilities for research. For those that remain in the research domain, the trend is toward consolidation, with a need for very expensive mainframes and supercomputers. For example, as seen in table A-10, the median cost for a top priority computer for computer science facilities ($200,000) is far greater than the cost of the top-needed computer for any other discipline.

Perceived Limitations of Current Instrumentation

The Question
Respondents were asked, "Are there any important subject areas . . . in which investigators in this (unit) are unable to perform critical experiments in their areas of research interest due to lack of needed equipment?"

Findings
In 1993, slightly more than half of all respondents (56 percent) reported that there were subject matters in which their faculty investigators were unable to perform critical experiments because needed equipment was lacking (figure 8). There has been a marked decrease in the percentage of respondents reporting this deficiency since 1983-84, when 74 percent of the respondents reported such limitations. The greatest decrease between 1983-84 and 1993 occurred in computer science, from 96 percent in the earlier survey, to 44 percent in 1993, the lowest percentage reported in any field.

Although there has been steady improvement in the availability of research instruments to academic researchers, the continued lack of specific instruments is a significant limitation for the scientific community, where a majority of respondents in all fields but computer science reported an inability to perform critical experiments because of a lack of instruments. The greatest proportional need occurred in other, multidisciplinary fields (66 percent of respondents), agriculture (65 percent), and physics/astronomy (64 percent) and chemistry (64 percent).



Footnotes

[12] It is important to note that respondents were asked to rate the capability of instrumentation for the existing faculty currently in place at the institution. Although it might be in the best interest of any individual institution to need different equipment to hire a new faculty member, on a national basis it was not appropriate to include this kind of need in the aggregate.