Division of Information and Intelligent Systems
National Science Foundation*
4201 Wilson Boulevard
Arlington, Virginia 22230 USA
Data from 3909 respondents to an Internet survey questionnaire provide the first insights into public perceptions of nanotechnology. Quantitative analysis of statistics about agreement and disagreement with two statements, one positive and the other negative, reveals high levels of enthusiasm for the potential benefits of nanotechnology and little concern about possible dangers. The respondents mentally connect nanotechnology with the space program, nuclear power, and cloning research, but rate it more favorably. In contrast, they do not associate nanotechnology with pseudoscience, despite its imaginative exploitation by science fiction writers. Qualitative analysis of written comments from 598 respondents indicates that many ideas about the value of nanotechnology have entered popular culture, and it provides material for an additional 108 questionnaire items that can be used in future surveys on the topic. The findings of this exploratory study can serve as benchmarks against which to compare results of future research on the evolving status of nanotechnology in society.
Key words: nanotechnology, nanoscience, sociology, public opinion, Internet, questionnaire, survey
A September 2000 conference on the societal implications of nanoscience and nanotechnology, organized by the National Science Foundation, concluded that it is essential to involve the social sciences early in the development of these fields (Roco & Bainbridge, 2001; cf. Roco & Tomellini, 2002). It is important to understand the evolving socio-cultural context in which research at the nanoscale is funded, the societal needs that nanotechnology may satisfy, and the popular misconceptions that nanoscience education will have to overcome. This article is an initial exploration of public opinion about research and engineering on the nanoscale.
To initiate social-scientific exploration of this topic, I placed three nano-related items into Survey2001, a complex questionnaire study sponsored by the National Geographic Society and the National Science Foundation. Two of these were so-called 'Likert'' items (Bainbridge, 1989: 93) - statements with which the respondent can express agreement or disagreement: 'Human beings will benefit greatly from nanotechnology, which works at the molecular level atom by atom to build new structures, materials, and machines.' 'Our most powerful 21st-century technologies - robotics, genetic engineering, and nanotechnology - are threatening to make humans an endangered species.'
The first of these statements was adapted from the formal definitions of nanotechnology offered in a number of publications of the National Nanotechnology Initiative (National Science and Technology Council, 2000; Siegel et al., 1999; Roco et al., 2000). The second was quoted from Bill Joy's (2000) famous article in Wired magazine that warned against the supposed dangers of nanotechnology and related fields. For each statement, respondents were asked to choose one of five responses: strongly disagree, disagree, do not know, agree, or strongly agree. The third nano-related item presented the two statements again and offered a space where the respondent could write his or her views freely. This approach combines both quantitative and qualitative methodologies in a manner that is especially effective in pilot studies designed to open up a new area of social-scientific research.
Survey2001 was administered over the World Wide Web, beginning in October 2001, and consisted of a number of modules that were given to different subsets of respondents in four languages (English, Spanish, German, and Italian). Here we focus on data from the English-language version. The two agree-disagree items belonged to a module of 30 science-related items, and 3,909 people responded to them all. Web administration offers many advantages to the survey researcher, and this approach is the subject of much debate and development at the present time.
The chief disadvantage of web-administration is that most Internet surveys are not able to employ random samples of respondents. Random sampling accomplishes two things. First, it makes it easy to extrapolate results to the general population, for example predicting the outcome of an election on the basis of a pre-election poll. Second, some of the methods of statistical analysis commonly applied to surveys assume the data came from a random sample, notably tests of statistical significance. However, random samples have become prohibitively expensive in recent years, for many research topics, because public willingness to answer telephone surveys has collapsed in response to the explosion of aggressive telemarketing (El Boghdady, 2002).
A major drawback of random samples for research on public attitudes toward science is that only a small fraction of the population is attentive to scientific issues - that is, both interested and well informed. For example, the report of a 2001 survey of public attitudes toward science, conducted by the Science Resources Statistics (SRS) division of the National Science Foundation, says, '...for most issues covered by the NSF survey, fewer than 10% of the public could be considered attentive' (National Science Foundation, 2002: 7-9; appendix table 7-7). Since just 1574 people responded to the SRS survey, only roughly 157 (10% of 1574) of them had well-grounded opinions about science and technology. For a given cost, web-based surveys typically have larger numbers of respondents. The science module of Survey2001, with 3909 respondents or about 2.5 times as many as the SRS survey, will have a much greater number of respondents attentive to science and technology issues, because people recruited to web-based surveys through universities or National Geographic are more educated and more science-savvy than the average.
Modern science policy is not determined by voters in elections or referendums, nor by random samples of the population, but through informal processes in which knowledgeable opinion leaders influence decision makers in government and industry. Thus, it may be more important to study the views of knowledgeable people who want to express their considered opinions, than the views of the inarticulate and inert general public. This would be especially true for exploratory studies that seek to develop hypotheses and appropriate measures in new areas, because it is always possible to carry out an expensive survey of a random sample of the population to verify results, once research in a new area has been established on a firm footing.
Much effort is currently being invested in developing appropriate analysis techniques for web surveys (Witte et al., 2000; Best et al., 2001). Research has shown that web-based surveys can do a good job of measuring relationships between variables, even when their respondent samples imperfectly reflect the precise level of each variable in the general population (Bainbridge, 2002). Web surveys also have a tremendous potential to launch new fields of study, by developing the tools for future research.
We will begin the first quantitative analysis of data about public attitudes toward nanotechnology, by examining how the two agree-disagree items relate to each other and to eight other items in the same module, six of them concerning conventional science and technology, and two concerning pseudoscience. As Table 1 shows, in response to the first statement asserting that 'human beings will benefit greatly from nanotechnology,' fully 57.5% of the 3909 respondents agreed. In contrast, only 9.0% agreed with Bill Joy's assertion that nanotechnology is 'threatening to make humans an endangered species.' This is a vast difference in the levels of agreement, and thus we would expect to see much more support for nanotechnology than opposition even in a random sample of the general public.
|Statement||Percent who AGREE||Correlation (r) with item about:|
|Human beings will benefit greatly from nanotechnology, which works at the molecular level atom by atom to build new structures, materials, and machines||57.5||1.00**||-0.68*|
|Our most powerful 21st-century technologies - robotics, genetic engineering, and nanotechnology - are threatening to make humans an endangered species||9.0||-0.68*||1.00**|
|Funding for the space program should be increased||47.5||0.38*||-0.22*|
|Space exploration should be delayed until we have solved more of our problems on Earth||34.8||-0.35*||0.30*|
|Development of nuclear power should continue, because the benefits strongly outweigh the harmful results||34.2||0.31*||-0.31*|
|All nuclear power plants should be shut down or converted to safer fuels||32.1||-0.28*||0.33*|
|Research on human cloning should be encouraged, because it will greatly benefit science and medicine||31.7||0.37*||-0.29*|
|There should be a law against cloning human beings||52.3||-0.26*||0.22*|
|Some scientific instruments (e.g., e-meters, psionic machines, and aura cameras) can measure the human spirit||9.1||-0.02||0.05|
|Perpetual motion machines, anti-gravity devices, and time travel machines are physically impossible||29.3||-0.02||-0.03|
|*Statistically significant beyond the 0.001 level; significance can be conceptualized in terms of the likelihood that the coefficients could be reproduced by randomly reassigning responses across respondents, if the conventional interpretation in terms of a random sample of the population is considered inapplicable rather than merely approximate. **Autocorrelation of one item with itself.|
In addition to the percentages agreeing, Table 1 also reports the correlations between items. These are Pearson's r product-moment correlation coefficients, that are limited mathematically to the range from -1.00 through 0.00 to +1.00. The -0.68 negative correlation between the two items shows that there is a very powerful tendency for people who agree with one of these items to disagree with the other. In attitudinal survey research one seldom sees a correlation with an absolute value this large, except of course in the substantively uninteresting case of correlating an item with itself, which gives 1.00.
The next six items are similar pairs - about the space program, nuclear power, and cloning - in which one item represents support for the technology, and the other measures opposition. The correlations between these six items and the two nanotechnology items are all rather substantial by the standards of survey research, with absolute values in the range 0.22 - 0.38. Clearly, people who are in favor of nanotechnology also tend to support the space program, nuclear power, and research on cloning. Although opposition to nanotechnology is rare, it correlates with opposition to the three other technologies. Interestingly, more people agree with the pro-nanotechnology statement, and fewer agree with the anti-nanotechnology statement, than with the corresponding statements for the other technologies.
The bottom of Table 1 reports results for two statements about pseudoscience, to explore the possibility that people may mentally assign nanotechnology to this disreputable category, perhaps on the basis of wild ideas popularized in science fiction stories. However, the correlations between the two nanotechnology items and the two pseudoscience items are as close to zero as one would expect to get in survey research, with absolute values ranging from 0.02 to 0.05. In comparing correlation coefficients, statisticians square the numbers first, to get the fraction of the variance in one variable explained by the other. Thus, we would not compare 0.05 directly with 0.38, which is a ratio of 0.05/0.38 = 0.13, but first square them to get 0.0025/0.1444 = 0.02. A correlation of 0.05 is statistically indistinguishable from zero, even with as many as 3909 respondents, so we have to conclude that nanotechnology has no measurable connection with pseudoscience in people's minds, even as they connect it strongly with other kinds of genuine technology.
As other researchers make progress analyzing their own portions of Survey2001, it will be possible to examine relationships between the nanotechnology items and questions in other modules of the questionnaire, such as a number concerning attitudes toward environmentalism. For now, we must be content with the science items and some of the demographic variables that might explain variations in attitudes. Table 2 looks at four categorical variables: gender, education, political orientation, and age.
|Percent who agree with item about:|
|Less than 4 years of college||1186||52.2||10.6|
|Middle of the road||1037||59.7||8.0|
|Age (years) >45||1155||63.1||8.2|
|Age (years) 31-45||1271||56.6||7.6|
|Age (years) 21-30||1126||53.5||11.0|
|Age (years) <21||349||54.7||10.0|
The table reveals that fully 69.2% of the 1787 men agree with the pro-nanotechnology statement, compared with just 47.6% of the 2111 women. We see smaller differences for the other variables. Putting them together, we would expect to find the greatest support among older, college-educated men who have relatively right-wing political views. However, except for gender, the differences are not great. Thus, the majorities agree that 'human beings will benefit greatly from nanotechnology' even among those who did not complete college, those who stand on the political left, and among those aged 30 or younger.
Future research, including with the present dataset, will seek to understand the reasons for these differences and to discover other associations. Perhaps the best way to progress quickly now, is to examine the written responses to the open-ended item asking respondents to comment on the two agree-disagree statements.
Open-ended questionnaire items, where respondents freely write their opinions about a topic, are a good way to learn what ideas and issues are common in the culture to which they belong. But such items can also be an excellent step toward development of rigorous fixed-choice items for future surveys. For example, one study of popular beliefs about the space program began with a series of open-ended questions about the possible goals of space exploration (Bainbridge, 1991), and distilled from the mass of verbiage contributed by respondents fully 125 statements that became fixed-choice items in a later survey. Similarly, an open-ended item about the future incorporated in Survey2000, the predecessor of the current questionnaire, generated 2000 predictions about the year 2100 that became questionnaire items (Bainbridge, 2000; in press). Here, I will explain how the same method was applied to nanotechnology.
The open-ended nanotechnology item in Survey2001 was administered at random to 1/4 of the people who responded to the agree-disagree items, and 598 of them wrote comments in English. Not all respondents were prepared to provide thoughtful responses, of course. For example, a 39-year old woman from Ohio wrote: 'In all honesty, this is the first time I have ever heard of nanotechnology, so I simply cannot comment on it.' However, hundreds of others were knowledgeable in varying degrees and expressed a rich variety of views. A student wrote, 'Upon reading about nanotechnology for research projects for my university course I can see the potential this could have for medicine and for smaller technology such as computers based upon quantum mechanics.'
Other respondents explained that something about their personal situation shaped their views. A 56-year old woman from Texas explained, 'I am disabled, so I look forward to advances in science to help repair my body's cells.' A 58-year old male college graduate from Illinois commented, 'My cancer treatment has benefited from research, and will likely continue to do so.'
Some had a clear perspective on nanotechnology because of their educations or occupations. A 28-year-old man wrote, 'I have a PhD in physics and am a working engineer, and I strongly believe that scientific and technical research is important and beneficial to our society. Our society is very near-sighted in that it does not see any immediate benefit to many research endeavors, though in truth they may very well have far-reaching impact in the future. Take for instance the discovery of electricity; or thermodynamics, which led to the creation of combustion engines; or various physical properties of light and particles which can be utilized for the purpose of medical imaging, or even the radio or TV in your living room.'
Similarly, a 21-year-old woman wrote, 'I am currently working on a PhD in Materials Science & Engineering and my research strongly involves nanotechnology, so I of course think that this is an important direction to continue to explore and develop our understanding of. MEMS devices and other nanotechnology will help us develop smaller machines and the nano- research on materials will help us understand how to build stronger materials, making for stronger cars and buildings and many other applications. As for robotics etc threatening humanity, that possibility is only far off in the future. Our current abilities to build robotics and manipulate genes and do anything nano is so poor and limited that there is no danger from this area to kill off humans for many many more years. We are more likely to be killed off first from another world war involving atomic bombs or chemical warfare.'
To see how material like this can be used to create new, fixed-choice questionnaire items, consider this complex set of ideas contributed by a 33-year-old man from New Jersey: 'Nanotechnology, even though it is relatively new and its implications are not fully understood, should be a benefit. Nanotechnology may be able to build new medical devices for diagnosis and treatment of pathological states. It will lead to miniaturization of electronics and development of new materials. The only caution should be its military application and its threat for use in information gathering on people's civil liberties.' Each of the specific thoughts about the consequences of nanotechnology can be edited as a separate statement:
However, it is rare that one individual contributes so many different ideas. Often, several people express roughly the same idea, and I either select one who expressed it with superior clarity, or combine the words of several respondents. For example, here are comments from three different people who mentioned medicine: 'Nanotechnology most likely holds many benefits for the field of medicine as does biotechnology and GMO's.' 'I feel there is a great deal of potential in the use of nanotechnology, especially in the area of medicine.' 'I do think nanotechnology can be helpful and beneficial, in medicine as well as engineering.' I condensed these into a single sentence: 'Nanotechnology has great potential benefit for medicine.'
Each of the derived statements could then become a fixed-choice item in a questionnaire, with which respondents could agree or disagree. Then it will be possible to explore the clusters of ideas that inform public attitudes toward nanotechnology, and to analyze statistical correlations that would link them to other variables. Whatever the opportunities for future research, developing a list of potential questionnaire items in this way already outlines the public's ways of conceptualizing the topic.
It is interesting to note than none of the respondents warned about the science-fiction notion of self-reproducing nano-robots or otherwise identified a way that nanotechnology could harm humans directly. Rather, the few critics worried that it would be misused by some people to harm other people, exacerbating existing social inequalities and conflicts. In contrast, respondents collectively were aware of quite a number of direct benefits and found a number of ways to express confidence that nanotechnology would help human beings achieve legitimate goals.
The findings of this initial, exploratory study can serve as benchmarks against which to compare results of future research. Many of the Survey2001 respondents had clearly heard something about nanotechnology, and others were able to react intelligibly to the brief description of the field conveyed by the two agree-disagree items. This work offers 110 questionnaire items that could be used in future social-scientific research, the original pair of statements and the 108 that were derived from respondents' reactions to them. Naturally, it would seldom be desirable to use all of these agree-disagree items, and some have rather similar meanings to each other. A selection can be made that is appropriate for the scientific goals of the particular project.
The last set of critical statements, especially numbers 102-108, suggest a reason why enthusiasm for nanotechnology appears to be slightly weaker at the left wing of the political spectrum. Apparently, some respondents see it first and foremost as a benefit for industry, that will take investment to realize, thus favoring the rich and powerful in society over the poor and powerless. In the comparable field of information technology, this concern has been expressed in the concept of the digital divide (Department of Commerce, 1999), the worry that disadvantaged groups will fail to benefit from computers and advanced communication networks. Perhaps concerns that nanotechnology will exacerbate inequality will evaporate as the benefits actually arrive, or there may be a role for government to ensure that these benefits are widely shared.
The statements do not clarify why women are so much less enthusiastic than men. Had we seen a number of military-related statements (rather than the one rather bland item 51 we did find), suggesting how nanotechnology might be used for attack or defense, we might have inferred that the gender difference on nanotechnology reflects an underlying difference in attitudes toward the military. But apparently the public has not yet become alert to the range of potential defense applications. Future research can explore the extent to which the gender difference merely reflects the far lower interest in engineering and the non-biological physical sciences among women (National Science Foundation, 2000).
The chief finding of this initial, exploratory study is that science-attentive members of the general public are very enthusiastic about nanotechnology, and a rather large number of ideas about its benefits have already entered popular culture. Over the coming years, social scientists in a variety of fields should employ a diversity of research methods and analytical theories to chart and understand the growing significance of nanotechnology for modern civilization.
*The views in this essay do not necessarily represent the views of the National Science Foundation. Survey2001 was partly supported by NSF grant 0082750 to James C. Witte at Clemson University. The author of this paper is an employee of NSF, so care was taken to make sure this work was done in an ethically proper manner, guided by the NSF Office of the General Counsel and with the permission of the author's supervisor. The author did not participate in the preparation or review of the grant proposal, received no support from this grant, and is not involved in management of the grant. The author was asked to join the Survey2001 team on the basis of his prior experience with web-based surveys, including Survey2000. His involvement is limited to designing questionnaire items, analyzing data, and writing reports like this one.
Bainbridge W. S., 1989 Survey Research: A Computer-Assisted Introduction. Wadsworth, Belmont, California.
Bainbridge W. S., 1991. Goals in Space. SUNY Press, Albany, New York.
Bainbridge W. S., 2000. 'Religious Ethnography on the World Wide Web.' In Jeffrey K. Hadden and Douglas Cowan (eds.) Religion and the Internet. JAI Press, Greenwich, Connecticut.
Bainbridge W. S., 2002. 'Validity of Web-Based Surveys.' In Orville Vernon Burton (ed.) Computing in the Social Sciences and Humanities. University of Illinois Press, Urbana, pp. 51-66.
Bainbridge W. S., 'The Future of Internet,' In Philip E. N. Howard and Steve Jones (eds.) The Internet and American Life. Sage, Thousand Oaks, California.
Best, S. J., B. Krueger, C. Hubbard, & A. Smith, 2001. 'An Assessment of the Generalizability of Internet Surveys' Social Science Computer Review 19, 131-145.
Department of Commerce, 1999. Falling Through the Net: Defining the Digital Divide. U.S. Department of Commerce, Washington, DC.
El Boghdady, D., 2002. 'Ears Wide Shut: Researchers Get Punished for Telemarketers' Crimes,' Washington Post, September 8, 2002; p. H1.
Joy, B., 2000. 'Why the Future Doesn't Need Us,' Wired, April 2000.
National Science and Technology Council (NSTC, Subcommittee on Nanoscale Science, Engineering and Technology), 2000. National Nanotechnology Initiative: The Initiative and its Implementation Plan. White House, Washington, DC.
National Science Foundation, Division of Science Resources Statistics, 2002. Science and Engineering Indicators - 2002. National Science Foundation (NSB 02-01), Arlington, VA .
National Science Foundation, 2000. Women, Minorities, and Persons With Disabilities in Science and Engineering: 2000. National Science Foundation (NSF 00-327), Arlington, VA.
Roco, M. C. & R. Tomellini, 2002. Nanotechnology: Revolutionary Opportunities and Societal Implications. Office for Official Publications of the European Communities, Luxembourg.
Roco, M. C., R. S. Williams, and P. Alivisatos, eds., 2000. Nanotechnology Research Directions. Kluwer Academic Publishers, Dordrecht, Netherlands.
Roco, Mihail C., and W. S. Bainbridge, eds., 2001. Societal Implications of Nanoscience and Nanotechnology. Kluwer Academic Publishers, Dordrecht, Netherlands.
Siegel, R. W., E. Hu, and Mihail C. Roco, eds., 1999. Nanostructure Science and Technology. Kluwer Academic Publishers, Dordrecht, Netherlands.
Witte, J. C., L. M. Amoroso & P. E. N. Howard, 2000. Method and representation in Internet-based survey tools: mobility, community, and cultural identity in Survey2000. Social Science Computer Review 18(2), 179-195.