https://onlinelibrary.wiley.com/doi/abs/10.1111/ssm.12051


Quando os cientistas mudam de idéia sobre uma ideia científica, diminui minha
confiar em seu trabalho. *
2. Os cientistas ignoram as evidências que contradizem seu trabalho. *
3. Teorias científicas são explicações fracas. *
4. Os cientistas mantêm intencionalmente seu trabalho em segredo. *
5. Podemos confiar que os cientistas compartilham suas descobertas, mesmo que não gostem de
achados.
6. Os cientistas não valorizam as idéias dos outros. *
7. Confio no trabalho dos cientistas para melhorar a vida das pessoas.
8. Os cientistas não se importam se os leigos entendem seu trabalho. *
9. Devemos confiar no trabalho dos cientistas.
10. Devemos confiar que os cientistas estão sendo honestos em seu trabalho.
11. Devemos confiar que os cientistas estão sendo éticos em seu trabalho.
12. As teorias científicas são confiáveis.
13. Quando os cientistas formam uma hipótese, eles estão apenas adivinhando. *
14. As pessoas que entendem mais a ciência confiam mais na ciência.
15. Podemos confiar na ciência para encontrar as respostas que explicam o mundo natural.
16. Confio que os cientistas possam encontrar soluções para nossos principais problemas tecnológicos.
17. Não podemos confiar nos cientistas porque eles são tendenciosos em suas perspectivas. *
18. Os cientistas se protegerão mesmo quando estiverem errados. *
19. Não podemos confiar que os cientistas considerem idéias que contradizem as suas. *
20. Os cientistas de hoje sacrificam o bem-estar de outras pessoas para promover sua
pesquisa.*
21. Não podemos confiar na ciência porque ela se move muito lentamente. *

I Just Don’t Trust Them: The Development and Validation of an Assessment Instrument to Measure Trust in Science and Scientists Louis Nadelson Boise State University Cheryl Jorcyk Boise State University Dazhi Yang Boise State University Mary Jarratt Smith Boise State University Sam Matson Boise State University Ken Cornell Boise State University Virginia Husting Boise State University Trust in science and scientists can greatly influence consideration of scientific developments and activities. Yet, trust is a nebulous construct based on emotions, knowledge, beliefs, and relationships. As we explored the literature regarding trust in science and scientists we discovered that no instruments were available to assess the construct, and therefore, we developed one. Using a process of data collection from science faculty members and undergraduate students, field testing, expert feedback, and an iterative process of design, we developed, validated, and established the reliability of the Trust in Science and Scientist Inventory. Our 21-item instrument has a reliability of Cronbach’s alpha of .86, and we have successfully field-tested it with a range of undergraduate college students. We discuss implications and possible applications of the instrument, and include it in the appendix. Topics such as climate change, vaccination, biological evolution, and genetically modified foods are increasingly sparking reactions in the general public that revolve around issues of trust in both science and scientists (Ipsos MORI, 2011; Scientific American, 2010). The trust ante may be increased when considering emotionally laden or highly personal topics, such as those associated with health and vaccines (Rousseau, Sitkin, Burt, & Camerer, 1998). Contributing to decreased levels of trust may be popular news stories (or personal perception) of researchers manipulating data, engaging in potentially unethical practices, using questionable methodologies, and withholding results (Crocker & Cooper, 2011; Kennedy, 2008; Tourney, 1992; Ziman, 1991). Further, levels of trust may be influenced by a lack of understanding of the fundamental tenets of the nature of science, such as the tentative nature of knowledge and the reliance on empirical evidence to describe the natural world (Lederman, 2006; McComas, 1998). Trust in science and scientists can influence social and political agendas (Conway & Oreskes, 2012; Gauchat, 2012) manifested by the rejection of scientific findings regardless of empirically supported rationale, and by reduced motivation to learn more about science or the related data (Gauchat, 2012). In education, trust in science and scientists may affect learning of science, as students with low levels of trust may avoid taking science courses and teachers with low levels of trust may elect not to teach science content (Ipsos MORI, 2011). Thus, there is a need to document conditions where issues of trust in science and scientists are at play. An assessment of trust in science would allow researchers to explore relationships between trust and a range of personal characteristics, such as level of education and engagement in science, and personal worldviews, such as political philosophy or religiosity. In our search of the literature, we were unable to locate a validated, reliable instrument for assessing personal levels of trust in science and scientists, particularly an instrument that is sciencegeneral and void of foci on specific science content or context. Thus, a lack of an assessment of trust in science and scientists, combined with our interests in the influence of trust on education and societal interactions, motivated us to develop and validate our Assessment of Trust in Science and Scientists. Our science domain general instrument is a unique contribution to the field that can be used across domains, contexts, and trust foci. We begin with a review of the pertinent research on trust, the associated influences and ramifications in science. We then introduce our research goals and present our instrument development process and results. Finally, we follow with a discussion of implications, applications, directions for future research, and study limitations. 76 Volume 114 (2) Review of Literature Emotional and Rational Approaches to Trust We began the process by attempting to define “trust,” which involves personal belief and/or knowledge in the dependability, honesty, aptitude, and robustness of situations, ideas, people, or persons (Simpson, Weiner, & Oxford University Press, 1989). Through our review of the literature, we became aware of the multifaceted nature of trust and the contextual and domain influences on how scholars define the construct (Blair & Stout, 2001; Mayer, Davis, & Schoorman, 1995; Romano, 2003). In our attempt to locate a pertinent theoretical framework for trust, it became increasingly obvious that the concept of trust is very complex and multifaceted, so much so that we concluded that no single framework that can adequately encapsulate trust in a meaningful way. Further, we determined it was important to examine many possible trustrelated sub-constructs that would impact perceptions of trust in science and scientists, which is why we adopted a more broad and comprehensive approach to framing trust as a construct. We embraced the position of Romano (2003), who detailed that trust consisted of multiple characteristics with no one characteristic being sufficient to define trust. We maintained the notion that trust should be considered as an attitude in response to the “. . . general question of when or whether trust is warranted, where warranted is broadly construed to include justified, well-grounded, and plausible” (McLeod, 2011, para. 2). Although McLeod’s (2011) position indicated that rational thinking is involved in situations of trust, trust is also laden with emotions and guided by feelings, morals, and faith (Dunn & Schweitzer, 2005; Romano, 2003). Trust, being multifaceted, includes several components of emotions or feelings, and rational thought (McKnight, Cummings, & Chervany, 1998). Similar to other constructs based on feelings and emotions, levels of trust may be subjective and very influential on decision-making (Isen, 2008). Using subjective trust-based decisionmaking on scientific issues in the public realm (e.g., genetically modified foods, hydraulic fracturing, vaccines, climate change) may be at odds with the objective, datadriven nature of science, and may lead to decisions that stifle the associated scientific development, research, and applications (e.g., stem cell research in the United States). As an example, the different perspectives of trust in science have a strong impact on vaccine research, development, and application. The public debate regarding the potential benefit or harm of vaccines is framed by both emotional and empirical contexts for trust in the associated science (Keelan, Pavri, Balakrishnan, & Wilson, 2010). Individuals subscribing to the perceived deleterious effects of vaccines (e.g., the proposed link to autism) tend to rely on emotions as their basis for (lack of) trust in vaccines and the associated science (Salmon et al., 2005). Often, individuals who perceive vaccines as dangerous are mistrusting of the science or scientists that present evidence contradicting their views (Omer, Salmon, Orenstein, deHart, & Halsey, 2009). Individuals using objective thinking are more likely to express high levels of trust in vaccines, relying on and applying evidence that indicates vaccines are fundamental to fighting many diseases (Larson, Cooper, Eskola, Katz, & Ratzan, 2011).Yet, as is with the nature of trust, Larson et al. (2011) reported that people are likely to use a combination of emotional and rational thinking when considering their levels of trust in vaccines and the associated science or scientists; expressing fear regarding the dangers while recognizing the potential benefits for fighting disease. The opposite may be true as well. People relying on emotions may hold high levels of trust in science or scientists, while those taking a more data-based or rational approach may hold low levels of trust science or scientists. Examples may be found in the supernatural or metaphysical (Shermer, 2002) or with alternative medical approaches such as faith healing (Barnes, Powell-Griner, McFann, & Nahin, 2004). Of course, these examples also raise issues regarding the validity or merits of the associated science. Regardless, emotionally based trust can result in different outcomes than rationally based trust. We argue that a measure of trust in science and scientists should be context and trust element neutral (i.e., should not contain emotionally laden items). The vaccine example illuminates how levels of trust or basis for trust may shift because of the context and situation (Clark & Payne, 1997; Sheppard & Sherman, 1998). Thus, we contend that due to the nature of trust, people are likely to determine their levels of trust using a combination of trust components, such as emotions and rational thought, which in some contexts are found to be highly interrelated (Clark & Payne, 1997; Cummings & Bromiley, 1996). Items that provide context or focus on a facet of trust may trigger affective reactions impacting perceptions of trust. While the interplay between subjective and objective reasoning and context for trust may be interesting to study, our goal was to create an instrument that could be used with an array of contexts while maintaining the flexibility for researchers to use our instrument with selected facets of trust. Trust in Science and Scientists School Science and Mathematics 77 Trust and Credibility The perceived credibility of the person (such as scientists) or process (such as science) may substantially influence perceptions of trust (Doney & Cannon, 1997; Mayer et al., 1995). Hardwig (1991) asserted that perceived levels of trust depend on the perception of the intellectual character and actions of the individual or institution that is the subject of trust. Credibility may also be associated with relevancy, as trust may be associated with perceived usefulness or pertinence of work to others (Bocking, 2004). Although there is debate regarding the possible conflation of credibility with trust (Romano, 2003), there are certainly indicators that people do take elements of credibility (e.g., integrity, consistency, motive) into account when determining level of trust (Butler, 1991). The influence of perceptions of credibility on levels of trust of scientists and science are at play when people personalize the potential ramifications of the work of scientists and of the process of science (Rousseau et al., 1998). For example, the credibility of science or scientists is more likely to be questioned if it involves issues of personal health. Thus, levels of trust based on credibility may come into play with medical procedures, pharmaceuticals, stem cell research, and diagnostic procedures, particularly under conditions in which our knowledge is limited, the conditions are personal, and individuals are potentially more vulnerable. We took credibility and vulnerability into consideration as essential perspectives in our development of the trust in science and scientists survey. Awareness of the potential influence of credibility and vulnerability on levels of trust associated with specific science issues motivated us to take steps to assure our items were void of potential confounding contexts. We wanted to avoid using items that could illicit responses that are contextualized and not reflective of a general level of trust in science and scientists. Trust and Epistemology Our worldviews, perceptions of knowledge, and thoughts about the plausibility of ideas can have a significant influence on our level of trust in others and in their associated institutions (Sinatra & Nadelson, 2011). Hardwig (1991) contended that epistemologies and the members of the associated communities are at play when considering levels of trust. That is, people outside of a particular community may not understand the epistemic structures and processes of that community, and may therefore lack trust for the work of the individuals within the community (Guha, Kumar, Raghavan, & Tomkins, 2004; Lederman, 2006). For example, those who do not understand the process of scientific research, the scientific community, and the nature of scientific epistemological development are not likely to trust science (Sinatra & Nadelson, 2011; Conway & Oreskes, 2012; Hardwig, 1991; Miller, 2008). Addressing the connection between trust and understanding of the nature of science, Munby (1982) and Norris (1992) promoted the notion that education should increase appreciation for the value of scientific knowledge and ideas to enhance the basis by which people determine their levels of trust in science and scientists. The potential for perceptions of scientific knowledge to influence trust in science and scientists provided justification for integrating related items into our trust instrument. We maintain that unlike the other domain-specific contexts of science we have discussed, perceptions of scientific knowledge are applicable across scientific domains and therefore should be integrated into an assessment of trust of science and scientists. Again, our goal was to design an instrument capable of assessing general levels of trust in science and scientists void of domain-specific content or contexts that may lead to contextualized perceptions of trust. However, we also recognize that there are constructs of science (e.g., scientific knowledge structures) that can be generalized to all areas of science, and therefore should be addressed in a measure of trust in science and scientists. Trust and Trustworthiness We define trustworthiness as the extent to which an individual finds a situation or people (or a person) worthy of trust; trustworthiness is thus an attribute people assign to situations or individuals and use as a basis for determining trust (Barney & Hansen, 2006). Considering trust in science and scientists is so closely related or even inseparable in this aspect, we believe it is logic to measure the trust in both science and scientists together in one instrument. It is important to draw the distinction between perceptions of trustworthiness and actual trustworthiness, as these two conditions may not be aligned (Rolin, 2002). For example, people may express a lack of trustworthiness in geoscientists with regard to climate change, but express high levels of trustworthiness in geoscientists with regard to where to drill for oil. Yet, the geoscientists are likely to be applying the same level of scientific ethical standards and morals as they work to draw conclusions in both situations resulting in a high level of actual trustworthiness meeting the standards of the scientific community. Thus, within the scientific community, science and scientists may maintain high levels of trustworthiness based on their ethics and morals regardless of their research focus, especially if the morals or ethics are considered to be consistent with scientific community norms and expectations. However, actions such as the politicization of science Trust in Science and Scientists 78 Volume 114 (2) (Conway & Oreskes, 2012; Gauchat, 2012) may substantially influence perceptions of trustworthiness of science and scientists based on the context of their work, regardless of conformity to professional ethics, morals, or standards. As Gauchat (2012) reported, worldviews and political orientation can substantially influence perspectives of trustworthiness, regardless of the actual situation. The compartmentalization of trustworthiness of geoscientists based on their scientific focus reflects the possible disconnect between perceptions of trustworthiness and actual trustworthiness. As a result, trust for science may shift based on views or knowledge of trustworthiness, particularly if trust is based on perspectives of trustworthiness related to scientific focus and not the processes of science. The potential conflation of actual and perceived trustworthiness due to factors such as political worldviews suggests an instrument assessing overall trust in science should provide takers with domain-general items. We maintain that items focused on domain-specific contexts (e.g., climate change) may trigger perspectives of trustworthiness and potentially promote responses that are not consistent with general trust in science and scientist. The influence of politics on perceptions of trustworthiness of science has been accompanied by the complexity created by the merging of science with other domains such as economics, geography, and history (Ludwig, Mangel, & Handdad, 2001). Similarly, influencing trust in science is the potential embedding of science within structures that have established issues of trust, such as private business and industry and governmental agencies (Gallup, 2013). The merging of science with other domains has the potential to conflate expertise, epistemologies, and knowledge, and to create opportunity for conflict of ideologies, values, and priorities (Cech & Leonard, 2001)—all of which may influence how individuals evaluate the trustworthiness of science and scientists. The potential for variations in perceptions of trustworthiness due to individual characteristics and experience, and conditions external to science (e.g., economic and geographic influences), guided our development of our trust in science assessment. We considered the potential compartmentalization of trustworthiness based on specific conditions, and determined that a general measure of trust in science should contain domain-general items. Yet, the potential compartmentalization provides justification for using a domain-general trust in science instrument in conjunction with measures of personal perspectives (e.g., political orientation) as these are potentially critical indicators of level of trust of science and scientists. Likewise, the potential conflation of science with other domains and the institutions in which science resides provide motivation for developing a general trust in science instrument and using the tool with other measures to determine how confounding variables are related and under what conditions trust in science and scientists may shift. Measuring Trust The extant literature reporting measures of trust in science and scientists include one item from a questionnaire (Gauchat, 2012) or have been part of opinion polls (Ipsos MORI, 2011; Scientific American, 2010). The lack of an extant instrument that could be used by the science education research community motivated our survey development. Based on evidence that people tend to use combinations of trust components or elements when determining trust (Butler, 1991; Romano, 2003), we argue that items aligned with specific elements of trust may be responded to using other elements of trust or a combination of elements that are not representative of the targeted facet of trust. Therefore, facet-specific trust items (e.g., based on emotions, rational thought, or credibility) are not likely to produce results truly representative of perspectives of trust from the targeted facet of trust, and therefore, are not particularly useful. Supporting our position is research on trust instrument development that concludes measures of trust change depending on the context, and the identifiable facets shift with the research process and focus (Butler, 1991; Clark & Payne, 1997; Currall & Judge, 1995; McAllister, 1995). Our position is supported by Romano (2003), who reported that attempts to segregate items by elements of trust produced results that were reflective of a general measure of trust. Thus, we determined that a measure of trust in science and scientists should have a domaingeneral perspective with efforts taken to assure items are neutral in terms of content or reliance on a specific component of trust. It is the responsibility of the researchers who use our instrument to identify the context or facet of trust in which they are interested in assessing trust, and interpret the outcomes of levels of trust according to the focus of their research. Implications for Trust in Science As scientific developments and findings about issues such as climate change and genetically modified foods move from the science community into the general society, the general public is exposed to situations in which they may be asked to make decisions associated with the issues (Litva et al., 2002). If these scientific developments are extremely complex (e.g., hydraulic fracturing, commonly referred to as “fracking”) and difficult to comprehend (e.g., stem cells), then people may rely on their feelings of Trust in Science and Scientists School Science and Mathematics 79 trust when responding to scientific issues (Ipsos MORI, 2011; Scientific American, 2010). As we discussed earlier, trust is an attitude and therefore may be influenced by emotions that may impact engagement in objective evaluation of issues, potentially leading to decision-making based on subjective perceptions of trust. As a result, subjective decision-making (using trust as the motivation) regarding science or scientists could lead to decreased financial and societal support for science that may impede potentially highly beneficial or necessary developments in science (e.g., alternative energy) (Bozeman & Sarewitz, 2005). Trust in science and scientists may in part be the responsibility of the scientific community (Lidskog, 1996; Turney, 1996), as science is part of the greater society. One of the principle tenets of the nature of science is that science is culturally embedded such that science cannot be extracted from influences and constraints of culture (Lederman, 2006; McComas, 1998). Further, the tenets of the nature of science provide recognition of science as a human endeavor (Lederman, 2006; McComas, 1998), which provides additional support for the potential interest in public trust in science and scientists. Given the cultural grounding of science, trust in science is likely to be influenced by a number of variables, including scientists themselves. Turney (1996) argued that scientists can and should influence the levels of trust in their work by educating others about their work and the processes of science. Similarly, Millstone and Zwaneberg (2000) contended that the general public, which is the main consumer of science when it enters the societal realm, is likely to be mistrusting of science because of a lack of understanding of science due to a lack of communication by scientists. Our instrument can help inform endeavors aimed at increasing understanding of science by providing a means to assess shifts in trust because of educational and awareness interventions that increase understanding of science. The perceptions of scientists and science as being untrustworthy may also result from a variety of media that are unrelated to the actual practices of science (Tourney, 1992). The possible influences of fictional portrayals of scientists on television, in movies, through the internet, in books, and other media may lead to strong, yet potentially inaccurate, understandings of scientists (Rahm & Charbonneau, 1997; Wyer, Schneider, Nassar-McMillan, & Oliver-Hoyo, 2010). Finson (2010) reported that the stereotyping of scientists has persisted longitudinally, indicating that issues of trust in scientists may be a perennial issue due to media influence. The potential public perceptions of scientists’ influence on societal attitudes toward science provides motivation for determining trust as a possible indicator of perceptions of scientists and the related attitude toward science (Finson, 2001; Finson, Riggs, & Jesunathadas, 1999; Ipsos MORI, 2011). Thus, we contend that the association between perceptions of scientists and attitudes toward science may be manifested in levels of trust of both science and scientists. We acknowledge science as a methodology and scientists as the people doing science. However, we argue that when most people encounter the term “science,” they do not think methodology but rather the body of knowledge about the natural world that includes domains such as chemistry, biology, physics, and geoscience, the domains that are labeled as “science” in school curriculum. Similarly, we contend that when people encounter the term scientist, they think of those doing research in domains such as chemistry, biology, physics, and geoscience. Therefore, we argue that there is rationale for combining these in an instrument assessing trust, as most people consider science and scientists to be coupled because of exposure to the portrayal of science in schools. Further, there is justification for the combining science and scientist based on the grouping that has taken place in public opinion research (Masci, 2009) and other research reported in the literature (e.g., Jones, Howe, & Rua, 2000). In summary, what we have found is that trust in science and scientists can have profound influence on policy, funding, and even legislation, and yet trust is a multifaceted construct that includes aspects of emotions, perceptions of credibility, perceptions of trustworthiness, world views, and knowledge. Although trust can be conceived generally, interpreting the particular influences or indicators of trust require a lens through which these facets are considered, and personal characteristics of study participants are likely necessary to explain levels of trust. Therefore, there is justification for assessing trust in science and scientists on a general level and using personal data (e.g., levels of science education or political orientation) to provide possible explanations for levels of trust. Methodology Research Goals Informed by the literature, news of public perceptions of science, and by knowledge of working with postsecondary students, we set out to create an instrument that could be used to assess students’ trust in science and the work of scientists. Specifically, the goals of our research were: • Identify domain-general aspects of the processes of science or the work of scientists that may involve aspects of trust, Trust in Science and Scientists 80 Volume 114 (2) • Create a set of items that assess a domain-general level of trust in science and scientists, • Establish the validity of our instrument with experts, using their feedback to refine our items, • Field-test our survey to establish the instrument’s reliability. Instrument Development Initial item formation. This work was conducted by an interdisciplinary research team, made up of six science, technology, engineering, and math (STEM) faculty (a geoscientist, a chemist, a biologist, a biochemist, a sociologist, and a mathematician) who were all engaged in bench or field research in their STEM fields. The team also included two education faculty who were engaged in STEM education research. The team was assembled as part of a systemic STEM initiative at a metropolitan research university during the 2011–2012 academic year and met bi-weekly to collaborate on STEM education research. We began our instrument development with discussions of aspects of science and the work of scientists that people might question in terms of trust. For example, from the literature, it is apparent that there are misconceptions of several aspects of science, such as theories of knowledge structures and the tentative nature of science (Lederman, 2006; McComas, 1998; Miller, 2008). The potential lack of understanding of knowledge structures and the tentative nature of science are common in society and may be manifested as lack of trust in science and scientists (Miller, 2008). Thus, we determined that some items in our instrument should focus on aspects of the nature of science, particularly the tentative nature of knowledge in science. In addition to the literature and general misconceptions of the nature of science, we used elements of issues of trust in science and scientists that are associated with public communication or are commonly presented in the in the news, (e.g., issues of trust in relationship to research techniques, processes, or integrity). As a team, we discussed these aspects of trust to determine the possible root(s) of trust in science and scientists. For example, we hypothesized that people who have trust issues with climate change science have concerns that scientists may be withholding data, falsifying data, or not considering all available evidence. Based on our conversations, knowledge of issues of trust in science found in the news, interactions with the public, and perceptions based on exchanges with students, we formed items such as: “Scientists ignore evidence that contradicts their work” and “We should trust that scientists are being honest in their work.” Using this process of examining reactions to science and the focus on the work of scientists, we generated a list of 21 forward- and reversed-phrased trust items. We mixed the direction of the questions not as dyads, but to prevent response sets (e.g., all “Strongly Agree”). The items are presented on a five-point Likert scale ranging from “1” representing “Strongly Disagree” to “5” representing “Strongly Agree.” We selected a five-point scale as it provides better discrimination of answers and less potential bias than a four-point scale (Garland, 1991). Establishing validity. As a team, we critically examined each of the items to assure they had a primary focus on trust in science and scientists. We also took steps to assure our items maintained a generalized emphasis on the processes and tenets of science related to elements of trust in science and scientists. As a team, we read each item and then discussed both the merits and justification for the item, with each member of the team providing input. The various paradigms and knowledge levels of the team members increased the assurance that our items were not only focused on trust, but also reflected the general issues of trust with respect to the tenets of science and work of scientists. Once we completed our drafting of items and internal review, we shared our instrument with five external researchers with expertise in measurement, science education, and assessing attitudes, and asked them to provide feedback regarding our items and the extent to which they reflected a focus on trust in science and scientists. Based on the feedback from our five external researchers, we made minor adjustments to the language of our items to further emphasize the focus on trust. All five of the external reviewers responded that the items on the questionnaire were reflective of trust in science or scientists. The experts also indicated that they thought adults would be able to effectively respond to the items using their personal references of trust in science and scientists without needing for further information or clarification. Therefore, we feel that we adequately established and confirmed the construct and content validity of our instrument. Field testing—round one. We began field-testing our instrument by administering it to 75 undergraduate college students enrolled in an introductory geoscience course. Our convenience sample was drawn from this course, which was designed for non-science majors to allow students from other disciplines to fulfill a core course requirement in the natural sciences. Of the 60 students who completed the demographics and survey instruments, 75% were non-science majors, 10% were natural science majors, 7% were engineering majors, 5% were health profession majors, and the remaining 3% were undeclared. The students had an average of 2.22 science courses (SD = 1.36), an average of .28 mathematics courses (SD = .83), Trust in Science and Scientists School Science and Mathematics 81 and an average of 2.13 English courses (SD = 1.10). The sample was on average 22.81 years old (SD = 4.98) and was composed of 57% females and 43% males. We administered a paper version of our instrument along with a general demographic survey to the students at the beginning of one of their course meetings. Students required approximately 10 minutes to complete the survey. Following the data collection, we entered the responses into SPSS (IBM Corp, 2011) for conditioning (reverse coding answers) and analysis. A reliability analysis of the results revealed a Cronbach’s alpha of .84, which we interpret as a good to very good level of reliability. The 60 participants had an overall average of trust of 3.57 (SD = .43) on a five-point Likert scale, which indicates a slightly above an ambivalence level of trust in science. The corrected item total correlations analysis revealed the lowest levels of contributions to the reliability for the items stating: “You have to know the right people to really find out what scientists know.” and “I trust scientists can find solutions to our major technological problems.” Using the item analysis as a guide, we decided to restructure these items. For example, we restructured the first item in the previous sentence to read, “Scientists intentionally keep their work secret” as this statement captures the essence of the item while simplifying the language. In addition to the restructuring of a few items, we also made some minor substitution in the language of two other items to increase their clarity. Field testing—round two. Once we completed the modifications of the items in our instrument, we again field-tested it with another convenience sample of undergraduates who were enrolled in a range of undergraduate level courses. Of the 314 undergraduates who returned surveys, 301 completed all items (or nearly all—with a few having skipped one or two items). The 301 students were on average 23.86 years old (SD = 7.11) and were 53.8% females and 46.2% males. The students had an average of 3.06 years of college (SD = 1.91) and had taken an average of 4.46 science courses (SD = 4.96). The students held an average level of religiosity of 5.05 (SD = 3.36) on a 10-point scale with 1 being “low commitment” and 10 being “high commitment.” Similarly, on a 10-point scale of political orientation scale, with 1 being liberal and 10 being conservative, our participants had an average political orientation of 5.27 (SD = 2.56). Our reliability analysis of our revised instrument revealed a Cronbach’s alpha of .86, which we interpret as a good to very good level of reliability. Our item analysis using the corrected item total correlations revealed the lowest levels of contributions to the reliability coming from the item stating, “It’s hard for the everyday person to access scientists’ findings” and similar to our first round the item, “I trust scientists can find solutions to our major technological problems.” Restructuring of a few items did not have a major influence on our instrument performance as a whole. The means and standard deviations (SDs) of the items were fairly consistent, indicating that the participants were fairly consistent in their responses and perceived the forward and reverse items effectively. The item means, SDs, and corrected item total correlations are presented in Table 1. As we examined the corrected item total correlations, which ranged from .34 to .63, we determined that all of the items in our instrument contributed positively to the overall reliability of our instrument (see Table 1) supporting the internal consistency of our instrument (Field, 2005). Further, the corrected item total correlations were relatively moderate, indicating that there are no instances of over-correlation, which suggests that our participants likely answered the items using various elements or combination of trust facets to form their perception of trust. The 301 participants had an overall average of trust of 3.53 (SD = .47) on a five-point Likert scale, which at first glance may indicate a level of trust between ambivalence and somewhat trusting of science and scientists. However, upon further analysis we can see that there are differences Table 1 Item Means, Standard Deviations, and Inter-Item Correlations for the Inventory of Trust in Science and Scientists Item M SD Inter-Item Correlation Trust_1 3.60 1.05 .42 Trust_2 3.40 1.02 .52 Trust_3 3.92 .87 .54 Trust_4 3.23 .95 .42 Trust_5 3.27 1.01 .37 Trust_6 3.74 .92 .48 Trust_7 3.65 1.00 .41 Trust_8 3.15 .95 .36 Trust_9 3.47 .88 .47 Trust_10 3.54 .85 .45 Trust_11 3.48 .92 .34 Trust_12 3.38 .82 .48 Trust_13 3.56 .99 .37 Trust_14 3.81 .85 .37 Trust_15 3.80 .86 .42 Trust_16 3.76 .82 .36 Trust_17 3.62 .91 .63 Trust_18 3.50 .88 .40 Trust_19 3.40 .91 .57 Trust_20 3.27 .91 .46 Trust_21 3.90 .88 .52 Trust in Science and Scientists 82 Volume 114 (2) in response based on number of college-level science classes, religiosity, and political philosophy (see Table 2). For example, for students with less than two college-level science courses (N = 92) held an average trust in science of 3.43 (SD = .40), while their peers with eight or more classes (N = 66) held an average trust in science of 3.73 (SD = .49). Our results indicate that as students take more science courses, they tend to have a higher level of trust in science. Similar results were found for the ends of the political philosophy and religiosity scales. Thus, our analysis revealed a differential response based on education, worldview, and political perspective, which is consistent with the literature. As a further verification of the validity of our instrument, we conducted a correlational analysis using level of trust and our measures of personal characteristics that are documented to be associated with trust in science and scientists (see Table 3). Our analysis revealed that trust in science was related to religiosity (p < .01), such that as religiosity increases, trust decreases. Similarly, we found a significant relationship with political philosophy (p < .05), such that liberals had higher trust in science and scientists and conservatives had lower trust in science and scientists. We also found a positive correlation between trust in science and scientists and the number of college-level science classes (p < .01), and the years of college (p < .05). These findings suggest that there is a differential response to our items based on our participants’ worldview (religiosity), political orientation, knowledge of science, and general level of education, all potential influences on levels of trust in science and scientists. The associations we found further confirm the validity of our instrument. Discussion We set out to develop the first (to our knowledge) assessment of the level of trust in science and scientists with established validity and reliability. Using our knowledge of trust from the literature, personal experience with students and the general public, stories in the news and media, and feedback from external experts, we generated a set of items and vetted them with experts in measurement and science education. After our first round of field testing, we made slight modifications to our instrument and distributed it to a diverse group of university students. The results of our analysis revealed a Cronbach’s alpha of .86, indicating a good level of instrument reliability. Our comparison of composite scores of trust with years of college and number of college-level classes indicated positive relationships such that as years of college increased and the number of science classes increased trust in science increased. We attribute the positive relationships to the likelihood that people with more education are more likely to have a deeper understanding of science and work of scientists, and are more likely to have engaged in critical examinations of scientific issues. Similarly, more science courses would lead to a deeper level of understanding of scientific tenets and processes through interactions with scientists, engagement in doing science, and reviewing the work of scientists. We speculate that years of college and number of science courses are likely to have provided students with experiences leading to a deeper understanding of science that confirms consistency in science and supports perceptions of merit for the work Table 2 Average Levels of Trust for Subgroups Based on Number of Science Classes, Religiosity, and Political Philosophy Group N Level of Trust M (SD) Greater than seven college-level science courses 66 3.73 (.49) Less than two college-level science courses 92 3.43 (.40) Religious commitment less than two 77 3.74 (.48) Religious commitment greater than eight 66 3.33 (.49) Political orientation less than two (liberal) 33 3.50 (.53) Political orientation greater than eight (conservative) 33 3.37 (.52) Table 3 Correlations Between Trust and Participant Personal Characteristics Trust in Science and Scientist Religious Commitment Political Philosophy Number of College Level Science Courses Years of College Trust in science and scientist — −.33** −.14* .24** .14* Religious commitment — .41 −.06 −.02 Political philosophy — −.07 −.07 Number of college Level science courses — .57 Years of college — * p < .05; ** p < .01. Trust in Science and Scientists School Science and Mathematics 83 of scientists, which in turn lead to deeper levels of trust. The relationships between education and trust in science provided further support for the validity of our instrument, and supported the notion that increased education is likely to shift trust in science, resulting in more engagement in science. The link between trust in science and scientists and engagement in learning science provided further justification for the need to address student trust in science. Our analysis also revealed inverse relationships with political orientation and religiosity. More conservative students and students with higher religiosity are likely to hold lower levels of trust in science. The consistency of this observation with the literature (Conway & Oreskes, 2012; Gauchat, 2012; Miller, 2008) further supported the validity of the items in our instrument and underscores the need to determine students’ level of trust as they are being taught science because levels of trust may influence their engagement in learning. Our results suggest that our instrument would be useful and appropriate for assessing the level of trust of college students and perhaps adults in the general population. Since our instrument is the first of its kind, it may be useful for a range of applications extending beyond those indicated in our study. Further, since we took strides to assure that our instrument was a measure of trust in science and scientists on a general level, it lends itself well to use in specific contexts, with specific content, or with a focus on a specific facet of trust. For example, it may be used in a course focused on the nature of science, or with curriculum associated with the philosophy or history of science. Of particular interest may be the assessment of trust of non-science majors engaging in a foundational curriculum designed to increase their knowledge and appreciation of science. We anticipate that our instrument may also be used as a means of assessing the impact of interventions intended to increase confidence and understanding of science and scientists, and the subsequent influence on their trust in science. We encourage researchers to explore ways in which our instrument can inform intervention studies and to seek novel situations to study various aspects of trust in science and scientists. Our instrument will be useful in research as another indicator of students’ attitude toward science in general. In research examining students’ ideas about controversial aspects of science such as biological evolution, climate change, genetically modified foods, or vaccines, gauging students’ trust in science may provide additional insight to explain their perceptions. Our instrument will also be useful for longitudinal studies to determine how trust may shift over time with education or conditions, particularly once we modify the instrument to make it accessible and appropriate for K-12 students. The important role of trust in science and peoples’ perceptions of scientists is apparent. Trust in science and scientists can impact learning, decision-making on science-related issues that enter the public realm, and how the work of science and scientists are viewed as a community. Being able to assess trust may shed insight into the perspectives of learners, the general public, and specific groups outside the scientific community, which could be useful for curriculum development, public relations, and communication. Limitations and Directions for Further Research As with any study involving self-report, our study is limited by the nature of the data collection process. However, given the personal nature of trust as an attitude and the efforts we took to develop an instrument that is context- and content-neutral, we anticipated that the participants were honest in their responses. Our data are consistent with the literature. However, delving deeper into how students go about determining trust and how trust might shift with contexts or content are excellent directions for future research. Our instrument provides an ideal mechanism for conducting this line of research. Another potential limitation of our study is our sample. Our participants were from a single institution. However, they were drawn from a broad cross section of courses and represented a wide range of science knowledge and interests. Because of the diversity of backgrounds (e.g., science majors and non-science majors), we feel that we had the variation needed in our sampling to effectively establish the reliability of our instrument. Further, the associations we found with education, religiosity, and political orientation provide additional support for the instrument validity. Regional and institutional variations in levels of trust in science and scientists are certainly excellent directions for future research. A similar limitation of our instrument is the focus on postsecondary students during its development. Appropriately modified versions of our instrument may also be useful in K-12 education to assess students trust in science and scientists. To assure the instrument is appropriate for the K-12 education level, we would first need to solicit input on the instrument from K-12 teachers prior to piloting its use with elementary through secondary students. It is possible that some of the instrument language would need to be simplified to align with the knowledge and experiences of K-12 students. Vetting the instrument and adjusting it for use with K-12 students is an excellent direction for future research. Trust in Science and Scientists 84 Volume 114 (2) Conclusion Trust in science and scientists can have profound implications at both personal and societal levels. Yet, the seemingly simple construct of trust is deceivingly complex to define and explain. Our goal was to develop the first instrument we are aware of that uses multiple items to assess trust in science and scientists on a general level. The results provided by the use of our instrument could give researchers insight into the relationship between participants’ levels of trust based on context, investigative lens, participant characteristics, and other variables as efforts are made to determine how trust may shift over time or based on conditions. We hope others will find our instrument useful, and will provide feedback on ways to increase its effectiveness and broaden its use. References Barnes, P. M., Powell-Griner, E., McFann, K., & Nahin, R. L. (2004). Complementary and alternative medicine use among adults: United States, 2002. Advance Data, 343, 1–19. Barney, J. B., & Hansen, M. H. (2006). Trustworthiness as a source of competitive advantage. Strategic Management Journal, 15(S1), 175–190. Blair, M. M., & Stout, L. A. (2001). Trust, trustworthiness, and the behavioral foundations of corporate law. University of Pennsylvania Law Review, 149(6), 1735–1810. Bocking, S. (2004). Natures experts: Science, politics, and the environment. Piscataway, NJ: Rutgers University Press. Bozeman, B., & Sarewitz, D. (2005). Public values and public failure in US science policy. Science and Public Policy, 32(2), 119–136. Butler, J. K., Jr (1991). Toward understanding and measuring conditions of trust: Evolution of a conditions of trust inventory. Journal of Management, 17, 643–663. Cech, T. R., & Leonard, J. S. (2001). Conflicts of interest-moving beyond disclosure. Science, 291(5506), 989. Clark, M. C., & Payne, R. L. (1997). The nature and structure of workers’ trust in management. Journal of Organizational Behavior, 18(3), 205–224. Conway, E. M., & Oreskes, N. (2012). Why conservatives turned against science: Remember when environmental protection was a bipartisan effort? Chronicle of Higher Education. Retrieved from http:// chronicle.com/article/The-Conservative-Turn-Against/135488/ Crocker, J., & Cooper, M. L. (2011). Addressing scientific fraud. Science, 334(6060), 1182. Cummings, L. L., & Bromiley, P. (1996). The Organizational Trust Inventory (OTI): Development and validation. In R. Kramer & T. Tyler (Eds.), Trust in organizations (pp. 302–330). Thousand Oaks, CA: Sage. Currall, S. C., & Judge, T. A. (1995). Measuring trust between organizational boundary role persons. Organizational Behavior and Human Decision Processes, 64, 151–170. Doney, P. M., & Cannon, J. P. (1997). An examination of the nature of trust in buyer-seller relationships. Journal of Marketing, 61(2), 35–51. Dunn, J., & Schweitzer, M. (2005). Feeling and believing: The influence of emotion on trust. Journal of Personality and Social Psychology, 88(6), 736–748. Field, A. (2005). Discovering statistics using SPSS (2nd ed.). London: Sage. Finson, K. D. (2001). Investigating preservice elementary teachers’ selfefficacy relative to self-image as a science teacher. Journal of Elementary Science Education, 13(1), 31–42. Finson, K. D. (2010). Drawing a scientist: What we do and do not know after fifty years of drawings. School Science and Mathematics, 102(7), 335–345. Finson, K. D., Riggs, I. M., & Jesunathadas, J. (1999). The relationship of science teaching self-efficacy and outcome expectancy to the Draw-aScience-Teacher-Teaching Checklist. Paper presented at the annual international meeting of the Association for the Education of Teachers of Science, Austin, TX. Gallup (2013). Trust in government. Retrieved from http://www.gallup.com/ poll/5392/trust-government.aspx Garland, R. (1991). The mid-point on a rating scale: Is it desirable? Marketing Bulletin, 2(1), 66–70. Gauchat, G. (2012). Politicization of science in the public sphere: A study of public trust in United States, 1974 to 2010. American Sociological Review, 77(2), 167–187. Guha, R., Kumar, R., Raghavan, P., & Tomkins, A. (2004). Propagation of trust and distrust. In Proceedings of the 13th international conference on World Wide Web, 17–22 May, New York (pp. 403–412). New York: ACM. Hardwig, J. (1991). The role of trust in knowledge. Journal of Philosophy, 88(12), 693–708. IBM Corp (2011). IBM SPSS statistics for Windows, Ver.20.0. New York: IBM Corp. Ipsos MORI (2011). Public attitudes to science 2011. Department for Business, Innovation and Skills. Retrieved from http://www.ipsos-mori.com/ Assets/Docs/Polls/sri-pas-2011-summary-report.pdf Isen, A. (2008). Some ways in which positive affect influences decisional making and problem solving. In M. Lewis, J. M. Haviland-Jones, & L. Feldman Barrett (Eds.), Handbook of emotions (pp. 548–573). New York: Guilford Press. Jones, M. G., Howe, A., & Rua, M. J. (2000). Gender differences in students’ experiences, interests, and attitudes toward science and scientists. Science Education, 84(2), 180–192. Keelan, J., Pavri, V., Balakrishnan, R., & Wilson, K. (2010). An analysis of the Human Papilloma Virus vaccine debate on MySpace blogs. Vaccine, 28(6), 1535–1540. Kennedy, D. (2008). Science and its discontents: An evolutionary tale. Research and Occasional Paper Series. Center for Studies in Higher Education, University of California, Berkeley. Larson, H. J., Cooper, L. Z., Eskola, J., Katz, S. L., & Ratzan, S. (2011). Addressing the vaccine confidence gap. The Lancet, 378(9790), 526–535. Lederman, N. G. (2006). Students’ and teachers’ conceptions of the nature of science: A review of the research. Journal of Research in Science Teaching, 29(4), 331–359. Lidskog, R. (1996). In science we trust? On the relation between scientific knowledge, risk consciousness and public trust. Acta Sociologica., 39, 31–56. Litva, A., Coast, J., Donovan, J., Eyles, J., Shepherd, M., Tacchi, J. . . . Morgan, K. (2002). “The public is too subjective”: Public involvement at different levels of health-care decision making. Social Science & Medicine, 54(12), 1825–1837. Ludwig, D., Mangel, M., & Handdad, B. (2001). Ecology, conservation, and public policy. Annual Review of Ecology and Systems, 32, 481–517. Masci, D. (2009). Public opinion on religion and science in the United States. The Pew Research Center. Retrieved from http://www.pewforum.org/ science-and-bioethics/public-opinion-on-religion-and-science-in-theunited-states.aspx Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995). An integrative model of organizational trust. Academy of Management Review, 20(3), 709–734. McAllister, D. J. (1995). Affect- and cognition-based trust as foundations for interpersonal cooperation in organizations. Academy of Management Journal, 38, 24–59. McComas, W. F., Ed. (1998). The nature of science in science education– rationales and strategies. Dordrecht, The Netherlands: Kluwer Academic Publishers. McKnight, D. H., Cummings, L. L., & Chervany, N. L. (1998). Initial trust formation in new organizational relationships. Academy of Management review, 23(3), 473–490. Trust in Science and Scientists School Science and Mathematics 85 McLeod, C. (2011). Trust. In E. N. Zalta (Ed.), The Stanford encyclopedia of philosophy. Stanford, CA: Center for the Study of Language and Information, Stanford University. Retrieved from http://plato.stanford.edu/ archives/spr2011/entries/trust/ Miller, K. R. (2008). Only a theory: Evolution and the battle for America’s soul. New York: Penguin Group. Millstone, E., & Zwaneberg, P. (2000). A crisis of trust: For science, scientists or for institutions? Nature Medicine, 6, 1307–1308. Munby, H. (1982). The place of teachers’ beliefs in research on teacher thinking and decision making, and an alternative methodology. Instructional Science, 11, 201–225. Norris, S. P. (1992). Practical reasoning in the production of scientific knowledge. In R. Duschl, & R. Hamilton (Eds.), Philosophy of science, cognitive psychology, and educational theory and practice (pp. 195–225). Albany: State University of New York Press. Omer, S. B., Salmon, D. A., Orenstein, W. A., Dehart, M. P., & Halsey, N. (2009). Vaccine refusal, mandatory immunization, and the risks of vaccine-preventable diseases. New England Journal of Medicine, 360(19), 1981–1988. Rahm, J., & Charbonneau, P. (1997). Probing stereotypes through students’ drawings of scientists. American Journal of Physics, 65(8), 774–778. Rolin, K. (2002). Gender and trust in science. Hypatia, 17(4), 95–118. Romano, D. M. (2003). The nature of trust: Conceptual and operational clarification (Doctoral dissertation, Louisiana State University). Retrieved from http://etd.lsu.edu/docs/available/etd-0130103-070613/unrestricted/ Romano_dis.pdf Rousseau, D. M., Sitkin, S. B., Burt, R. S., & Camerer, C. (1998). Not so different after all: A cross-discipline view of trust. The Academy of Management Review, 23(3), 393–404. Salmon, D. A., Moulton, L. H., Omer, S. B., Dehart, M. P., Stokley, S., & Halsey, N. A. (2005). Factors associated with refusal of childhood vaccines among parents of school-aged children: A case-control study. Archives of Pediatrics & Adolescent Medicine, 159(5), 470–476. Scientific American (2010). In Science We Trust: Poll results on how you feel about science. Scientific American, 303, 56–59. Sheppard, B. H., & Sherman, D. M. (1998). The grammars of trust: A model and general implications. Academy of Management Review, 23(3), 422– 437. Shermer, M. (2002). Why people believe weird things: Pseudoscience, superstition, and other confusions of our time. New York: Freeman. Simpson, J. A., Weiner, E. S. C., & Oxford University Press (1989). The Oxford English dictionary. Oxford: Clarendon Press. Sinatra, G. M., & Nadelson, L. (2011). Science and religion. Ontologically different epistemologies? In R. S. Taylor & M. Ferrari (Eds.), Epistemology and science education. Understanding the evolution vs. Intelligent Design controversy (pp. 173–193). New York: Routledge. Tourney, C. P. (1992). The moral character of mad scientists: A cultural critique of science. Science, Technology, & Human Values, 17, 411–437. Turney, J. (1996). Public understanding of science. The Lancet, 347(9008), 1087–1090. Wyer, M., Schneider, J., Nassar-McMillan, S., & Oliver-Hoyo, M. (2010). Capturing stereotypes: Developing a scale to explore US college students’ images of science and scientists. International Journal of Gender, Science and Technology, 2(3), 381–415. Ziman, J. M. (1991). Reliable knowledge: An exploration of the grounds for belief of science. New York: Cambridge University Press. Appendix Trust in Science and Scientists Inventory Directions: Rank your level of agreement to each of these statements on the scale provided. Strongly Disagree Disagree Neutral Agree Strongly Agree 1. When scientists change their mind about a scientific idea it diminishes my trust in their work.* 2. Scientists ignore evidence that contradicts their work.* 3. Scientific theories are weak explanations.* 4. Scientists intentionally keep their work secret.* 5. We can trust scientists to share their discoveries even if they don’t like their findings. 6. Scientists don’t value the ideas of others.* 7. I trust that the work of scientists to make life better for people. 8. Scientists don’t care if laypersons understand their work.* 9. We should trust the work of scientists. 10. We should trust that scientists are being honest in their work. 11. We should trust that scientists are being ethical in their work. 12. Scientific theories are trustworthy. 13. When scientists form a hypothesis they are just guessing.* 14. People who understand science more have more trust in science. 15. We can trust science to find the answers that explain the natural world. 16. I trust scientists can find solutions to our major technological problems. 17. We cannot trust scientists because they are biased in their perspectives.* 18. Scientist will protect each other even when they are wrong.* 19. We cannot trust scientists to consider ideas that contradict their own.* 20. Today’s scientists will sacrifice the well being of others to advance their research.* 21. We cannot trust science because it moves too slowly.* * Reverse coded item. Trust in Science and Scientists 86 Volume 114 (2)