The theory of planned behavior is a well-established theory of human behavior, which relies on the premise that a person’s intention to engage in a behavior is the single best predictor of whether that person will in fact engage in that behavior. According to the theory, intentions, and consequently, behavior are influenced by three sets of variables: attitude towards the behavior, subjective norms, and perceived behavioral control ( Ajzen, 1991). Attitude towards the behavior is the degree to which performance of the behavior is positively or negatively valued by an individual. This attitude is in turn determined by behavioral beliefs—the subjective probability that the behavior will produce a given outcome—that link the behavior of interest to expected outcomes. Behavioral beliefs are based on personal experience, information sources and inferences. Subjective norms reflect an individual’s perception of social pressures to engage or not to engage in a behavior. This variable is linked to the degree to which certain groups of people approve or disapprove of the individual performing the specific behavior and how social pressure informs subjective perception about the particular behavior. Perceived behavioral control refers to a person’s perception of her ability to perform a given behavior. This variable is linked to the perceived ability to engage in a behavior and to the ability to do so effectively ( Ajzen, 1991).

Planned behavior theory in particular has provided a fertile framework for empirical research in scientists’ participation in public engagement activities ( Besley et al., 2013; Dudo et al., 2014; Dudo & Besley, 2016; Poliakoff & Webb, 2007). Poliakoff & Webb (2007: 254) have applied this approach to investigate the determinants of scientists’ intentions to participate in public engagement activities. Based on an ‘augmented’ version of the theory of planned behavior, the two authors show that scientists’ intentions are determined by past experiences with public engagement, attitudes towards engaging with the public, perceived control that scientists can exercise on public engagement activities, and beliefs about colleagues’ participation in public engagement. Poliakoff and Webb also conclude that career recognition, availability of time and other constraints do not significantly predict intentions to engage in public engagement of science activities. Using data from two large surveys of scientists from the United States and the United Kingdom, Besley et al. (2013) found that a personal commitment to the public good and feelings of personal efficacy and professional obligation can predict scientists’ involvement in outreach activities. Dudo et al. (2014) found that predictors of nanoscientists’ involvement in science communication include perceiving public communication as important for the welfare of society, seeing professional benefits from generating publicity about their research, and spending more time using online tools. These results were in great part replicated by Aykurt (2016), who conducted a similar study among nonscientists in Denmark. Interestingly, the study also found that a lack of time is an obstacle to engaging in science communication using online tools ( Aykurt, 2016: 44). Finally, in researching how scientists prioritize science communication objectives, Dudo & Besley (2016) found that informing the public, exciting the public, building public trust about science, and defending science from public misinformation are objectives that received the largest support among study participants. The fact that these objectives are important to scientists certainly has implication to their intention to engage in science communication: scientists may be more willing to engage if they perceive that science communication efforts prioritize the right objectives.

With this framework in mind, we analyze how the three variables deployed by the theory of planned behavior contribute to our understanding of scientists’ intentions to participate in science communication activities. Our analysis is based on qualitative data that we gathered in conjunction with a large survey of scientists conducting basic research in the biomedical field in Italy, the United Kingdom, and the United States ¹ . The survey focused on the attitudes and beliefs of scientists towards policies that could incentive the types of basic research with the highest likelihood of creating public health benefits ² . While the survey did not focus directly on science communication, many study participants shared their perception of science communication activities in a Comments section that they could fill out after completing the survey. The section reads as follows: ³

Participants were recruited by email, and, after consenting to participating, asked to confirm their status as basic researchers and identify their position/role within the respective organizations. Respondents could indicate their geographical location, but not their institution. They could skip any questions, including the Comments field ⁴ . Out of the 7,786 invitations that were sent between August 24, 2015 and October 10, 2015, 885 recipients filled out the entire survey and, among them, 145 participants filled out the Comments field ⁵ . Comments are presented, along with all responses to the entire survey, in the Final report in the Data availability section. The majority of 145 respondents were male (64%) and PIs (71%). In total 60% were based in the US (50 in Los Angeles-San Diego and 34 in New York City), with the remaining split between the UK (36) and Italy (20) ( Table 1).

All Comments were imported into MAXQDA (ver. 12.1; http://www.maxqda.com/), a qualitative data analysis software, to be analyzed. We then conducted a three-step analysis the data. In Step 1, we coded segments ‘ in vivo’ based on the ‘first-step coding,’ a code system that we had used to analyze qualitative data in a previous study exploring similar issues ( Boggio et al., 2016). The ‘first-step coding’ schema is available as Supplementary File 1. During Step 1, we discarded 12 of the 145 answers because respondents either commented on the survey design or merely thanked us for conducting the study. In Step 2, we reviewed the results of the first-step coding to identify clusters of data that could lead to further insights when analyzed in conjunction with the theory of planned behavior. In Step 3, we proceed to analyze the data clusters, identified as a result of Step 2, based on the three main parameters of the theory of planned behavior: attitudes, subjective norms, and (perceived and actual) behavioral control. This analysis was conducted by sifting through the material, highlighting text, terms, and phrases, writing notes, and reorganizing coded segments along the three parameters. Finally, we reviewed the results, reflected on the data, and constructed narratives that would convey study respondents’ attitudes towards science communication activities. While focused primarily on the qualitative data, the analysis is mindful of the quantitative findings of the survey and we occasionally refer to the quantitative findings to provide the appropriate research context for our analysis.

Initially, we analyzed responses that provide insights as to the degree to which scientists positively or negatively value engaging in science communication activities. Overall, respondents expressed positive attitudes towards science communication.

With regard to content of science communication, many respondents indicated that science communication should be preoccupied to correcting misunderstandings about the social value of science and to fostering a deeper understanding of the public value of science. Since society often fails to fully appreciate basic science’s potential to contribute to social outcomes, science communication should firstly aim at “increasing people awareness of the contributions of basic research to our daily lives” (#191) ⁶ and foster “much more interaction between society, scientists and sufferers” (#160). As one respondent put it, science communication must answer the question “What is the point of this research?” so that “everyone can understand why they are doing this research and why it will be a benefit to us” (#222). Secondly, science communication is also an opportunity to address negative perceptions of science. “With the current challenges about data replication in the popular press, good communication has never been more vital” (#116). Patients, some scientists worried, “do not understand the real role of bioscience for the general health [and think negatively of] animal experimentation, drugs in general and natural cures (that there are not miraculous cure of cancer)” (#133). Thirdly, many respondents also indicated that, if science were better understood, the public would be more willing to support basic research. To foster public support of science “the benefits of basic research are fed back to the people who in the end are funding the research” (#216). Others favored communication efforts that focus on “non-instrumental” aspects of science (#118 mentioning astronomy as “very popular, even though it has no direct economic benefit”), the complexity and mystery of scientific inquires (#176: “science is a process that takes time and hard work and great thoughtfulness, not just a received body of ‘facts’ that one is supposed to memorize”), and science as “awesome and fun” (#150). The last respondent also added that “we need to get away from the pure health focus of research: in neuroscience all the promises of the decade of the brain were bad public policy. Let us not repeat that mistake over and over again!” (#150). “Good science is similar to painting or composing music, but the tools are different” (#166).

Our respondents expressed different attitudes depending on the audience of science communication. Exchanges with clinicians and patients received an assessment that is more favorable. By contrast, the relationship with policymakers, politicians, and the media were mostly framed in negative terms (we discuss policymakers, politicians, and the media under ‘perceived behavioral control’). With regard to clinicians, more than 20 of the 133 comments remarked that interactions between basic scientists and clinicians are beneficial in two ways. First, scientists can benefit from clinicians’ feedback. Due to the “constant patient and clinician contact,” the clinical environment may generate “a good amount of feedback into the direction of the science” (#184). In addition, interactions with basic scientists may reorient clinicians’ focus from “treatment of symptoms … to understanding disease [since] basic research tends to want to understand how things work” (#139). Second, clinicians can become agents of science when communicating with other users (colleagues and patients) (#215: “MDs have key role to play in supporting/communicating basic research”). Similarly, interactions with patients are perceived as positive. To stimulate these exchanges, study participants supported the proposal to relocate some basic researchers into hospitals and suggested that scientists should get actively involved with clinical care to generate more opportunities for knowledge exchanges ⁷ . They also noted though that more needs to be done for science communication to happen in the clinical environments. “Simply having basic scientists at hospitals does not help if they are not actively involved with clinical problems” (#244). Cultural barriers may be an obstacle to knowledge exchanges, since clinicians and scientists come from different cultural worlds, and their exchanges would improve if their educational paths—e.g. a PhD in biology or an MD for clinicians—overlapped more.

A different picture emerges from the analysis of comments involving the media and policymakers. Media are perceived as agents of distortion: the information that the media feed to the public does not accurately portray how scientific research is conducted and how it contributes to the betterment of society. Some imputed this state of affairs to the fact that “the media are for the most part corporate and can say/show what they want” (#146) and to the fact that the media engage in “false equivalence giving equal weight to dubious if not downright false opposing viewpoints” (#141). Against this background, the media are nonetheless perceived to play a critical role in science communication efforts.

The media should thus “promote well-informed discussion of basic research discoveries and their contribution to knowledge in the media” (#230) with “accuracy of portrayal” (#108). Policymakers are also perceived negatively. The respondents questioned policymakers’ ability and willingness to understand and embrace science. “The ignorance of a large portion of the general public is terrible, especially members of Congress” (#208).

Overall, negative attitudes seem to be grounded on scientists’ perception that the public lack sufficient scientific literacy. In fact, many respondents expressed the wish to have policymakers—politicians in particular—with better science credentials and more willingness to understand science (“Need more policy makers who are scientifically literate, with some experience in basic science not simply owning medical degrees” #107). To address the scientific literacy deficit, respondents proposed various improvements that could change their attitudes towards communicating knowledge to nonscientists. Improving education in science and technology seems to be a clear target to enhance science communication. Curriculum reform at all levels of education was proposed by numerous respondents. Curricula, some participants suggested, should emphasize how the scientific process works and how scientific findings benefit society. Scientific literacy would improve “if future generations are more aware of the scientific process” (#234). Scientific education should start at an early age—“get kids into science. That’s the only way to change the mindset of the generations” (#220). Also, “make sure all children have introduction to biology, chemistry, physics and human physiology; emphasize scientific process and benefit of scientific research” (#111). Curricula should also “showcase how basic science has paved the way for translational science” (#129). Children’s interest should be also fostered outside the classroom by having the right kind of programs and initiatives (“Bring back Bill Nye, The Science Guy, to public television; and/or promote TV/movie with scientists as protagonists to encourage more children to pursue STEM education and careers” #120). In addition to triggering interest in science, these initiatives would forge more informed citizens and better policymakers (“An unprepared mind will be unprepared for the proposed policies” #199). To some, curriculum reform is arguably more important than having career scientists engage the public since “much promotion of research by even highly regarded scientists does not serve the knowledge/long term interests of the public that well” (#235).

We then turned to analyzing responses that address the normative beliefs and subjective norms held by bioscientists. These are comments that concern whether bioscientists are expected to engage in science communication activities and how individuals perceived those expectations. Participants seem to be in agreement that bioscientists are expected to engage with the public. Furthermore, many argued that bioscientists ought to become more engaged with science communication efforts and take greater ownership of these efforts. This is part of their professional obligation, especially where research is publicly funded. “Since the public funds the research in many cases, scientists have an obligation to educate the public and to deliver new knowledge to them” (#111). Bioscientists must get involved “in the training of media-relations personnel (i.e. journalists), of both trainees in the discipline (e.g. classroom settings of journalism schools) and of active journalist (e.g. through seminars or workshops)” (#177).

Respondents also identified various institutional and professional opportunities for greater engagement with science communication. Organizing meetings and open houses, an option directly asked about in our survey, is one. Institutions could organize “periodic open house type of event where basic scientists could explain their science to general public, entrepreneurs, clinicians” (#170). These events are also seen as opportunities for scientists to reflect on their work and “think about potential societal benefits just by interacting with people in different background and educational qualification” (#170). Professional organizations are seen as science communication tools: “professional societies could do more societal outreach” (#154). However, some argued, these organizations need to change their approach to be effective: “Scientific societies need PR and even something like political operatives to expose industry and anti-science groups for their agendas and questionable credentials. Instead of not engaging or expecting some sort of even playing field, scientific associations need to get media savvy and do more to reclaim what science stands for and represents” (#141). Finally, respondents mentioned publication outlets as valid communication tools. Open access journals in particular offer opportunities for disseminating science and for the public to appreciate that public funds are used in ways that lead to beneficial social outcomes—or “to see their tax dollars at work” as one respondent put it (#200). The idea of an online forum “where people feel safe to ask ‘stupid’ questions” (#210) was also brought up.

Finally, we analyzed answers to find insights concerning bioscientists’ behavioral control. This variable refers to perceived and actual ability to engage in science communication activities. Bioscientists’ lack of skills to communicate science effectively was mentioned several times as a source of inability of bioscientists to communicate science effectively. As a solution, respondents proposed offering proper training to scientists, but also choosing only the most “effective communicators to engage with the public and media” (#243). “Train early career scientists to communicate with the public [and] include training in communication tools and public speaking during postdoctoral training” (#196). More importantly, respondents felt that they are unable to participate in public engagement of science because of perceived environmental constraints, such as time and resources. To them, these constraints constitute actual barriers to greater involvement with science communication. “Overburdened schedules” (#154) force bioscientists to struggle with keeping up with the day-to-day work of scientific research. Demands for greater commitment to science communication efforts means asking them to reduce the time and energy spent responding to the demands of today’s academic environment, where ‘publish or perish’ is the norm. Scientists’ perception is that it is simply impossible, especially considering that engaging with the public does not lead to immediate returns in terms of productivity and career outcomes. “You are raising the possibility of more sessions, more discussions, more meetings (etc.) for persons who already do research into the late hours, are involved in classroom teaching, do university administration, and instruct people in the lab” (#181). This cannot be done without increasing funds, available to scientists, labs and institutions, to pay for outreach and communication efforts, or setting up directs rewards (in the form of career advancements, credit recognition in grant reviews or prizes) for scientists who engage in such efforts. Otherwise, the risk is that further demands may “drive good scientists away from science” (#181).