Technical difficulty and concern about trolls had negative effects in both models

The data came from 661 respondents to a statewide survey fielded between May and July 2016, which achieved an overall response rate of 28% . In addition to UC Cooperative Extension professionals, the survey included people from organizations that are part of the knowledge network engaged in outreach, education and communication: producer groups, nongovernmental organizations, consultants, resource conservation districts, government agencies and others. While some respondents did manage farms, we were not targeting farmers but rather those who develop and deliver information to farmers. It would be useful for future research to extend the survey to farmer populations, specific consultant groups such as pest control advisors, and agricultural knowledge networks in other countries and U.S. states. Using the framework of the diffusion of innovation theory, our analysis tested the following hypotheses: extension professionals who perceive a greater relative advantage are more likely to adopt ICT; less likely to adopt ICT are extension professionals who perceive ICT as incompatible with their values and social norms, extension professionals who perceive ICT as too complex or time consuming and extension professionals who are uncertain about how to measure ICT effectiveness or strategically target audiences. In addition, we tested how demographic factors may be linked to ICT adoption, with the expectation that the patterns would be similar to the patterns in the general population. To pursue the possibility of an extension gap,ebb and flow table we tested whether UC employees had a lower ICT adoption rate relative to other types of extension professionals.We tested our hypotheses by first constructing dependent variables for the number of ICT platforms used and the frequency of social media use. The survey assessed ICT adoption by asking respondents if they used blogs, websites, email, mobile applications, Facebook, Twitter, Pinterest, Instagram or LinkedIn to communicate or learn about sustainable agriculture. We constructed a yes/no variable for each ICT platform. To zero in on the frequency of use for just the social media platforms , we followed the Pew Internet Survey in establishing the following categories of use: several times a day, once a day, a few days a week, every few weeks, or less often.

Importantly, the focus was on using social media for professional communication about sustainable agriculture, not personal use of social media. To analyze how perceptions about the attributes of ICT are related to ICT behavior, we constructed a social media frequency scale that calculated the average frequency of social media use across all five platforms, plus an “other social media” category. The scale ranged from 1 = do not use any social media to 6 = use all social media several times per day, with mean = 2.01. To calculate the number of total ICT platforms used, not just the social media platforms, we summed the number of platforms respondents checked; the numbers ranged from 0 to 9, with mean = 3.81. We then estimated multivariate models with social media frequency and total number of ICT platforms as dependent variables, and the four attributes of innovations as independent variables . Respondents’ perception of social media’s relative advantage was measured by averaging their responses to six statements related to its capacity to reach audiences and help extension professionals coordinate professional activities. These statements form a reliable scale ranging from 1 = strongly disagree on all statements to 5 = strongly agree on all statements, with mean = 3.71. Respondents’ perception of compatibility was measured in their responses to four statements: social media risks spreading fake news, there are positive incentives for its use, most colleagues use it, and it involves too much interaction with trolls. The response options did not form a reliable scale, so we included each statement as a separate variable in the analysis. Perception of social media complexity was measured in responses to these four statements: it takes too much time, it’s technically difficult to use, best practices are well known, and the large number of platforms is confusing. Again, the response options did not form a reliable scale, so we included each statement as a separate variable in the analysis. Lastly, respondents’ perception of the measurability of social media’s effectiveness was assessed. One statement suggested it was easy to measure effectiveness; the other statement suggested it was easy to identify appropriate audiences for social media. Responses were averaged into a reliable scale that ranged from 1 = strongly disagree on both statements to 5 = strongly agree on both statements, with mean = 2.32. To test the effect on ICT adoption of users’ demographic characteristics, we used the same dependent variables as described above, and we estimated the same models using the following demographic independent variables: sustainability attitude , age , income , UC system , male and education .

Figure 1 reports the overall adoption rates for the ICT platforms. In decreasing order of use, email was used by 92% of the respondents, followed by websites and Facebook , with Instagram and Pinterest having the lowest adoption rates. While the results for the most popular platforms echo the results for the general population, Twitter and LinkedIn were used relatively more by extension professionals because they are specifically intended for information dissemination and professional networking. Figure 2 reports the average temporal frequency indicated by each respondent for using just social media. For the general population, Pew reports that 55% of Facebook users and 23% of Twitter users access their accounts several times per day. In contrast, our sustainable agriculture stakeholders in California access Twitter and Facebook at the lower rates of once a day or a few days per week. The lower frequency of use for LinkedIn most likely reflects that the content changes more slowly than the events communicated on Twitter and Facebook. Figure 3 plots the coefficients from an ordinary least squares regression model for social media frequency and a Poisson model for number of ICT platforms , with the variables arranged in order of decreasing magnitude from the social media frequency model. Relative advantage had the strongest positive relationship with both the frequency of social media use and number of ICT platforms. Respondents who thought most of their colleagues used social media also used more ICT platforms, more frequently.Interestingly, the attributes of innovations were less important for the number of ICT platforms used than for the frequency of social media use, where the capacity to measure effectiveness and identify the audience had a positive influence and lack of time and concern about fake news had strong negative effects. The existence of professional incentives, confusion about the number of platforms and knowledge about best practices were unimportant in both models. Figure 4 suggests some demographic variables behaved in ways consistent with the general population: ICT and social media use was higher among female, younger and wealthier respondents. The results also corroborated the technology gap described earlier: UC employees used fewer ICT platforms and used social media less frequently than other respondents. More educated respondents also used social media less frequently, which contrasts to the general population, where educational levels are positively correlated with social media use. In the context of professional extension activities, UC system employees or those with advanced degrees may be stressed for time, perceive social media as incompatible with norms of scientific knowledge sharing or lack professional incentives.

The breakdown of UC respondents was 35% campus faculty, 13% Cooperative Extension specialist, 22% Cooperative Extension county advisor, 6% other academic title, 10% student/ postdoctoral scholar, 11% staff and 3% other. Sustainable agriculture stakeholders in both developed and developing countries are quickly catching up to the information revolution that has transformed society in the 21st century. Our results confirmed the usefulness of diffusion of information theory, which frames the debate about the benefits and risks of ICT in terms of innovation attributes related to relative advantage, compatibility, complexity, observability and trialability. Extension professionals clearly recognized the relative advantages for ICT in terms of quickly communicating with a more diverse and distant audience, but with less potential to coordinate on-the-ground activities. Extension professionals are more likely to capitalize on these relative advantages if their colleagues are also using ICT, and they have good tools for measuring effectiveness. The most important barriers for widespread adoption of ICT were time constraints, technical complexity and incompatibility between norms of scientific discourse and the reality of trolls and misinformation on the internet. These results support some concrete recommendations for organizations seeking to increase the use of ICT and make it more effective for extension professionals. Resources could be invested in developing a community of practice for aspiring ICT users interested in using ICT for outreach, with leadership from extension professionals with an established reputation for successful innovation. Communities of practice are one of the organizational concepts in e-eXtension, and are defined as informal networks of professionals with a common goal who regularly interact to share information and expertise . They can help creatively solve problems, transfer knowledge and develop professional skills and are effective where the network of individuals is distributed across many administrative units or system components, as is the case with extension professionals experimenting with ICT. A sponsoring organization can help foster a community of practice by identifying potential members,flood table providing organizational infrastructure for interaction and measuring effectiveness with appropriate metrics. Such a community of practice should document the potential advantages of ICT and provide information about best practices. It would increase awareness about how many extension professionals are using ICT, which would help create a community norm. The community of practice should include a diverse set of stakeholders, including digital technology specialists from outside of agriculture who are knowledgeable about different types of tools, altmetrics and social media strategies that are effective in digital communication. It is also important for agricultural extension organizations, including land-grant universities, to establish clear guidelines for recognizing the value of ICT as an extension tool that complements traditional communication strategies and ways of extending knowledge. If extension professionals know what counts in terms of documenting their professional activities for job performance evaluation, they are less likely to be confused and view ICT as a risky endeavor. Developing such guidelines would benefit from consultation by outside specialists with expertise in digital tools and measurement. The risks of misinformation and credibility may be some of the most important for extension organizations to address at a strategic level.

Such organizations typically desire to be perceived as impartial providers of evidence-based information. Social media platforms recognize that legitimate knowledge exists outside of Cooperative Extension but also provides a gateway for misinformation. At the individual level, the risk is not so much that a particular extension professional may make a mistake in communicating their own research, but rather they may unintentionally spread misinformation from others and be required to invest additional resources in sorting accurate social media information from misinformation. At the institutional level, social media’s democratization of information creates the fear of messages being corrupted, misinterpreted or simply lost in the wash of real information and misinformation. In both cases, it is important to avoid damaging the reputation of providing high-quality science communication. However, Bastos et al. provide some evidence that may mitigate these fears. Examining the topology of Twitter networks connected with UC Agriculture and Natural Resources Twitter users, Bastos et al. found that communities focused on specialized agricultural topics formed centralized networks in which a relatively small number of expert nodes collected and broadcast information to a large audience. In other words, relative to more general users and nonspecialists, technical experts become more central in the online networks and serve as important information hubs for specialized and technical issues. This suggests that social media communication is not completely incompatible with the traditional extension goal of providing hubs of expertise. Further research is needed to increase confidence in our results and recommendations. More systematic comparison between extension professionals and their clientele would corroborate the extent and nature of the technology gap. While our findings are relevant for extension professionals involved in sustainable agriculture, it would be important to generalize the research to other types of agricultural sectors, compare different commodity groups, directly survey farmers and extend the research to natural resource managers. Using big data approaches to monitor the dynamics and effectiveness of communication is also an important effort going forward.