Assessment of Digital Technology Adoption and Access Barriers Among Agricultural Service Providers and Agricultural Extension Professionals

Jessica L. Drewry^1,*, John M. Shutske², David Trechter³, Brian D. Luck¹

Published in Journal of the ASABE 65(5): 1049-1059 (doi: 10.13031/ja.15018). Copyright 2022 American Society of Agricultural and Biological Engineers.

¹Biological Systems Engineering, University of Wisconsin, Madison, USA.

²University of Wisconsin, Madison, USA.

³Agricultural Economics, University of Wisconsin, River Falls, USA.

^*Correspondence: jdrewry@wisc.edu

The authors have paid for open access for this article. This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License https://creative commons.org/licenses/by-nc-nd/4.0/

Submitted for review on 10 January 2022 as manuscript number ITSC15018; approved for publication as a Research Article by Associate Editor Prof. Brian Steward and Community Editor Dr. Naiqian Zhang of the Information Technology, Sensors, & Control Systems Community of ASABE on 24 June 2022.

Highlights

• Agricultural service providers and Extension professionals were surveyed on technology use and barriers.
• Data privacy and security and hiring qualified IT staff are barriers in the industry.
• Surveys point to major opportunities for digital agriculture professional development and education.

Abstract. As agriculture becomes an increasingly data-driven sector, it is important to understand the technology use and barriers by the service providers who work closely with producers. This is especially important as these individuals can serve as drivers for technology adoption by farmers and ranchers. Although many surveys have looked at technology adoption by producers, little data on agricultural industry service providers exists. Surveys of Extension Professionals (EP) in Iowa, Michigan, North Dakota, Ohio, and Wisconsin, and several categories of Agricultural Service Providers (ASP) in Wisconsin were administered in 2018 to determine satisfaction with internet service and to learn more about information and technology usage and potential barriers to its adoption within the industry. In addition, follow-up informational telephone interviews were conducted with 13 Wisconsin EP. A paper survey was administered to ASP, which had a response rate of 35% (N=462). A similar online survey was administered to EP having an estimated 23% response rate (N=223). The frequency of transmission of data, information, and recommendations was high but specialized for both groups with more frequent use in crop production versus livestock (including dairy). The most commonly-cited barriers associated with the adoption of digital technologies by ASP were related to attracting and hiring well-qualified IT employees, training for both employees and clients, and data security. While data security was a commonly cited barrier, 16% of ASP worked for companies that did not have written policies on data privacy. ASP and EP consider their clients to be adept at using text messages, images, maps, and smartphones. They do not consider their clients to be skilled in their use of digitally accessed and used financial statements, spreadsheets, or laboratory analysis reports. Additional training should be better focused on these areas for ASP and EP which would allow them to be more effective in educating producers and to better realize the value associated with more thoughtful and widespread digital technology adoption.

Keywords. Agricultural data, Agricultural service provider, Extension professional, Internet access, Technology adoption, Technology barriers.

Adoption of technology can have many benefits for producers including increased production, reduced inputs, environmental protection, and increased resource efficiency. A United States Department of Agriculture (USDA) report analyzed data from the Agricultural Resource Management Survey and found that the adoption of GPS-based yield and soil mapping, guidance systems, and variable rate technologies slightly increased net returns and operating profits with the positive impacts of these technologies increasing with farm size (Schimmelpfennig, 2016). An analysis of grain farmers in Kentucky, USA found that the adoption of guidance systems increased net returns by up to 0.9% (Shockley et al., 2011). However, even with the potential to increase profits, adoption rates remain low within the industry. A USDA survey in 2010 found only 12% of farms under 243 hectares (the US average was 180 hectares in 2017 [USDA, 2021]) had adopted GPS soil/yield mapping, guidance systems, or variable rate technology (Schimmelpfennig, 2016). Another smaller national survey of crop input dealers found their rate of adoption of these technologies to be up to 78% across the dealers who participated, with adoption increasing since the inception of the survey in 2000 (Erickson et al., 2017).

Based on a survey of worldwide precision agriculture adoption, Lowenberg-DeBoer and Erickson (2019) argue that adoption has not been “slow,” but is instead restricted only to those uses which have a high perceived value by producers. To some extent, this argument assumes that individual producers have knowledge to adequately perceive and evaluate value. Similarly general studies of technology adoption show that “perception of usefulness” is a key factor (Adrian et al., 2005; Davis, 1989). Additional studies have found that privacy and security (61% citing these as a high or medium barrier) were also cited as barriers among crop, livestock, and dairy producers as it related to their adoption of digital agricultural technologies (Drewry et al., 2019). Other high-ranking concerns were the ability to keep up with technology change (55%), cost (54%), and lack of understanding about how to use data (51%) (Drewry et al., 2019). Interviews of German producers also found that many were concerned about data misuse and privacy as a result of the digital surveillance of individuals (Kutter et al., 2011). Additionally, a survey of sheep producers in the United Kingdom cited connectivity as a significant barrier in rural areas (Morris et al., 2017) which was also noted by Drewry et al. (2019).

Peer-to-peer interaction with other farmers/ranchers who had previously adopted technology and purposeful engagement with agricultural service providers has been found to increase technology adoption in agricultural producers. Interviews of German producers found that agricultural technology firms and other producers were the two most important direct influences for the spread of precision farming information. Professional literature and field days were the two most important channels to deliver this information (Kutter et al., 2011). Additionally, a survey of Danish and US producers found that crop advisors and fertilizer dealers were the main source of information on precision agriculture technologies (Fountas et al., 2005). Finally, a large scale survey of European Union producers found that among adopters of precision agriculture technologies, the use of a trusted and experienced advisor had a significant positive effect on adoption (Barnes et al., 2019). Given the key role these service providers play in adoption, it is important to understand the degree to which they are adopting and actively using technology as they serve this industry with products, information and services, and it is also useful to better understand their perceptions of producers’ abilities.

In addition to structural factors, demographic factors can play a role in technology adoption. Past work shows that the older the producer, the less likely they are to adopt technology (Charness and Boot, 2009). Demographic factors such as education, gender, and farm size likely play a significant role in technology adoption (Daberkow and McBride, 2003; Drewry et al., 2019; Michels et al., 2020; Pierpaoli et al., 2013; Schimmelpfennig, 2016). ASP and EP could play a role in helping agricultural producers to better understand the usefulness and economic value of technology, and the past studies cited above could be used to inform specific strategies for targeted groups (younger versus older producers, men versus women, more highly educated, etc.) when developing new delivery methods or educational material.

With respect to new delivery channels, online education and technical training through webinars, short courses, and video are often viewed as a potentially efficient method to train a large number of individuals who are dispersed in remote locations with the cost for delivery and travel being relatively lower as compared to one-on-one contact or contact in small group settings (meetings, in-person conferences, etc.). Online technical education could be used as a tool to increase interactions between key groups that appear from past research to be important to producer-level technology adoption (such as enabling producer-to-producer learning, advisor-to-producer interaction, and peer learning and sharing among ASP and EP). There are few studies on the efficacy of online learning for employees in the agricultural service sector; however, more generalized studies of undergraduate students have found little to no negative effect on course performance and learning in online vs. face to face instruction across a range of demographically-diverse groups (Jaggars and Bailey, 2010; Neuhauser, 2002). A study of performance gaps between online and face-to-face instruction at community colleges did find achievement gaps; however, these gaps narrowed for older learners (Xu and Jaggars, 2014). In a study of in-service training for EP, face-to-face instruction was perceived as more favorable than online training, but achievement levels of skill and knowledge were comparable if the training included highly interactive elements as a means to help learners have more direct experience with the principles being taught. (McCann, 2007).

As agriculture and the agriculture service industry become increasingly data-driven sectors with the ability to drive producers’ adoption of technology, these providers must be equipped to create, aggregate, transfer, and interpret digitally-collected, stored, and transmitted data. In addition, they need to be proficient enough to inform, answer technical questions, demonstrate value, and educate agricultural producers as part of their professional roles. This includes Agricultural Service Providers (ASP) such as crop consultants, equipment dealers, financial specialists, and University Extension Professionals (EP).

While many articles have studied technology adoption by producers (Daberkow and McBride, 2003; Michels et al., 2020; Morris et al., 2017; Paustian and Theuvsen, 2017; Pierpaoli et al., 2013; Schimmelpfennig, 2016), few have looked at the use, barriers, and attitudes toward technology use by ASP and EP. A better understanding of the current level of technological knowledge, data transfer, and customer proficiency could aid in developing better training methods and other strategies to bridge the important gaps between service providers and producers, ultimately influencing the application of new digital technologies and information on individual production operations.

The objective of the study presented here was to learn more from ASP and EP about their professional use of digital technologies and to better understand how they viewed and perceived their clients’ use of agricultural and informational technologies, understand limitations to technology adoption, and guide the training needs of those who work in businesses or educational institutions that provide goods, services, technical information, and other products as they perform their work in service and support of agricultural producers. Surveys (traditional paper and online) were designed to better understand current technology use, satisfaction with existing data-related services, utilization effectiveness, primary modes and types of data transfer between providers and producers, and perceived barriers to adoption. In addition, in-depth interviews with EP were conducted to understand their and their clients’ experiences and attitudes toward technology adoption more deeply. These interviews also were used to validate and/or corroborate survey findings.

Materials and Methods

Extension Professionals

Following Institutional Review Board (IRB) approvals from the University of Wisconsin-Madison and University of Wisconsin-River Falls, a 22-question online survey was sent via email to EP in Iowa, Michigan, North Dakota, Ohio, and Wisconsin in 2018. The survey was administered as a sample of convenience. Following one-on-one conversations between authors and leaders in the North Central state land grant university Extension offices, an email link was sent by the state-level Extension agricultural program leader. This is an administrative leadership position (with some of these individuals also holding the title of Associate Dean) located at each university. The decision to use these individuals as a primary point of contact was based on their level of name recognition and authority within each state. These individuals agreed to send the link and consent information to EP in their respective states with an agricultural focus. Individuals who were contacted by their respective program leader included local (county, regional, etc.) educators and state/campus-based specialists. Planned reminders (either two or three depending on the state) were sent to the agricultural program leader contact by email asking them to remind survey participants by email to respond over the course of one month. Increased responses were seen after each prompt.

In addition, 30 to 60-minute informational telephone interviews were conducted with 13 Wisconsin EP. Interviewees were selected from the list of individuals to whom the email survey was originally sent. Authors discussed and made attempts to ensure that the Wisconsin EP were representative relative to the survey population with respect to gender, role (primary commodity/production system served), and years of service in their role. This included individuals with multiple decades of service and some who were relatively new in their roles. Interviews followed a pre-prepared and IRB-approved script that contained questions on the themes that included their use of technology in extension programs, data transmission to/from producers, their barriers to adoption, and technology training for themselves and their role in providing training. Interviewees were also asked open-ended questions in a semi-structured interview format to elicit their thoughts and ideas as it related both to their use and their clients’ use of technology. The purpose of the interviews was to gain a deeper insight and provide additional context to enrich survey responses and to better understand their personal stories and experiences with technology. Interviews were concluded when no new ideas of concepts were shared by the interviewees (Francis et al., 2010).

Agricultural Service Providers

A 23-question, IRB-approved paper survey was designed by a team that included experts from the University of Wisconsin at River Falls Survey Research Center along with University of Wisconsin-Madison and University of Wisconsin Division of Extension. Members of the project team also had informal input on survey design and questions through small group and individual conversations, conferences, and other special events that informed some of the survey elements. For example, in a small group meeting with a group of crop service providers, regulatory officials, and others, issues of data privacy and ownership were identified as a general thread of interest that is likely not well understood by the industry. Additionally, in a small group meeting with a group of agricultural cooperative IT personnel, issues of being able to find qualified IT “talent” willing to work for a smaller company in rural areas were cited as highly significant and limiting concerns that the group would want to explore more deeply. The survey mailing list was systematically assembled from a variety of known, credible sources including:

• A mailing list purchased from Certified Crop Advisers for 630 Wisconsin addresses.
• A mailing list from the University of Wisconsin Center for Cooperatives for approximately 200 managers, COOs, CEOs, and other key contact people that that Center mails to regularly (program announcements, newsletters, etc.).
• A mailing list from the USDA Natural Resource Conservation Service (maintained by program staff at the University of Wisconsin Extension Service for 147 Technical Service Providers.
• A mailing list from the University of Wisconsin Extension Service for 127 personnel involved with nutrient management planning, custom manure pumping/hauling, and those involved in other nutrient/manure-related activities.
• A mailing list from the WI Agribusiness Association for 292 agricultural industry points of contact, managers, CEOs, COOs, and other leadership.

After duplicate entries were removed, these sources yielded contact information for 1,363 individuals.

Surveys were administered in 2018. The team used the Dillman survey method (Dillman, 2000) included an initial hardcopy mailing of the survey with a cover letter and consent information. Then, a postcard reminder was mailed to non-respondents 14 days after the first mailing. A second full mailing of the survey and letter was sent four weeks after the first mailing to non-respondents. Returned surveys were scanned and the data was entered by the University of Wisconsin-River Falls Survey Research Center. The survey instrument used in this study is provided as supplementary material (available at https://doi.org/10.13031/20465919.v1).

Statistical Methods of Survey Analysis

To analyze the data, independent variables were grouped into the following classes of demographic factors:

• Gender: men versus women (two categories).
• Age: those under 45 years of age versus those 45 and older (two intervals).
• Education: those with a high school diploma or less, those with some college or a 2-year degree, and those with a 4-year degree or higher (three ordinal variables).
• Years Employed in Agriculture: those working in the field for less than one year, those working 1-10 years, and those working more than 10 years (three intervals).
• Region of Wisconsin (For ASP): northern, northeastern, south central, southeast, west central (identified by the Wisconsin Department of Natural Resources [five categories]).

Respondents who indicated being retired, their primary occupation was farming, and duplicate entries were removed from the analysis. Additionally, “other” responses were reclassified to the appropriate category if applicable.

Frequency analyses were conducted to explore data trends and to identify potentially significant parameters for survey responses (SAS Institute Inc., 2011). Chi square statistics and p-values were reported for all tests of categorical variables. Logistic regression was used to model categorical and ordinal responses of the explanatory demographic variables (SAS Institute Inc., 2011). Internet satisfaction (ordinal), frequency of data transmission (ordinal), characterization of client proficiency with technology (categorical), and barriers to adoption (ordinal) were assessed. To categorize the characterization of ‘proficiency,’ the median of the summed scores for all 11 categories was used to separate ‘high’ and ‘low’ rankings of producer ability, with a score of 1 rating the client’s effectiveness as ‘not well’ and a score of 4 rating their effectiveness as ‘very well’. Results were presented as a point estimate of the odds ratio along with 95% confidence intervals.

Results and Discussion

Respondent Demographics

Extension Professionals

The EP survey link was sent to approximately 960 individuals and 223 were returned and usable. The precise number of survey recipients contacted was somewhat difficult to estimate with precision since the team relied on the personnel in each state’s email list and numbers were estimated by that state’s Extension administrative contact (typically the state extension “agricultural program” lead). During any given window of time within a state, there is generally some group of EP in flux as a result of retirements, moving to other positions, and other forms of attrition. Based on the estimated population of 960 EP, the overall response rate was 23% with an estimated 47% response rate at North Dakota State University, 45% at the University of Wisconsin, and 15% at Iowa State University. Respondent demographics are summarized in table 1.

The telephone interviewees included EPs with a mix of crop and livestock (including dairy) specializations. While common among some extension services, extension personnel focused on consumer horticulture issues were not included in the interviews. The interviewees were slightly more male (62%, N=13) as compared to the percentage of male respondents in the full survey. Forty-six percent of the interviewees (N=13) had 10 or fewer years of experience as compared to 49% of those in Wisconsin who completed the full survey.

Agricultural Service Providers

Of the 1,363 ASP surveys administered, 60 surveys were returned as undeliverable and 462 surveys were returned and usable, resulting in a response rate of 35%. After the first mailing, 66% of the total usable responses were received. An analysis considering demographics of respondents from only the first and second mailing indicated statistically significant differences in education levels. The respondents of the second mailing had a slightly lower education level as compared to the first (X²(3, N = 467) = 19.0, p <0.0008), but otherwise the respondents were similar over the various mailings, suggesting a relatively low non-response bias (Sheikh & Mattingly, 1981). Counties with high levels of agricultural production tended to have higher relative levels of respondents, an indication of regional representativeness (fig. 1). Most respondents mailed surveys from WI zip codes; however, a few were mailed from outside states in counties near the state’s border (Iowa, Illinois, Michigan, and Minnesota).

Demographics of survey respondents are summarized in table 2. Gender breakdown was comparable to the Bureau of Labor Statistics estimates, which estimated in 2018 that 75% of those employed in agriculture and related fields identify as male (BLS, 2020). The median age of respondents was between 45 and 54 years with the majority (72%) between 35 and 64. The ASP respondents had a higher degree of formal education as compared to the United States adult population. In 2015, the Current Population Survey reported that 33% of adults had a four-year college degree and 12% reported an advanced degree (Ryan & Bauman, 2016). Position groups were merged into four groups for cross comparisons: High level management including General Manager/CEO, Chief Operating Officer, and Chief Information Officer/IT Lead; Mid-level managers including Data Manager and Local Regional Manager; Workers including Non-managerial Sales/consulting and general employees; and Other.

Internet Service Satisfaction

Overall, internet service satisfaction was relatively high for both ASP and EP. The median non-mobile and mobile satisfaction for ASP and EP was 4 on a scale from 1 to 5, with 1 being very dissatisfied and 5 being very satisfied (tables 2-6). This was similar to the findings of high satisfaction among agricultural producers (Drewry et al., 2019). Overall, the differences in satisfaction levels with home and mobile connections was not found to be significant by location, gender, age, education, or years working in an agricultural profession using logistic regression (all p values greater than 0.05). Non-mobile mean satisfaction was significantly higher for EP that ASP (Z=8.06, p= <0.0001). This could be a result of land-grant universities having made investments in extension-related network technologies over the last decade and being in locations such as county courthouses or on university campuses. The lowest median satisfaction level (3) was for the cost of service. Although overall satisfaction was high, interviewees noted many problems with connectivity that interfered with extension activities or the adoption of technology by producers.

Extension Professionals were asked about their mobile connection status and the source of the funds that they use for their smartphones and service plans (table 7). Only 2% reported not having a mobile device. The majority of EP with a device (68%) pay for the phone and mobile data plan personally versus being paid for by their employer. Interviewees did not feel limited by poor internet access and indicated that they have developed new and creative ways to deliver and access information. However, the interviewees noted that access to mobile data had improved their programming and interactions with clients over time. ASP were not asked about funding for their mobile access and devices.

Technology Use and Digital Information Transfer

Extension Professionals

Extension Professionals were asked about the frequency of the transmission of data, information, and recommendations sent to their clients. Transmission was categorized as never, monthly or less, or frequently (at least most weeks). Crop production data, production-related information, and recommendations were the categories most sent with 77% reporting transmission of these categories. Transmission related to livestock was less common with 65% reporting some transmission (fig. 2). Similar trends, although a lower frequency of transmission, were seen for transmission from clients to EP. The frequency of data transmission was not found to be significant as a function of the EP demographic variables (such as age and gender) in the logistic model. All p values were greater than 0.05. This could indicate that institutional culture and encouragement to use current technology may be a stronger driving force than individual demographic characteristics such as age, gender, or position type as it relates to the frequency of data transmission. The higher usage of the internet for crop-related data transmission could be associated with many factors. For example, interviews with EP cited data transfer needs that pertained to a wide multitude of content-related issues such as nutrient management planning and transmission of yield data for decision making. Other factors cited in the differences between crop versus animal production data transmission included the fact that there is a more intense level and peak data generation/transfer during the growing season for crops, whereas animal-related questions might be more spread out over time. Additionally, the quantity of crop-related data might also be greater overall as compared to livestock as most livestock operations are also in the crop production business due to feed production. These differences are likely not due to privacy concerns among livestock producers because livestock producers appear to be no more likely to cite privacy and security as technology use barriers as compared to others (Drewry et al., 2019).

Figure 2. Number of respondents who reported transmission of data, information, and recommendations to clients electronically in each of the categories by extension professionals.

Data from each individual EP survey respondent was analyzed to determine the frequency of data transmission within any of the survey’s content categories that are included in figure 2 (safety, finance/management, natural resources, etc.). While there was a high percentage of respondents reporting never transmitting data within specific categories, fewer than 5% of respondents never transmitted any data, information, or recommendation electronically across all categories. This suggests that while transmission can be rather frequent, it tends to be specialized (fig. 3). On average, EP use computers for electronic data transfer 76% of the time and smartphones 23% of the time.

Figure 3. Frequency of any data, information, and recommendations transmitted via the internet by extension professionals.

Extension professionals reported a high level of contact with clients, with 31% reporting sending data, information, or recommendations daily. This was an interesting finding because a survey of more than 1,000 Wisconsin producers found that only 3.3% of farmers were in “regular” contact with extension (Drewry et al., 2019). There are a few potential explanations for this discrepancy in contact. One plausible explanation is that even though EP have this stated high level of digital contact, it is not possible to be in direct communication with anything close to the majority of agricultural producers in a specific window of time given the large number of producers relative to the number of EP. In Wisconsin, there are more than 60,000 farm operations and approximately 150 EP (including campus and county/regional); a 400:1 ratio which would be of similar magnitude in other Midwestern states. An additional possible explanation is that producers do not consider the full range of EP when they consider their contact with extension. For example, a local producer within a particular county will likely see their local/county extension educator as an EP, but that same person may not see a campus-based extension specialist the same way. Also, producers may not consider content that is published in newsletters, magazines, or other print/electronic media as being connected to EP even if they may be aware that a land-grant university is the original source.

Extension professionals were more likely than ASP to send (Z=4.63, p <0.0001) and receive (Z=2.32, p =0.02) crop data and to send (Z=4.45, p <0.0001) and receive (Z=4.48, p <0.0001) livestock data, information, and recommendations. Interviewees shared that they were primarily transmitting information and recommendations rather than receiving data from clients. They also shared that transfer of agronomic data typically occurred “as a last resort” when a specific problem on farm could not be identified or solved by an ASP. Some EP also discussed that data is often transferred from producers when they are working as partners with EP on cooperative research projects.

Extension professionals were also asked about the frequency of technology use on mobile and non-mobile devices (table 8 and table 9). Email was the most frequently used technology on both mobile and non-mobile devices. Creating videos (capturing/producing) was the least-cited type of use on these devices. A survey of German producers found that 58.5% owned a smartphone and with approximately half were using the phone for agricultural-production related purposes (Michels et al., 2020) as compared to 76% of EP in this survey. All of the interviewees stated that their use of email communication with clients had increased substantially in recent years, often replacing one-on-one interactions and phone calls. However, phone calls were still common and text messaging is also increasing. Many reported that text messaging was their most reliable form of communication with clients, while only a few that did not use it as a form of communication. Given that 68% of EP are paying for their own plans, this could become a barrier to communication. The low use of mobile and non-mobile devices to create video content was cited by interviewees to occur primarily because of a lack of training and comfort. For example, one interviewee stated, “I've got a camera, I've got a remote microphone, I’ve got editing software; it's just the time. And it's also difficult to do by yourself.” Additionally, the ability to develop timely information, especially during the growing season, was another frequently cited obstacle.

Figure 4. Number of respondents who reported transmitting data, information, and recommendations to clients electronically in each category by agricultural service providers.

Figure 5. Frequency of data, information, and recommendations of all types transmitted to/from clients electronically by agricultural service providers.

Agricultural Service Providers

Agricultural service providers were asked about the frequency of their transmission of data, information, and recommendations to their clients. Crop production related transmissions were common with 74% reporting some transmission. Dairy and livestock transmissions were less likely to be sent or received over the internet with 52% reporting some livestock-related transmission (fig. 4). Similar trends, although a lower frequency of transmission was seen for transmission from clients to ASP. When looking at the transmission of any data type, we find that fewer than 10% never transmit data with clients (fig. 5). As with EP, ASP use of digital transmission of information is common, yet highly specialized, which accounts for the large number of responses in the “never” response for some categories among individual respondents. The frequency of data transmission was also not found to be significant as a function of the demographic variables (all p values were greater than 0.05), similar to EP.

Agricultural service providers were asked about which devices were used to transmit the data, information, and recommendations. They used a computer for 56% of transmissions and smartphones for 28%. The bulk of transmission was done via computers, which was logical given potential for large data sets and that most farm data management software is computer rather than mobile based. However, this may shift as the number of mobile applications increase and begin to offer more decision-making capabilities.

Respondents were surveyed about the policies and personnel related to data management and privacy. Of the 462 returned surveys, 370 worked for companies with written policies on data ownership and 390 for companies with written policies on data privacy. Having a written policy on data privacy was highly correlated with written policies on data ownership r(350)=0.75, p<0.0001. Having a person whose position duties include customer data management was significant, but not highly correlated with written policies on data ownership r(352)=0.53, p<0.0001 or privacy r(368)=0.38, p<0.0001. Given the concern expressed by agricultural producers about privacy and security (Drewry et al., 2019; Kutter et al., 2011), it was expected that many companies would have written policies on both privacy and data ownership, and this expectation may account for the high adoption rate of these policies.

Barriers to Technology Adoption

Agricultural service providers were asked about barriers to technology adoption. The majority cited hiring people with agricultural technology expertise (58%) and attracting potential employees to the community (54%) as moderate to big problems (table 10). None of the top three cited barriers were significant by demographic factors of respondents. Data security was cited as a big (9%) or moderate (25%) problem by just over one-third of respondents. Data security is a commonly cited barrier to technology adoption by agricultural producers (Drewry et al., 2019; Kutter et al., 2011). A survey of producers in the United Kingdom found that the most commonly cited barriers to adoption of agricultural decision support tools were performance, ease of use, peer recommendation, cost, habit, age, IT education, facilitating conditions (such as internet access), and ability to meet compliance requirements (Rose et al., 2016). Many of these themes overlap with this study’s findings. The EP interviewees said that clients were more concerned about data privacy and security with respect to transmitting and receiving financial information as compared to agronomic data.

The ability to transmit information was cited as a moderate problem by 27%, even though internet satisfaction was high among ASP. The top five cited ‘big’ problems were not associated with any demographic factors. In telephone interviews with EP, themes of internet connectivity issues and lack of perceived value among producers were frequently brought up when discussing barriers to technology adoption. The lack of perceived value has also been cited as an important barrier by other researchers (Lowenberg-DeBoer & Erickson, 2019). While Extension professionals may not be the ‘most frequently’ contacted source for information by producers who have been surveyed, they do tend to have the quality of being highly-trusted by producers as it pertains to science-based topics and issues (Prokopy et al., 2015). This may uniquely position them to educate producers on the value of technology adoption where such value has been demonstrated by unbiased and high-quality research. As one EP interviewee stated, “we can give them all these resources, but I think part of our job in extension is you're telling them how they can use this research, these resources, and technology … and show that they actually do have a benefit to what you're doing.”

Training for both ASP/EP providers and producers could help reduce problems related to understanding and using agricultural data. Both ASP and EP preferred face-to-face training (table 11 and table 12). For ASP, those under 44 years of age prefer face-to-face meetings (40%), conferences (28%) and on-line publications/videos (25%). Those 45 years of age and older overwhelmingly prefer face-to-face training (44%) with webinars (18%) and on-line publications/videos (13%) a distant second and third, respectively. However, these age-based differences were not significant (?²(2, N=465)=23.3 p=0.20). The absence of significant trends indicates that results are likely due to personal preference. This survey looked only at the learning preferences of respondents, not the effectiveness or cost of those alternatives. Several studies have found little to no difference in learning between online and face-to-face instruction suggesting this might be an area for the industry to focus more effort on in the future (Jaggars & Bailey, 2010; McCann, 2007; Neuhauser, 2002; Xu & Jaggars, 2014).

Extension professionals were asked to rate their interest in receiving and providing professional development on topics connected to information technology on a scale from 1 to 4, with 1 being not interested and 4 being very interested. Extension professionals were significantly more likely to want to receive (2.9 ± 0.9) than to offer (2.5 ± 1.0) educational programs and activities that focus on using information technology t(221)=7.91, p<0.0001. Given that 42% of ASP consider information technology training for their clients to be a moderate or big problem, providing information technology training to producers could be a valuable service provided by EP. However, EP may require more professional development before feeling competent and confident enough to provide this service. Given the expansion of mobile applications with more sophisticated interfaces and algorithms, the need for information technology training will likely increase (Mendes et al., 2020).

Neither EP nor ASP consider producers to be skilled at using or interpreting digitally-delivered financial statements, spreadsheets, laboratory reports (e.g. feed, soil, or milk analysis), or in effectively using smartphone applications (figs. 6 and 7). Although surveys have found that producers use smartphone frequently, specifically for agricultural applications, (Drewry et al., 2019; Michels et al., 2020), EP and ASP do not consider producers’ use of mobile application technology to be proficient. However, both groups consider producers skilled at effectively using text messaging, pictures/ images, maps, and smartphones in general. This has the potential to place a financial burden on EP given that the most effective means of communication must come on their personal devices, which 68% pay for personally. EP, as compared to ASP, consider producers to be more proficient at use of word documents (Z=8.43, p <0.0001), images/ pictures (Z=4.65, P <0.0001), with videos (Z=4.38, P <0.0001), and smartphone applications (Z=2.16, P=0.031). ASP, as compared to EP, consider producers to be more proficient with laboratory analysis/reports (Z=-3.28, p <0.0011) and financial statements (Z=-4.42, p <0.0001) although neither group considers them to be proficient overall in these categories. For ASP the median of the summed scores for all 11 categories of 25 was used to separate ‘high’ and ‘low’ rankings of producer ability. Those younger than 45 were 2.5 times more likely to rate producers as having a ‘high’ ability (?²(4, N=438)=15.7 p<0.0001). For EP, the median summed score of 27 was used to separate rankings. EP location was found to be the only significant demographic factor in rankings of producer ability and proficiency in using various types of digital technologies and applications (?²(2, N=203)=14.6 p<0.0157). Michigan State and North Dakota State Universities were both more likely to rate producers in the ‘high’ ability group (table 13). While there were some significant differences between demographic groups, most all ASP and EP were consistent in their ratings of producer ability. It is important to note that the COVID-19 pandemic may have served as a catalyst to improve ASP, EP, and producer proficiency in digital technologies and increase the frequency of digital communication between groups (Baffoe-Bonnie et al., 2021; Samuel, 2021).

When discussing producers’ use of data, one EP interviewee stated about producers, “they are collecting lots of data but they're not really digesting it, looking at all of the data saying, what does this really mean?” This same opinion was conveyed in similar ways by many interviewees. In interviews with EP, use of social media was also a common discussion theme. One interviewee’s response summarized the general consensus well, by stating, “I think most (not every) producer is pretty good at having social media. I don't know if they're always using social media to the best benefit of their farming operation, but they have it as a step in the right direction.” As agriculture becomes more data driven, it will be important that producers become more proficient in collecting, interpreting, and making actionable decisions based on their data. Interviews of EP suggested that their clientele is not proficient at data interpretation, and this limits the effective technology use by clients. Their overall sense was that they are more proficient at data collection; however, they may lack the skills of interpretation, analysis, and making the decisions that must follow, placing that burden onto service providers who may or may not be fully equipped with the tools that they need.

Conclusions

Respondents were surveyed about the frequency of data, information, and recommendation transmission. Crop data were most frequently transmitted with 75% of ASP and 77% of EP reporting some level of transmission. There were also a high number of respondents who reported never transmitting certain types of data indicating that transmission is specialized among both groups. EP were more likely than ASP to transmit crop and livestock data, information, and recommendations. EP interviewees indicated that these exchanges were primarily of information and recommendations rather than raw/less-refined data and did not find one type of producers more likely to share agronomic data than another.

Both EP and ASP consider their clients to be adept at using text messages, images, maps, and smartphones. Alternatively, they do not consider their clients to be adept at using digitally-delivered financial statements, spreadsheets, or laboratory analysis/ reports. This is consistent with interviewees’ perception that producers are becoming proficient in collecting data but not interpreting and acting on that collected data. These areas should be a focus of training by EP and ASP to improve producer’s decision-making abilities.

Agricultural service providers were asked about barriers to technology adoption in their professions. The most cited problems were related to attracting and hiring employees, training for both employees and clients, and data security. Interestingly, 32% cite transmitting data as a moderate or big problem, yet median satisfaction of mobile and non-mobile internet was rated as 4 on a scale of 1-5, indicating that the majority were satisfied. Additionally, while data security was a commonly-cited barrier, 16% of ASP worked for companies without written policies on data privacy. Another barrier cited as a moderate or big problem was IT training for their clients (42%) and their staff (28%) indicating the ASP were not equipped to handle these training needs. In addition, EP are not interested (or perhaps lack confidence) in offering training on information technology and are less likely to want to offer than receive training. Another barrier to producer education and potential technology adoption may be the reliance of EP on personal smartphone and data plans (68% pay for their own plan). While EP interviewees state they have developed solutions for poor or nonexistent data connections, they also gave many benefits to access during producer interaction. Lessening that financial burden may improve communication and programming between EP and producers.

Training in agricultural and information technologies could fill the needs of ASP and EP, especially with respect to increasing agricultural technology experience to replenish the aging workforce. Additionally, training in the interpretation and making actionable decisions with data could help ASP and EP to assist agricultural producers to make effective use of the increasing amounts of data that farms are generating to improve decision making. However, developing such training is no trivial task given the breadth of agricultural and information technologies and the many different software applications used to manage agricultural data. Developing online and/or interactive training may be able to fill the need for training in the agricultural service industry. The issue associated with recruitment of IT talent into agricultural businesses and sometimes into rural agricultural communities is complex and deserves further conversation. Similar issues are prevalent in healthcare and teaching, where it can be difficult to recruit practitioners into rural or other underserved industries. One possible approach could be to develop specialty tracks that could include incentives and opportunities through internships, residencies, and other programs for engineers, technologists, and others with IT training and experience in ways that have been done with healthcare professionals (Patterson et al., 2019) and teachers (Guha et al., 2017).

Acknowledgments

Support for this work was provided by the UW-Consortium for Extension and Research in Agriculture and Natural Resources (CERANR) and National Institute of Food and Agriculture, United States Department of Agriculture, Hatch project 1022349.

References

Adrian, A. M., Norwood, S. H., & Mask, P. L. (2005). Producers’ perceptions and attitudes toward precision agriculture technologies. Comput. Electron. Agric., 48(3), 256-271. https://doi.org/10.1016/j.compag.2005.04.004

Baffoe-Bonnie, A., Martin, D. T., & Mrema, F. (2021). Agricultural extension and advisory services strategies during COVID-19 lockdown. Agric. Environ. Letters, 6(4), e20056. https://doi.org/10.1002/ael2.20056

Barnes, A. P., Soto, I., Eory, V., Beck, B., Balafoutis, A., Sánchez, B.,... Gómez-Barbero, M. (2019). Exploring the adoption of precision agricultural technologies: A cross regional study of EU farmers. Land Use Policy, 80, 163-174. https://doi.org/10.1016/j.landusepol.2018.10.004

BLS. (2020). Labor force statistics from the current population survey. 2020. United States Bureau of Labor Statistics. Retrieved from https://www.bls.gov/cps/cpsaat11.htm

Charness, N., & Boot, W. R. (2009). Aging and information technology use: Potential and barriers. Current Directions Psychol. Sci., 18(5), 253-258. https://doi.org/10.1111/j.1467-8721.2009.01647.x

Daberkow, S. G., & McBride, W. D. (2003). Farm and operator characteristics affecting the awareness and adoption of precision agriculture technologies in the US. Precis. Agric., 4(2), 163-177. https://doi.org/10.1023/a:1024557205871

Davis, F. D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS quarterly, 13(3), 319-340. https://doi.org/10.2307/249008

Dillman, D.A., 2000. Mail and Internet Surveys: The Tailored Design Method. 2nd Edition. John Wiley Co., New York.

Drewry, J. L., Shutske, J. M., Trechter, D., Luck, B. D., & Pitman, L. (2019). Assessment of digital technology adoption and access barriers among crop, dairy and livestock producers in Wisconsin. Comput. Electron. Agric., 165, 104960. https://doi.org/10.1016/j.compag.2019.104960

Erickson, B., Lowenberg-DeBoer, J., & Bradford, J. (2017). 2017 Precision Agriculture Dealership Survey. 1-27. Purdue University. Retrieved from https://agribusiness.purdue.edu/precision-ag-survey

Fountas, S., Blackmore, S., Ess, D., Hawkins, S., Blumhoff, G., Lowenberg-Deboer, J., & Sorensen, C. G. (2005). Farmer experience with precision agriculture in Denmark and the US Eastern Corn Belt. Precis. Agric., 6(2), 121-141. https://doi.org/10.1007/s11119-004-1030-z

Francis, J. J., Johnston, M., Robertson, C., Glidewell, L., Entwistle, V., Eccles, M. P., & Grimshaw, J. M. (2010). What is an adequate sample size? Operationalising data saturation for theory-based interview studies. Psychol. Health, 25(10), 1229-1245. https://doi.org/10.1080/08870440903194015

Guha, R., Hyler, M. E., & Darling-Hammond, L. (2017). The teacher residency: A practical path to recruitment and retention. Am. Educator, 41(1), 31.

Jaggars, S., & Bailey, T. R. (2010). Effectiveness of fully online courses for college students: Response to a department of education meta-analysis. https://doi.org/10.7916/D85M63SM

Kutter, T., Tiemann, S., Siebert, R., & Fountas, S. (2011). The role of communication and co-operation in the adoption of precision farming. Precis. Agric., 12(1), 2-17. https://doi.org/10.1007/s11119-009-9150-0

Lowenberg-DeBoer, J., & Erickson, B. (2019). Setting the record straight on precision agriculture adoption. Agron. J. https://doi.org/10.2134/agronj2018.12.0779

McCann, B. M. (2007). The effectiveness of extension in-service training by distance: Perception versus reality. J. Ext., 45(1).

Mendes, J., Pinho, T. M., Neves dos Santos, F., Sousa, J. J., Peres, E., Boaventura-Cunha, J.,... Morais, R. (2020). Smartphone applications targeting precision agriculture practices-A systematic review. Agron., 10(6), 855. https://doi.org/10.3390/agronomy10060855

Michels, M., Fecke, W., Feil, J.-H., Musshoff, O., Pigisch, J., & Krone, S. (2020). Smartphone adoption and use in agriculture: empirical evidence from Germany. Precis. Agric., 21(2), 403-425. https://doi.org/10.1007/s11119-019-09675-5

Morris, W., Henley, A., & Dowell, D. (2017). Farm diversification, entrepreneurship and technology adoption: Analysis of upland farmers in Wales. J. Rural Studies, 53, 132-143. https://doi.org/10.1016/j.jrurstud.2017.05.014

Neuhauser, C. (2002). Learning style and effectiveness of online and face-to-face instruction. Am. J. Distance Education, 16(2), 99-113. https://doi.org/10.1207/S15389286AJDE1602_4

Patterson, D. G., Andrilla, C. H. A., & Garberson, L. A. (2019). Preparing physicians for rural practice: Availability of rural training in rural-centric residency programs. J. Graduate Med. Education, 11(5), 550-557. https://doi.org/10.4300/JGME-D-18-01079.1

Paustian, M., & Theuvsen, L. (2017). Adoption of precision agriculture technologies by German crop farmers. Precis. Agric., 18(5), 701-716. https://doi.org/10.1007/s11119-016-9482-5

Pierpaoli, E., Carli, G., Pignatti, E., & Canavari, M. (2013). Drivers of precision agriculture technologies adoption: A literature review. Procedia Technol., 8, 61-69. https://doi.org/10.1016/j.protcy.2013.11.010

Prokopy, L. S., Carlton, J. S., Arbuckle, J. G., Haigh, T., Lemos, M. C., Mase, A. S.,... Angel, J. (2015). Extension's role in disseminating information about climate change to agricultural stakeholders in the United States. Climatic Change, 130(2), 261-272. https://doi.org/10.1007/s10584-015-1339-9

Rose, D. C., Sutherland, W. J., Parker, C., Lobley, M., Winter, M., Morris, C.,... Dicks, L. V. (2016). Decision support tools for agriculture: Towards effective design and delivery. Agric. Syst., 149, 165-174. https://doi.org/10.1016/j.agsy.2016.09.009

Ryan, C. L., & Bauman, K. (2016). Educational attainment in the United States: 2015. U.S. Departemnt of Commerce Economics and Statistics Administration. Retrieved from https://www.census.gov/library/publications/2016/demo/p20-578.html

Samuel, S. D. (2021). COVID-19 pandemic and beyond: Utilizing digital ICT tools for agricultural innovation transfer to farmers. SSRN. https://doi.org/10.2139/ssrn.3809014

SAS Institute Inc. (2011). SAS/STAT(R) 9.3 User’s Guide, Second Edition. SAS Institute Inc.

Schimmelpfennig, D. (2016). Farm profits and adoption of precision agriculture. United Stated Department of Agriculture. Retrieved from www.ers.usda.gov/publications/err-economic-research-report/err217

Sheikh, K., & Mattingly, S. (1981). Investigating non-response bias in mail surveys. J. Epidemiol Community Health, 35(4), 293-6. https://doi.org/10.1136/jech.35.4.293

Shockley, J. M., Dillon, C. R., & Stombaugh, T. S. (2011). A whole farm analysis of the influence of auto-steer navigation on net returns, risk, and production practices. J. Agric. Appl. Econ., 43(1), 57-75. https://doi.org/10.1017/S1074070800004053

USDA. (2021). Farming and Farm Income. 2022. Retrieved from https://www.ers.usda.gov/data-products/ag-and-food-statistics-charting-the-essentials/farming-and-farm-income/

Xu, D., & Jaggars, S. S. (2014). Performance gaps between online and face-to-face courses: Differences across types of students and academic subject areas. J. Higher Education, 85(5), 633-659. https://doi.org/10.1080/00221546.2014.11777343