Search , Seek , Share : A National Survey Assessing Americans ’ Information Channels and Sharing Behaviors During a Pandemic

The spread of accurate and inaccurate information happened quickly in the early stages of the COVID-19 pandemic and understanding how this occurred is important to prepare for communication of future disease outbreaks. The purpose of this study was to understand Americans’ information seeking and sharing behaviors during the early stages of COVID-19 and was guided by the following objectives: identify passive sources/channels of information; identify active sources/channels of information; and describe how frequently and across which channels/sources the U.S. public shared information about COVID-19 in early stages of the pandemic. Results indicated people first found information about COVID-19 from personal communication but turned to national and international organizations if they were to actively seek information. Scientists and universities were some of the least sought after and shared sources of information. The sources shared most were from the Centers for Disease Control and Prevention and the World Health Organization. Implications from this research are a need for communicators to use grassroots communication efforts during a crisis, to actively share information early during a crisis, to share information outside of traditional academic networks, and to collaborate with sources inside and outside of traditional Extension networks.


Introduction and Problem Statement
The presence of infectious, zoonotic diseases is increasing and could pose significant impacts to the agricultural industry, particularly in sectors of the industry that involve animals. Policies and disease management will need a more proactive approach in order to address zoonotic diseases while ensuring an abundant food supply (Rohr et al., 2019). The full impact of an emerging infectious disease was exemplified during the COVID-19 pandemic. The novel coronavirus disease  was first identified in Wuhan China in late 2019 and spread rapidly through more than 160 countries, reaching the United States in late January. The impacts of COVID-19 surpassed the usual impacts of typical disease outbreaks and affected areas of science, entertainment, politics, economy, education, and more in countries around the world (Sahin et al., 2020). The World Health Organization (WHO) declared COVID-19 as a pandemic in March, followed by the United States declaring COVID-19 a national emergency, which allowed Americans to access federal support. Over the next few weeks through March and April, the United States experienced school closures, stay-at-home orders, economic turmoil, and individual and community interventions (American Journal of Managed Care [AMJC] Staff, 2020).
COVID-19 impacted the health of people around the world, as well as impacted the economy through job and income loss due to widespread business closures and decreased spending while people were staying at home. Additionally, COVID-19 brought about changes in public purchasing and consumption behaviors. The shelves at grocery stores were often emptied as people stocked up on household goods and cooked more frequently at home (Clements, 2020). The purpose of the current study was to understand the public's information seeking and sharing behaviors during these early stages of COVID-19.

Theoretical and Conceptual Framework
Information seeking behaviors, information channels, and sources impact a person's understanding of risk communication and, therefore, guided the conceptual framework of this study. The perception of a risk can impact the goals of an organization, making effective risk communication particularly important for health organizations or state and federal governments during a public health crisis (Covello & Sandman, 2001). The groundwork for trust in risk communication is laid between the community and the organization dealing with the risks involved (Telg, 2019).
Communication messages during a public health crisis should be carefully crafted, and the source and channel used for the message should be thoughtfully matched (Liu et al., 2011). Communication channels are the method in which a message is shared, including print media, online media, and social media; communication sources refer to the sender of the message, which could be an organization or an individual. Providing communication messages on targeted channels is only half of the process. Understanding the seeking and processing tactics that individuals use when confronted with messages can provide insight on the impact of the message. Bates (2005) described differences between active vs. passive information as simply the act of an individual acquiring information about a topic versus passively taking in information. Bates (2005) further defined differences through types of behaviors associated with each. Awareness and monitoring are often considered passive behaviors, whereas searching and browsing are considered active behaviors. Curiosity and browsing often lead to directed searching or actively seeking information to answer specific questions (Bates, 2005). Kahlor et al. (2006) stated that active information seeking tends to be more purposeful and leads to more systematic processing of information.
As the agricultural industry proactively prepares for future disease outbreaks or pandemic situations, crisis communication should be considered. Two-way communication is particularly important during a pandemic or public health crisis, such as COVID-19. Covello (2003) recommended seven best practices for communication in a public health crisis, which include the following: (a) involve stakeholders as partners; (b) listen to all parties involved; (c) be truthful and transparent; (d) collaborate with other credible sources; (e) provide information to the media; (f) be compassionate and clear; and (g) plan ahead.
The information sources and channels used to communicate during a crisis warrant consideration as these can impact an individual's perception of the crisis (Nguyen et al., 2017). During natural disasters, the public has found traditional news sources, such as printed news, to be more credible (Endsley et al., 2014). COVID-19 occurred in the most technologically advanced time in history, which enabled the transfer of information to spread quickly through traditional and emerging channels. Rapidly increasing infection rates signified a need to provide knowledge and best practices "at a pace equal to or better than the spreading epidemic" (Chan et.al., 2020, p. 1). Popular social media channels were utilized including Facebook, Instagram, and Twitter, all of which are commonly used to disseminate information. Early in the pandemic, the social media platform TikTok included users posting videos of a coronavirus dance, which featured moves that mirrored the steps for preventing the spread of COVID-19 (Wang, 2020). The platform also had inaccurate and culturally insensitive messages related to COVID-19 (Kenyon, 2020). The WHO saw this as an opportunity and joined the platform to spread valid health communication related to COVID-19 (Brown, 2020).

Purpose
The purpose of this study was to understand the U.S. general public's information seeking and sharing behaviors during the early stages of the COVID-19 pandemic, specifically to learn what information sources were passive and which were active in an effort to understand how to reach the public with purposeful communication during a zoonotic disease pandemic. The following objectives guided this study: 1. Identify the passive sources/channels of information used by the U.S. public in the early stages of the COVID-19 pandemic. 2. Identify the active sources/channels of information used by the U.S. public in the early stages of the COVID-19 pandemic. 3. Describe how frequently and across which channels the U.S. public shared information about COVID-19 in the early stages of the pandemic.

Methods
An online survey research design was utilized. A third-party company, Qualtrics, was consulted to obtain a nonprobability opt-in sample of U.S. residents 18 years of age or older. This approach is commonly used to make population estimates (Baker et al., 2013) and has become common in research examining public opinion of emerging issues due to increased access to internet, relatively low cost of online surveys, and higher response rates (Dillman et al., 2014). An online link was distributed to 4,935 U.S. residents during the second and third weeks of March, 2020. Attention filters (e.g. "select strongly agree for this answer") were used to identify respondents not paying attention. Respondents who did not complete all items of the instrument or did not select the appropriate answer to attention filters were excluded from analyses. Useable responses were obtained from 1,512 residents (31% participation rate). Potential exclusion, selection, and non-participation biases can limit the use of nonprobability samples (Baker et al., 2013). To minimize impacts, post-stratification weighting methods (Kalton & Flores-Cervantes, 2003) were executed post hoc to balance results based on 2010 Census data to accurately approximate to the population (Baker et al., 2013).
An original researcher-designed questionnaire was the instrument for this study; it was reviewed for face and content validity by a panel that consisted of faculty and staff at the University of Florida and its One Health Center of Excellence. The instrument was reviewed for content accuracy, clarity of wording, readability, and survey flow (Colton & Covert, 2007). A pilot test (n = 50) was conducted to examine preliminary data distribution and ensure functionality of built-in survey logic. This instrument was a part of a larger study; four sections of the questionnaire used to meet the objectives of this study: (a) information search frequency; (b) active and passive use of information sources; (c) information sharing frequency; and (d) use of information-sharing networks and sources. Respondents' information-search frequency was assessed using a single item to gauge how frequently they sought information about COVID-19 the month prior to the study. Responses were collected using a 5-point ordinal scale: 0 = never; 1 = rarely (1-2 times); 2 = occasionally (3-4 times); 3 = often (5-6 times); 4 = very often (more than 6 times). To identify the channels through which respondents had passively received information about COVID-19, they were asked to indicate, by checking all that apply, where they had seen or heard information about COVID-19 during the past month. Active information-seeking behavior was assessed using 24 items designed to capture respondents' likeliness of use for both the source and method of delivery (e.g., "Department of Health websites," "social media posts from the Department of Health," etc.). Responses were collected using a 5-point Likert scale: 1 = very unlikely; 2 = unlikely; 3 = neither likely nor unlikely; 4 = likely; 5 = very likely. To assess information-sharing frequency, respondents were first asked to indicate if they had shared any information related to COVID-19 the month prior to the study (1 = yes; 2 = no). Respondents who had shared information were then asked how frequently they had done so. Responses were collected using the same, previously mentioned frequency scale. Use of networks and sources when sharing was assessed by first asking respondents to indicate, by checking all that apply, which networks they used to share information (e.g., Facebook). Respondents were then asked to list the top three sources of the information they had shared.
Data were analyzed using the SPSS26. Analyses consisted of descriptive statistics (e.g., means, standard deviations, modes, frequencies, and percentages). Open-ended responses of the top three sources of information shared were analyzed in Excel using Glaser's constant comparative method (Glaser, 1965) to identify themes.

Findings
Passive Channels/Sources of Information In order to understand how the public found out about information related to COVID-19, it was important to understand the passive ways someone saw information about the virus in the early stages of the pandemic. The passive items were provided in a "check all that apply" format, with a mix of the most common sources and channels. Because a filter question for completing the survey was "have you heard of the coronavirus disease COVID-19?", all respondents had heard about COVID-19. 9.8% Note: Responses were collected using a "check all that apply" multiple-response format, so percentages do not add up to 100%.

Active Sources/Channels of Information Used by The Public about COVID-19
The next piece to fully understand how people received information about COVID-19 was to understand their active search for information. Researchers asked how often people actively sought information about COVID-19 over the month, i.e., late February through early March, 2020. The vast majority of people had actively searched for information related to COVID-19 (f = 1,431; 96.4%) to some extent, with the largest number of people searching very often for information (f = 626, 41.4%; see Table 2). Next, researchers sought to understand where respondents were searching for information. All respondents, including the 5.4% who indicated they had not searched for information in the month prior to the study, were asked to indicate the likelihood they would seek information related to COVID-19 from select sources/channels. The highest means for actively seeking information were all for national or international organizations' websites involved in health:  Table 3). Responses were collected using a 5-point Likert scale: 1 = very unlikely; 2 = unlikely; 3 = neither likely nor unlikely; 4 = likely; 5 = very likely.

How the Public Shared Information about COVID-19
The next step in understanding information spread was to determine if people were sharing information about COVID-19 and, if so, how they were sharing information. Slightly more than two-thirds (f = 1,045; 69.2%) of respondents said they had shared information about COVID-19 in the month prior to this study (see Table 4). On average, respondents who indicated they had shared information about COVID-19 did so occasionally (M = 2.59, SD = 1.06). When asked specifically what personal and social networks they used to share information, the majority of people said they used a form of personal communication (n = 831, 54.9%). The next highest percentage was for Facebook (n = 404, 26.7%), and the lowest was for TikTok (n = 44, 2.9%). Full results are in Table 5. 2.5% Note. The "Other" category included WhatsApp, Snapchat, Reddit, Skype, Hangouts, and other forms of personal communication The last step taken to understand how people shared information during the early stages of COVID-19 was to ask the 1,045 respondents who indicated they had shared information to list the top three sources they used to do so. Diversity in channels and sources shared were great (see Table 6). The largest number of people said they used the CDC or WHO (f = 679). The CDC and WHO had to be included in the same category because a large portion of people who said they used the CDC and WHO included these as one source in the form of CDC/WHO or similar instead of listing these as distinctly different sources. This was followed by news outlets (f = 435) and conversations with people (f = 412). A large number of sources in the middle range of use were highly specialized and ranged from social media (f = 246) and online sources (f = 148) to the government (f = 153) and doctors (f = 143). On the lower end were science and research (f = 22), blogs and maps/statistics (f = 12), email (f = 11), and religious sources (f = 7; see Table  6).

Conclusions, Discussion, and Recommendations
Both active and passive information-seeking behaviors among the American public early in COVID-19 were observed, as well as some key differences in the primary sources and channels of information when pursued actively versus consumed passively. These findings support Bates' (2005) recommendation to use multiple channels and sources to reach all audiences as part of an effective risk communication strategy. If members of the public were to actively seek information about COVID-19, they were more likely to do so from the websites of major national or international health organizations. While previous work indicated people found newspapers a more credible source during a crisis (Endsley et al., 2014), the results of the current study showed people were more likely to use online sources. With rapid changes occurring early in the COVID-19 pandemic, online sources may have been attractive to Americans due to the instantaneous nature of online channels and quickly updated crisis response information.
Respondents were not likely to actively seek information about COVID-19 from scientists, universities, and Extension systems directly, which provides further evidence for a need to work with health-focused organizations during a zoonotic disease outbreak. This also provides evidence that scientists and health communicators must move beyond using communication channels viewed only by people in academia. Effective science and health communication may also help a university or Extension program to establish a relationship before a public health crisis (Telg, 2019). Personal networks were also extremely important in the sharing process among those who shared information in the early stages of the COVID-19 pandemic. This provides further support for the use of grassroots communication efforts for risk communication in the early stages of a pandemic. Though people indicated they eventually would seek information from national and international organizations' websites, these health organizations could have been on the forefront of communication if they were actively pushing information in the beginning stages of the pandemic. While social media channels received average to little consideration from people when they were in the passive or active search process, people indicated more use of personal communication, Facebook, Twitter, Instagram, YouTube, and TikTok when it came to sharing information about COVID-19. While only a small portion of people were using TikTok during the early stages of the pandemic, its emergence as a channel/source of information during this pandemic was rapid (Wang, 2020) and may indicate an emerging trend in its use for health communication messages in the future.
This work supports the continued use of Covello's (2003) best practices for communication in a public health crisis. Involving stakeholders as partners and listening to all parties involved will allow communicators to connect through personal networks, which were the most used active and passive channels in this study. Collaboration with other credible sources is also supported by this work, as the top active search channels/sources were all health-focused organizations.
Scientists, universities, Extension programs, and communicators from many disciplines, including public health, science, and agriculture, will need to partner with health organizations in future zoonotic disease outbreaks.
Future research should investigate the trust of the public related to their primary and secondary passive and active sources of information to understand the role trust played in risk communication during COVID-19. Additionally, future research should explore the relationship between the public and popular sources of information, such as the CDC and WHO, to determine why people might react more positively to some sources compared to others. And lastly, research should be conducted to understand how these sources, channels, and behaviors changed throughout the pandemic.