Posted October 20, 2016
Photo: School of Library and Information Science faculty members who conducted research
projects after the 2015 flood in South Carolina
In the aftermath of the 2015 catastrophic flooding of several counties in South Carolina, the USC Office of the Vice President for Research announced an internal funding opportunity in support of faculty research to examine all aspects of community resilience. The faculty of the School of Library and Information Science (SLIS) received several grants to conduct multidisciplinary research projects. These studies showcase the expertise of SLIS faculty members and demonstrate the benefits of collaboration with scholars in different academic units. The research results have been integrated into the school curricula, enhancing courses at both the graduate and undergraduate levels.
Library and information science researchers are well positioned to investigate community development and engagement. They see social responsibility and community impact as central to scholarship, and through their research they seek to not only inform local communities, but to make on-the-ground impact for the better. The catastrophic flooding provided a platform for good scholarship and positive impact. While the results of these projects will be reported at multiple scholarly conferences and journals, the true impact is in the recommendations made to the library systems, local government, and federal agencies for the improvement of community engagement and services.
Public Libraries and Public Health
The first project was an investigation of the value of public libraries and their legitimacy as partners of public health agencies during a disaster. Motivated by a belief in the value of public libraries to their communities, the team consisted of Drs. Samantha Hastings, Jingjing Liu, and Feili Tu-Keefner and two SLIS doctoral students — Elizabeth Hartnett and Hassan Zamir.
Public libraries have a long history of providing community outreach programs and services, e.g., after Hurricane Katrina, when the local public libraries played vital roles in disaster response and recovery. However, research on the partnership between public libraries and public health agencies during a disaster is scarce.
The research team wanted to see whether public libraries proactively provided comprehensive health information regarding emergency planning and disaster preparedness and also to know the technology used by the librarians and community members during and after the South Carolina floods.
The team, partnering with the Richland Library and the South Carolina State Library, focused on Richland and Orangeburg Counties, which were hard hit by a series of floods. They used a framework for risk communication preparedness and implementation recommended by public health experts. The methodology included focus-group meetings with public library administrators and librarians, one-on-one interviews with community members, and an in-depth interview with a FEMA agent. In addition, a survey was conducted to investigate community members’ use of technology — such as social media — before, during and after the disaster.
The research documented the Richland Library’s main library and branches creating disaster recovery centers for FEMA after the flood. These centers enabled community members to use the public libraries to work with FEMA agents in filing damage claims online. Fourteen percent of all FEMA applications were filed at the Richland Library’s main and branch libraries. The Richland Library was a water distribution site; librarians took books, toys, and computers to shelters. On the Richland Library’s Facebook site, the library’s posts were shared 1,386 times — an average of 98 shares for each post. From October 4-12, the Richland Library’s Twitter account gained 242 new followers. This evidence affirms our belief in public libraries’ social value to their communities and their legitimacy as partners of public health agencies during the disaster.
The findings show that technology access was crucial to obtaining credible information and disseminating resources and services to the community. The Internet was predominantly used by librarians for this purpose. The survey results especially provided insights on public library users’ access to social media, among other information sources, during the flood. Community members used social media, TV, and text messages to receive and verify the alerts sent by public agencies. The major social media sites used were Facebook, Twitter, and YouTube. Community members said that Facebook was the most useful communication tool to access information and to stay connected with others. It was fortunate that the electricity stayed on, so that the Internet was available in many areas hit hard by flooding. Even when people were evacuated to shelters, they could use cell phones to access the Internet and social media.
The results of the study also show a discrepancy between reliable resources vital to consumers and the health information shared with them by the public libraries. Information and technology literacy issues created barriers for many community members in accessing FEMA applications and filing claims online. Public librarians were not fully prepared to provide the necessary information for adult users, especially through online venues, before and after the flood. The team recommends that library and information science education programs and health sciences libraries collaborate to design professional development programs and continuing education opportunities for public librarians.
Floods and People with Disabilities
The 1,000-year flood impacted many South Carolinians statewide, and among those most severely impacted were persons with disabilities. In this study, Dr. Clayton Copeland with faculty members from the USC School of Medicine, Drs. Robert Dawson and David Leach, investigate the effect of these catastrophic events on the lives of community members with disabilities. Although South Carolina rapidly responded to the crisis for the population as a whole, the state was largely unprepared to help persons with disabilities — a group that represents 14 percent of the total population. The focus of this research was on identifying the gaps in services and the best-practice models that can be used to effectively meet the needs of persons with disabilities.
In their research, the team used a mixed-methods approach, including surveys of flood survivors with disabilities, emergency responders and volunteers. They also conducted follow-up interviews with survey respondents. Surveys were distributed through Able SC, the United Way, Portlight and the American Red Cross. Data came from the responses of 123 community members with disabilities, as well as emergency responders, volunteers and employees of the Able SC.
Of the 123 respondents with disabilities, 67 percent were females and 33 percent males. Identification of disability can aid in developing future emergency plans. Sixty-eight percent of respondents reported having one disability, and 28 percent responded having two or more disabilities. The most frequently cited disabilities were physical disability (28 percent), chronic illness (18 percent) and psychiatric disability (16 percent).
The survey also addressed functional limitations, or the impact of disabilities upon activities of daily living. These functional limitations could correspond with difficulties respondents experienced during the crisis. A person who is deaf, for instance, could need closed captioning or sign language interpretation in order to access emergency information and could be in danger if appropriate accommodations are not provided. A person with a physical disability may have difficulty navigating inaccessible landscapes. Twenty-one possible functional limitations were reported. The four most frequently cited functional limitations were: walking (40 percent); standing (35 percent); concentration (34 percent); and memory (33 percent). This is an important finding in that it demonstrates persons may have difficulty with mobility in all areas including evacuation centers. This corresponds with respondents reporting that specific instructions on evacuating were not provided or were difficult to understand (50 percent) with 43 percent of persons saying they needed assistance in the evacuation. The findings of concentration and memory are significant because they highlight the need for clear instructions during the crisis and recognition that instructions may need to be repeated or duplicated with a memory aid (note card).
The majority of respondents, 63 percent, were displaced from their homes. Fifty-seven percent reported damage to their homes. Among respondents who did not evacuate, cited reasons included: an inability to leave the residence; underestimation of the severity of the flood and resulting damages; and fear that possessions would be stolen or damaged. The majority of respondents experiencing displacement from their residences also reported that they did not relocate to an emergency shelter. Of respondents who reported staying in an emergency shelter, the length of time individuals remained in the shelter varied from 1 to 38 days. Conditions of the shelters were rated as good or better by a majority of respondents (60 percent), with 80 percent responding that they had everything they needed at the shelters. Such reports are especially significant, as Centers for Independent Living and other agencies had numerous reports from evacuees with disabilities indicating a lack of access to shelters, inaccessibility of shelters, or being separated from their service dogs. Persons with disabilities may have experienced difficulties or lack of access in the emergency shelters. While quantitative data do not capture evidence of this, qualitative findings include reports from respondents who observed persons with disabilities experiencing difficulties.
Response to a disaster is crucial to ensure safety of persons with disabilities. The majority of individuals received information about the flood through television (63 percent) or through electronic media (20 percent). Most (70 percent) reported that the time in which government officials responded to the floods was reasonable or good. Only 13 percent of the persons with disabilities surveyed reported poor governmental response to the flooding; nine percent indicated that the government or local officials did not know how to effectively meet the needs of persons with disabilities. The majority of respondents said that they did not receive help from any individual or group. However, for those who did, 40 percent were helped by FEMA; 18 percent by church-based organizations and 16 percent by the American Red Cross.
Persons with disabilities were surveyed after the flood. The time between the flood event and the aftermath follow-up ranged from three months to six months. Immediately after the flood, respondents reported the following as most significant needs: water (47 percent); food (30 percent); medications (20 percent); medical care (16 percent); and access to assistive technology or medical equipment (10 percent). Frustration was a key theme identified in both the quantitative and qualitative findings. Respondents were frustrated by the amount of paperwork and forms that were not accessible for persons with disabilities. Numerous respondents reported not receiving enough monies to cover the cost of repairing their homes. Respondents described financial and emotional stress seven months after the initial crisis. During follow-up phone calls to respondents eight months after the crisis, there were reports of mold, continual water damage, and lack of repairs to places of residence. Respondents also reported frustration and feeling that they are in a cycle of being sent to different departments or service agencies without any definitive outcomes.
The researchers suggest several potential areas of improvement. Disability awareness and emergency planning training are essential for professionals and volunteers. There should be greater awareness by volunteers regarding best practices for working with persons with disabilities in times of crisis. Services should be strengthened or developed to ensure equal access to all resources. One specific concern was a lack of response or promised services by FEMA and other organizations six months after the disaster. Improvements should also be made with communication, and provisions of and accessible modes of communication and services. This research highlights the need for training before the next natural disaster and the need for collaboration between persons with disabilities and emergency response personnel.
Computing and Social Media in Disasters
The third research project was that of Dr. Amir Karami, a SLIS faculty member, and Dr. Zhenlong Li of USC's Department of Geography. Because the unpredictable nature of disasters makes it hard to develop a comprehensive strategic response plan, these researchers have proposed a high-performance computational framework to investigate disaster information disseminated via social media. They have used a five-step process to extract, through Twitter comments, patterns that will be useful in evaluating the 2015 disaster management strategy and developing a more effective plan for preparedness, response, and recovery in the future.
In this study mined the spatiotemporal patterns of more than one million tweets containing the senders’ experiences, needs, and opinions during and after the South Carolina flood. The data collection was based on the frequency of related hashtags or words. For example, negative feelings were expressed earlier in the month, when people were without drinking water, and positive feelings appeared in mid-October, when the problem was solved. Interestingly, negative feelings about “traffic” did not change during the research time frame (October 3-15, 2015.)
Using the framework developed in this proposal, the research team was able to collect, analyze, and map information from tweets for all South Carolina counties, thereby gaining a better understanding of the flood effects from October 3rd to October 15, 2015. The patterns of people’s experiences, needs, and feelings were extracted through sentiment analysis and text mining. Such patterns can help track needs and problems, and can aid in evaluating disaster management. This type of research can be applied to a variety of domains and can be used in the event of future disasters.
Scholarship in Action
Taken together, the three studies discussed here demonstrate the power of combining good research with real needs in a time of disaster to affect positive change. It also demonstrates the ability of academics to quickly engage in times of crisis and to work across traditional disciplinary borders. As Hurricane Matthew ravaged the coast and waterways of the south east, the power of real-world scholarship becomes obvious. Libraries, governmental relief agencies, and health care providers must work together to learn from each disaster and build tools for others to implement. While the flooding in South Carolina is called a 1,000-year event, in truth the ravages of nature strike across the country on an increasing basis. We must all be prepared.
By Feili Tu-Keefner, Ph.D., with R. David Lankes, Ph.D., Jingjing Liu, Ph.D., Samantha Hastings, Ph.D., Clayton Copeland, Ph.D., Robert Dawson, Ph.D., David Leach, Ph.D., Amir Karami, Ph.D., and Zhenlong Li, Ph.D.