Universities face a variety of crisis management problems by virtue of the fact that many of them function much like their own little city, from having meeting spaces and residential areas to community leadership and a police force. With on campus populations often in the thousands and total enrollments even higher, it is inevitable that student will be involved in one of these crisis situations. When that happens, the university carries a duty to ensure that the student is healthy, well, and able to continue their progress on their degree.
As students are seeking support from their networks and university staff to make it through the challenging dynamic of college, many student affairs professions are ill equipped to properly advise students. If the resources existed to understand what kind of situations students are involved in and how often those situations occurred, they might be better prepared to serve a student properly. Particularly in hands-on parts of higher education, like residence life, such data might inform trainings, in-services, and other professional development.
Gathering, analyzing, and reporting the data on incident rates in residence life might give a department a stronger ability to provide the highest level of service for their students.
Residence Life departments record after-hours on-campus incidents in reports that are reported to a staff listserv, which generates a significant dataset of incident responses. These datasets often go unused when they could be examined to gather an understanding of trends in the occurrence of various on-campus incidents.
As higher education is shifting resources towards both data analysis and increasing crisis management needs, stores of incident data from residence life may contain insights invisible to even the most honed of professional and human intuition. Without an existing model by which to analyze residential life critical incident data, this study seeks to describe the existing data from a large residential campus. The study may benefit practitioners and researchers with a useful tool for response practices and new crisis management data alike.
On Arizona State University’s Tempe campus each of the six residential areas are staffed by a graduate or entry-level housing staff member on-call, who provide second level coverage. From 5:00 pm to 8:00 am on weekdays and all day on weekends, this individual provides crisis management for their respective residential areas. They are referred items of concern by the student staff member on-call in the first level of coverage, or their supervisor, who serves in the third level. Each morning, second level on-call individuals send an incident report to all University Housing staff. Incident reports contain the what, where, and when of each incident during their shift. As a member of the second level of on-call for a residential community, I am interested in how an analysis of the incident reports might reveal patterns and trends that would allow second level responders to be better prepared and have greater access to key resources. It is my observation that there is a discernible pattern in the existing data and having the analysis to verify my observations would make it possible to advocate for potential policy changes. Policies around mental health support and substance abuse are of direct interest, as critical incidents containing either or both appear to be on the rise.
The purpose of this study is to describe the data contained in the incident reports from University Housing at Arizona State University Tempe’s six residential areas during the fall 2018 semester, from August 2018 to December 2018. Results from this study could be used to better understand the total number of incidents that occur, where and when they occur, and open discussion on what can be done to mitigate on-campus critical incidents.
After the mass shootings at Virginia Tech University in 2007 and Northern Illinois University in 2008, the need for research into higher education crisis management increased, as long-held positive perceptions of campus safety began to decline (Keller & Hughes, 2011). Threats from inside and outside campuses generated the need for comprehensive, well-developed crisis management plans (Catullo, Walker, & Floyd, 2009). Three categories of severities were developed in these plans: disasters, crisis, and critical incidents (Holzweiss & Walker, 2018). “Disasters” constitute the highest level of crisis management, events that affect an entire region, city, or locality in addition to the institution . The middle level is called “crisis” and are events that only affect the institution itself (Holzweiss & Walker, 2018). The third and final category, “critical incidents” contains the localized incidents on a college campus without broader impact .
The literature on the breadth of crisis management required of institutions of higher education shows the magnitude of the problem they face. Institutions must react to situations where students engage in risky behaviors with alcohol and other drugs, deal with mental health concerns, and many other unknown personal challenges (Bernat, Lenk, Nelson, Winters, & Toomey, 2014; Catullo, Walker, & Floyd, 2009; Keller & Hughes, 2011). Response plans must be contingent on staff, faculty, and student needs, as well as key decision makers and available resources (Catullo, Walker, & Floyd, 2009; Keller & Hughes, 2011). Despite the pressing need, assessments found that in the years between September 11th, 2001 and the shootings at Virginia Tech in 2007 institutions did not significantly advance in preparations to handle a crisis or disaster .
Despite being the lowest severity of the three crisis management categories, critical incidents represent a more frequent type of crisis management addressed often by junior student affairs professionals (Holzweiss & Walker, 2018). Given the right conditions, these critical incidents can turn into crisis situations that demand a response from senior staff members (Holzweiss & Walker, 2018). Despite the higher rate of occurrence and potential for escalation, few resources have been dedicated to studying these lower level crisis management situations (Keller & Hughes, 2011). This study represents an attempt to describe and analyze critical incidents through the lens of a residence life department where many such incidents occur regularly.
Residence life departments and professionals occupy a unique space in higher education. While most university staff and faculty leave to their personal residences after the workday, residential staff live among the on-campus population day-to-day and confront critical incidents in that setting (Golden, 2018). One of the most significant of those confrontations is the rise in mental health-related incidents (Trela, 2008). Many papers have been dedicated to research the sudden rise in mental health concerns in higher education and best practices to address the problem (Golden, 2018). Critical incidents related to mental health have a large impact on residence life professionals, who live and work amongst students in various stages of crisis (Canto et al., 2017). The severity of these mental health concerns necessitate that housing professionals train in counseling techniques, and programs with counselors-in-residence are gaining traction at various universities to provide additional support (Trela, 2008). When confronting the underlying causes of distress students also often choose to self medicate, causing drug and alcohol use related critical incidents (Trela, 2008).
The use of alcohol and other drugs in residential students leads to a variety of crisis management situations. Residence life may confront noise complaints, medical transports, sexual misconduct, injuries, vandalism, and academic struggles due to self-medication with drugs and alcohol (Bernat et al., 2014). Critical incidents can reduce the ability to cope with stress, and related social work research has found one solution is to restore balance by decreasing distress and burnout (Cacciatore, Carlson, Michaelis, Klimek, & Steffan, 2011). While crises and underaged drinking are typically referred to the police, critical incidents often are kept within the university judicial system (Bernat et al., 2014). For both the police and the university there is a lack of understanding on what sanction best educates and modifies behavior within those systems to achieve the goal of reducing distress (Bernat et al., 2014; Cacciatore et al., 2011).
Gaps in the literature exist as to the best way to address substance abuse and mental health crisis in residential populations. What types of critical incidents happen with the most frequency, to what kind of students, at what times also is unknown, limiting the opportunity to use that data to best devise support theories for practice. Higher education may have to rely on the knowledge generated in other service industries like social work, healthcare, and insurance to begin to build the body of knowledge necessary to learn from the various crisis management situations presented to all levels of university administration.
Given the apparent gap in the research regarding crisis management, particularly around critical incidents, it is difficult to gain a better picture of the level of need from researcher and practitioners for this kind of study. What has been found by surveying top-level administrators is that university administrations are underprepared for crises and disasters (Catullo, Walker, & Floyd, 2009). Crime and psychological research into college crisis management contain some limited findings but the need for additional examination to confirm theories (Bernat et al., 2014; Trela, 2008). Higher education research into entry level-professionals also shows that very little is known about critical incident response (Holzweiss & Walker, 2018). As a relatively new topic the research is growing, and many studies have begun to identify various actionable steps. Universities are now acting on these by forming crisis response teams, providing additional staff and training, and breaking down the information barriers for the cross section of potentially impactful departments (Keller & Hughes, 2011). The existing research shows the need for a better understanding of the overall data, space where I believe my study will work within to show the rates of critical incidence on-campus (Keller & Hughes, 2011). While critical incidents have not been the priority of the existing research, further studies may allow for a better understanding of the issues that escalate to become fully fledged crisis situations and beyond.
I will use secondary data from the on-call reports of the Arizona State University Tempe campus department of University Housing in this study. During the hours of 5:00 pm-8:00 am from Monday to Friday and 8:00 am-8:00 am Saturday and Sunday graduate and entry-level Housing staff serve in an on-call capacity, resolving various student, facility, and campus issues. After completion of each shift, at 8:00 am each day, the staff member updates all housing officials with an “On-Call Report” documenting all incidents in their area. Represented in this study are the incident reports from the six residential complexes of the Arizona State University Tempe campus. Data from August 2018 to December 2018 is contained in the dataset. The data for five of the six residential areas were reported by master’s degree holding, full-time, entry-level housing staff employed by Arizona State University. Reports from one of six residential areas were reported by master’s degree track, part-time graduate assistant staff members. Students in four residential areas are primarily first-year students, with the remaining two areas housing upper-division students. The majority of students housed on-campus are first-year students.
No human participants are used in the study. The secondary data source is generated from existing records which report the time, type, and frequency of after-hours incidents on the Arizona State University Tempe campus. The full dataset consists of approximately 850 total records from August 2018 to December 2018. These records were already in possession of the researcher as a member of the on-call rotation for one of the residential areas and were chosen for convenience. Individual records do not identify students using personally identifying information rather using the naming convention of designating various persons as “resident”, “student”, “guest”, or other appropriate categorizations. Student genders are also obscured with neutral singular or plural pronouns. Data reported is only representative of aggregated, categorized data and contains no student identifiers.
A quantitative, descriptive approach was chosen for this study in order to research the total occurrences of various critical incidents, their locations, dates, and times. Conversations with my supervisor indicated a lack of this data internally for ASU University Housing, thus this descriptive study seeks to provide greater understanding to practitioners for better understanding critical incident occurrence.
I will collect data via a secondary source, gathered through the emails that are sent each morning by on-call University Housing staff members. These emails consistently contain at least the following elements: Location, Date, Time, Incident Type, and Incident Description. The reported dataset will contain the Location, Date, and Time as-is from the reports. I will code Incident Type and Description into the following categories, simplified from University Housing general use categories: Alcohol and Other Drugs (AOD), Facilities Emergency, Facilities Non-Emergency, Fire Alarm, Mental Health, Non-Mental Health Hospitalization, Non-Roommate Physical or Verbal Altercation, Roommate Physical or Verbal Altercation, Theft, and Wellness Check. Incident topics that do not fit into these categories will be placed into an “Other” category unless their topic count exceeds five total incidents, at which point the topic will be added to the reportable dataset. Usage of the data was confirmed with Danielle Sosias, Assistant Director for Greek Leadership Village, in ASU’s Office of Fraternity and Sorority Life, also the researcher’s director supervisor.
I will randomly sample incident reports from the full dataset using a random number generator, for a total of 10% of the approximately 850 records. Results are intended for generalization to the Arizona State University Tempe campus during the fall semester, August 2018 to December 2018.
I will analyze reports with descriptive statistics of mean, median, and mode. I will present frequency tables for incident types, location, day of week, and time. Incident types are reported as raw numerical values and as a percentage of total incidents categorized as noted above. Locations are reported as the six locations where on-call reports come from: Barrett Complex (BRC), College of Liberal Arts and Science/McClintock/University Towers (MCUT), Greek Leadership Village (GLV), Hassayampa/Arcadia/Adelphi/Sonora (HAAS), San Pablo/Tooker House (TSB), Vista del Sol/Villas at Vista del Sol (VVDS). Time reporting will contain four equal blocks of time determined by best fit for the data, both as a raw number and as a percentage. Day of the week will contain the raw number and percentage counts of incident occurrence on each day of the week. Descriptive statistics are reported using both raw and percentage values for each of the measures. I will perform crosstabulations on the following two items: Incident Type and Incident Location (Table 1) and Incident Location and Incident Day (Table 2). Sample tables are included in the Tables section for proposed crosstabulations.
This research looks to create a better understanding of Arizona State University Housing critical incidents that benefits housing professionals by improving their crisis management tools. Within the department, it might allow for on-call staff members to seek out additional resources in the form of knowledge capital or supplemental staff for their location to better support the success of all residential students. This might allow several students to be afforded assistance to further their education in ways that might not have been noticed and resolved without this research. By sharing the findings with housing staff of all levels, it might create more awareness of the incidents that one should be prepared for. That awareness might also impact potential informational programs or outreach done to the community, which could have a broader impact for all residential students. For the topics of mental health and substance abuse, the researcher hopes that seeing the total amounts of those incidents will create an awareness and discussion around improved methods for resolving those concerns.
Limitations of this study include: only analyzing one semester’s worth of incidents, only using the short incident reports rather than full detailed logs, potential human errors or bias in the secondary data, and potential sampling errors. Limited amounts of reports were available to the researcher and time constraints prevent the analysis of more reports. The collected dataset concerned only the types and times of incidents, ignoring the human element involved in the incidents themselves. As the data was gathered using secondary sources, the reliability and biases of the individuals writing the incident reports may affect the results in unknown ways to the researcher. Finally, while the data was sampled randomly, errors in sampling such as the initial sorting of the data and coding reliability may limit the usefulness of the results.
Further research should explore a significantly longer time period, perhaps three to five years, coded with multiple researchers and using student data like gender and major to explore extended trends and a more generalizable model of crucial incident occurrence. More research could also be done in that vein on multiple comparable campuses, to discover if the trends are generalizable to many college campuses or only each individual campus. Finally, research indicated the need for predictive analysis of crisis management data, and a next step of this descriptive analysis could be gathering and analyzing critical incidents to generate a predictive model (Keller & Hughes, 2011). Such a model may be able to provide researchers with methods to explore the theory of optimal responses, as well as providing practitioners with a tool that would allow proactive, rather than purely reactive, responses.