key: cord-0905699-854rzbrx
authors: Dohaney, Jacqueline; de Roiste, Mairead; Salmon, Rhian A.; Sutherland, Kathryn
title: Benefits, barriers, and incentives for improved resilience to disruption in university teaching
date: 2020-05-29
journal: Int J Disaster Risk Reduct
DOI: 10.1016/j.ijdrr.2020.101691
sha: e825c5643ce153ea0e1130bc3fa8b72b9cc22d1a
doc_id: 905699
cord_uid: 854rzbrx

Pandemics, earthquakes, fire, war, and other disasters place universities at risk. Disasters can disrupt learning and teaching (L&T) for weeks to months or longer. Some institutions have developed business continuity plans to protect key organisational services and structures, allowing L&T to continue. However, little research touches on how academics, learners, and communities of practice might respond before, during, and after disasters and how their resilience to disruption can be fostered to reduce impacts on L&T. In this research, we investigated academics’ perceptions of building resilience to major L&T disruptions in the New Zealand context. Specifically, we explored how academics characterise a resilient academic and institution, and identified the benefits, barriers, and incentives to building resilience. We used a pragmatic theoretical approach with a mixed methods methodology, to categorise the results within three distinct levels (individual, school/department, and institution), supporting the design and implementation of resilience-building strategies for academics and institutional leaders. We found that support, community, leadership, and planning at universities are critical in building and inhibiting resilience. Participants reported several ‘high impact’ incentives, addressing multiple barriers, that could be used to kick-start resilience. Online and flexible learning are key opportunities for resilience-building, but universities should not underestimate the importance of face-to-face interactions between staff and learners. Our results provide a strong starting point for practitioners and researchers aiming to understand how universities can foster resilience to major disruptions and disasters on university teaching.

In this research, we investigated academics' perceptions of building resilience to major 32 L&T disruptions in the New Zealand context. Specifically, we explored how academics 33 characterise a resilient academic and institution, and identified the benefits, barriers, 34 and incentives to building resilience. We used a pragmatic theoretical approach with a 35 mixed methods methodology, to categorise the results within three distinct levels 36 ( individual, school/department, and institution), supporting the design and 37 implementation of resilience-building strategies for academics and institutional leaders.

We found that support, community, leadership, and planning at universities are critical 39 in building and inhibiting resilience. Participants reported several 'high impact' 40 incentives, addressing multiple barriers, that could be used to kick-start resilience.

Online and flexible learning are key opportunities for resilience-building, but 42 universities should not underestimate the importance of face-to-face interactions 43 between staff and learners. Our results provide a strong starting point for practitioners 44 and researchers aiming to understand how universities can foster resilience to major 45 disruptions and disasters on university teaching.

Earthquakes, fire, war, pandemics, and other disasters place universities at risk. Disasters can 50 disrupt learning and teaching (L&T) for weeks to months, and even longer. Such significant 51 disruptions are likely to multiply with increasing urbanisation, over-population and changing 52 extreme weather and health crises, in particular, the H1N1 flu pandemic in 2009 (Ekmekci & 77 Bergstrand, 2010) . 78

Indeed, the differing conceptual approaches embraced by individuals and institutions can 79 determine the framework implemented which will in turn influence their chosen actions and 80 priorities (Linkov and Trump 2019) . For example, academic continuity has often focused, 81 almost exclusively, on the use of digital technology to provide L&T during the event. 82

Ensuring a common understanding of resilience is, therefore, important in exploring its 83 conceptualisation and possible manifestations within higher education institutions. 84

As institutional scale resilience, or business continuity, has been a focus in the literature, the 85 effects of L&T disruption on individual teaching practice has received little attention. This 86 focus on the organisation is not unique to the academic sector and resilience is often explored 87 at societal, community, or organisational scales (Linkov and Trump 2019) . In universities and 88 other tertiary institutions, academics are responsible for how courses are delivered, materials 89

presented, and what content is included. Additionally, current institutional approaches tend 90 to be responsive, rather than preventive, and focus predominantly on reducing the 91 institutional economic risk following an event (e.g., wide scale online provision in the wake 92 of COVID-19 when face-to-face tuition was impossible). Understanding the role of resilience 93 in higher education and the difficulties faced by academics in making their practice resilient 94 facilitates ongoing resilience rather than responses to singular events. 95

This article explores academic and organisational resilience in university teaching, 96 specifically academics' perceptions of the benefits, barriers, and incentives to building 97 resilience. Our investigation is supported by two key research questions: 1) How do 98 university staff describe a resilient academic and a resilient institution?, and 2) What do 99 university staff describe as the benefits, barriers, and incentives to building resilience? We 100 explore both institutional and academic continuity but look further into individual perceptions 101 of resilience-building in the university context. 102

We applied a mixed methods methodology grounded in a post-positivist pragmatic approach 104 (e.g., Feilzer, 2010) using interviews, focus groups, and questionnaires. The key aim of the 105 research was to identify key problems when L&T is disrupted and provide solutions leading 106 to improved resilience. Approval for this research was granted by Victoria University of 107 Wellington (VUW) human ethics committee (#22950). In this section, we first describe the 108 context in which this research was undertaken, followed by the participants involved, and the 109 methods for data collection and analysis employed. 110

VUW is distributed across three campuses within Wellington in New Zealand and several 112 smaller locations across the region. It is situated within an active earthquake zone and is at 113 risk from several natural disasters (Johnston et al., 2013) , but, until 2010/11, the potential 114 impact to university L&T at VUW had not been well explored. 115

In 2010 and 2011, the Canterbury earthquake sequence occurred causing fatalities, significant 116 damage to the built and natural environment, and disruption of social and economic activities 117 across the Canterbury region in New Zealand (Potter, Becker, Johnston, & Rossiter, 2015) . 118

The scale of the disruption on the local university (the University of Canterbury) (Healey, 119 2011) , and other earthquake events in New Zealand, increased the awareness of potential 120 impacts to VUW. In response, the Pro-Vice-Chancellor of Sciences, Engineering, 121

Architecture and Design (SEAD) established a resilience Steering Group (i.e., a community 122 of practice and research project) to explore how we can prepare staff and learners and 123 promote L&T continuity in the event of significant disruption. The group ran from July 2014 124 to August 2017 and explored issues, such as post-disruption access to infrastructure and  125   services, communication, staff and student support, and existing resilience and contingency  126 plans. Here, we present our investigation into how the Steering Group, and other university 127 staff, perceived resilience to disruption and its improvement. The most likely scenario 128 envisaged was an earthquake but other possible disasters were also considered, including 129 pandemics. 130

Our participants were sampled from two groups of people. First, was the resilience Steering 132 Group (SG) consisting of 8 academics across the SEAD faculties (including the second and 133 third authors of this paper), 6 professional central-service unit staff, and a resilience 134 coordinator (including the first author of this paper). The SG were asked to participate in the 135 research via email, and ten took part (5 academics, and 5 professional staff). The second 136 group of participants were SEAD academics, not members of the Steering Group (NSG) (n= 137 8). The purpose of engaging this second group was to: 1) assess perceptions of those who had 138 not been immersed in the topic, 2) raise awareness of resilience to disruption across SEAD, 139

3) share ways to build resilience, and 4) gauge their initial level of 'buy-in' while being new 140 to the initiative. We aimed to sample across the disciplines, length of teaching experience, 141 and workloads. NSG were recruited via an institutional e-newsletter and L&T community e-142 mailing list. To meet our cross-discipline sample, three participants were recruited directly by 143 email. 144

All participants (n=18) were asked to complete a hardcopy questionnaire at the start of the 145 interview. Questions included: gender, age, nationality, ethnicity, discipline, teaching and 146 research workload, and teaching experience. Participants included equal numbers of women 147 (9) and men (9), a wide range of origin nationalities (11 from NZ and 7 overseas), and a 148 broad age-spread (30-65, median of 44). The ethnicity of the participants was predominantly 149

European New Zealander or European, and one Māori participant. 150

All participants were permanent (i.e., ongoing) staff (13 academic, 3 professional, and 2 151 teaching-only roles) across a wide range of academic disciplines: architecture and design (1), 152 atmospheric science (1), geography (1), biology (2), science history (1), chemistry (2), 153 software engineering (1), statistics (1), mathematics (2) , and psychology (2), educational 154 technology (3), and fine arts (1). Participants also reported a range of teaching experience (0-155 33 years, median of 11), mostly teaching-heavy course loads (15), with a range of 2-7 courses 156

per year and all but one of these 15 participants co-teach at least one course. Overall, 157 participants reported diverse workload proportions with teaching-dominant (6), service-158 dominant (5), or distributed evenly amongst research/teaching/service (7). 159

We interviewed participants in one-on-one and focus group formats (lasting 40-120 mins). 161 SG participants were interviewed first (2 focus groups, and 4 one-on-ones), followed by NSG 162 participants (1 focus group, 5 one-on-ones). The resilience coordinator (Dohaney) led the 163 interviews with the participants and was known to the SG but not known to the NSG, and had 164 personally experienced long-term L&T disruption during the Canterbury earthquakes. 165

The semi-structured interview protocol (see supplementary material) focused on (Part 1) 166 characterisations of resilience to disruption, and (Part 2) benefits, barriers, and incentives to 167 building resilience. Part 1 was used as a meaning-making experience for the SG, allowing us 168 to form characterisations of resilience to disruption for communication with the wider 169 community. Part 2 aimed to characterise the benefits, barriers and incentives of resilience-170 building. Part 1 results were coded, summarised, and shared with NSG participants to 171 introduce the resilience initiative and help them understand the context of the project, before 172 proceeding to Part 2. During the interviews, we asked interviewees to take several moments 173 to themselves to read and think about the questions and then write down (hardcopy) their 174 thoughts before discussing them. These participant notes (including the same question 175 prompts) were considered primary data and collected when the interview was completed. 176

Participant notes were collected to triangulate the verbal data and capture individual views 177 (within the focus group format). Interviews and participant notes were de-identified and 178 transcribed by an external research assistant prior to data analysis. 179

We initially approached the datasets holistically and then interrogated the data at different 182 levels (micro, meso, and macro). The first author completed the data analysis (using 183 ATLAS.ti qualitative coding software) using a conventional approach to content analysis 184 (Cohen et al., . Using a pragmatic approach, we documented a wide range of 185 unique perceptions prioritising comprehensiveness over depth of description. 186

The participant's notes were analysed first as they represent individual views and because 187 participants often wrote responses in bulleted lists, which helped to create quick, first-188 impressions of the data. Next, the interview data was coded, and initial codes assigned. The multilevel theory approach allows researchers to investigate phenomena within and between 201 major levels (or scales) of a system. This approach appealed to the researchers, as we 202 intended to operationalise our findings into resilience-building initiatives within the existing 203 levels of the university socio-political-system. We, therefore, organised and presented the 204 themes by level. This approach does not follow a systems analysis (i.e., identifying 205 vulnerabilities within a system) but retains the focus on academics' perspectives 206

Once a coding scheme was developed, the second and third authors checked the codes to 207 support code generation, clustering, and scheme refinement, while the fourth author (who had 208 no prior involvement in the project) checked the final code scheme through independently 209 coding a full transcript (randomly-chosen participant), resulting in minor additions to one 210 code category and rewording/clarification of another. Code frequencies were used to guide 211 our aggregation, clustering or parsing of information. Several codes had low frequencies but 212 clustering data in these areas reduced comprehensiveness of a resilience planning/building 213 solution thereby reducing its utility (i.e., supported by our pragmatic approach) to building a 214 solution for increased resilience. Code categories and sub-categories were checked for 215 statistical relationships between subpopulations (i.e., Did participants with more teaching 216 experience report more learner-centred benefits to resilience?) but no significant relationships 217 were found; consequently, the findings are not presented by subpopulation. 218

The results and discussion are presented within the four major themes identified in our data 220 analysis: our community of practice's characterisation of resilience to disruption, followed by 221 the benefits, barriers, and incentives to building resilience. The following sections of the 222 paper are intended to assist the thinking and planning of university academics, staff, and 223 leaders as they consider their own local context. Our tables (Tables 1-3), in particular, offer  224 individual academics, schools, and academic institutions many factors which influence 225 resilience to disruption. Longer excerpts of text supporting Tables 1-3 are available in the  226 online supplementary material. 227

Here, we present the SG's characterisation of academics and higher education institutions 229 resilient to disruption, within our institutional and cultural context. Both characterisations 230

were developed as shared ideals rather than rigid standards, or definitions, and should reduce 231 L&T disruption regardless of the severity of the disruption. 232

Resilient academics (in the order of most frequently mentioned codes): 233

• are flexible, adaptable, emotionally-resilient, collaborative, empathetic, open-minded 234 individuals (attributes); 235

• respond quickly during a disruption, are digitally literate, organised, prepared and 236 creative-thinkers (capabilities); 237

• have a sound awareness of their courses, learner-centred approaches, L&T delivery 238 options during disruptions, emergency protocols, and the wider institutional system 239 (knowledge). 240

Resilient institutions have the following characteristics (in the order of most frequently 241 mentioned codes): 242

• effective communication channels, 243

• a coherent crisis communication strategy, 244

• an established, coherent, L&T disruption plan across all levels of the institution, 245

• strong resilience-building leadership, 246

• existing emergency response plans and management, 247

• existing flexible, blended, and digital learning strategies, 248

• support for staff to undertake resilience-building initiatives, 249

• support for staff to develop digital literacy, 250

• effective and easy-to-use digital infrastructure, 251

• a strong sense of staff and learner community, and 252

• existing rewards/schemes to promote engagement with academic professional 253 development (aligned to resilience qualities, above). 254

The characterisations contain major themes that emerged throughout the study and include 255 (in no particular order): flexibility, communication, community, support, strategic planning, 256 preparedness, and leadership. Both sets of characteristics highlight flexible learning 257 pedagogies (i.e., L&T delivered through a variety of modes, contexts and settings; Errington, 258 2004 ); therefore, flexible learning research and initiatives may provide a critical remedy for 259 building resilience to disruption. However, the characterisations also include sustainable and 260 long-term resilience measures (Kapucu & Khosa, 2013 ) that shift the focus away from short-261 term (threat-dependent) solutions, advocated by researchers in business continuity 262 community. 263

It is also notable that these characterisations do not rely on a risk assessment, often 264 interwoven with more organisational scale approaches (e.g., Linkov et al. 2018) . We consider these characterisations as a starting place for understanding resilience to 273 disruption and useful to individuals and groups seeking to improve resilience at their 274 institution. No institution is alike, and these characterisations come from a specific university 275 within New Zealand. Each university and individual are likely to experience different 276 resilience dimensions to a greater or lesser extent. However, these sets of characteristics 277 provide a firm starting point which can be provided to other higher education institutions and 278 individuals to explore their understanding of resilience and upon which a customised plan, 279 taking into consideration their unique vulnerabilities and areas for improvement, can be 280 created. 281

Building from these characterisations of resilient academics and their institutions, we here 283 identify the benefits our participants perceived in preparing for disruption (i.e., increasing 284 resilience), the barriers faced when building resilience, and incentives to help overcome these 285 barriers. While the responses are specific to our institution, we expect these dimensions to be 286 Recognising these stages and acknowledging that each stage brings different challenges, we 291 conceptualise and investigate resilience as how individuals, schools, and universities perceive 292 both preparing for and responding to disasters. 293

Increasing resilience by preparing for a major disruption has benefits in the immediate and 295 longer-term aftermath of a disaster, for both institutions and individual academics. Table 1  296 shows the ways in which the participants anticipated that preparing for disruption will benefit 297 them in their everyday practice before, during, and after such events. 298 [Insert Table 1 here]. 299

Participants described 61 unique benefits to improved resilience to disruption. In Table 1 , 300 sub-categories are labelled with the number of unique mentions by participants (in italics, 301 n=18) and codes are labelled with proportion from the total unique benefits (not italics, x/61). 302

The most frequently mentioned benefits to developing resilience to disruption were 303 academic-centric, with academics… 304

• being more organised, simplifying their everyday L&T, and working on other things 305 aside from teaching, 306

• focusing on learning outcomes, rather than course logistics, 307

• feeling in control and emotionally prepared for a disruption, 308

• better supporting their colleagues, learners, and families during a disruption, 309

• knowing what to expect and things to consider during a disruption, 310

• having the ability to pass their course on to others in the event of an illness, and 311

• being encouraged to try new L&T strategies. 312

Most benefits focus on 'during the event', followed by 'before the event'. Academic 313 developers and leaders might use this 'dual-benefit' frame (i.e., preparing for disruption will 314 benefit you both now, and in the event of a crisis) to motivate and support individuals to 315 develop resilience over the long-term and with shorter-term disruptions (e.g., going on 316 sabbatical, illnesses). 317

We also differentiated benefits by key impacted groups (academics, learners and institution) 318

with the most frequently mentioned concerned with academics' competency and emotional 319 state, followed by benefits to L&T, and to the institution. With the dominant themes pointing 320 to a personal and professional development aspect, building resilience can be directly 321 beneficial to the individual, as well as the institution, and can be communicated this way. 322

Participants also report that building resilience could reduce the emotional impact of 323 disruptive events giving academics more control over what is happening, allowing them to 324 support their learners, colleagues, and families. Perhaps through resilience-building efforts 325 academics might better prepare themselves, emotionally and cognitively, and subsequently 326 reduce the impact of potential events. This is promising, as Kemp et al. (2011) report that 327 people with lower levels of emotional stability were more affected by the earthquakes in 328

Canterbury. 329

On the other hand, participants did not focus on the benefits to the university nor to students. 330

This tells us that the academic's perception of resilience-building described here would also 331 benefit from student engagement. Beaven et al. (2014) report that students want to be a part 332 of the solutions rather than 'dictated to' in times of crisis. 333

There were no significant differences in how the two groups (SG and NSG) described 334 benefits. Participants new to the idea of improving educational resilience noted the same 335 number and types of benefits as those who had spent two years working on the project. Also, 336 academics across the range of teaching experience identified the same number and types of 337 benefits. None of the SG or NSG participants were opposed to, uninterested in, or dismissive 338 of resilience-building. Promisingly, this similarity indicates that benefits to increasing 339 resilience are apparent to those without previous exposure to such thinking. It also signifies 340 that the characterisations developed with the SG above are relevant to those new to the 341 concept of resilience and will be a helpful framework to inform staff. 342

In conjunction with the characterisation of resilience for academics and institutions, the 343 benefits listed in Table 1 can have two obvious impacts: to communicate to academics why 344 they should undertake resilience-building initiatives, and to convince senior leaders of the 345 need to invest resources in such initiatives. 346

The number of benefits identified by each participant underlines the ease with which 348 resilience concepts can be communicated. However, barriers exist to building resilience. 349

Understanding these barriers can influence the planning and implementation of effective 350 resilience initiatives. The focus in this section is on barriers facing academics while 351 institutional barriers are explored in section 3.2.4. 352

The 18 participants described 56 perceived barriers to resilience-building across the different 353 levels of the institution (Table 2a- Table 2 are organised by  354 group and level, with codes listed in order of mention-frequency. Again, no significant 355 differences were reported in barriers by the two participant groups (SG and NSG) or 356 participants with varying levels of teaching experience. These findings represent strongly 357 consistent views. 358 [Insert Table 2 here]. 359

The most frequently mentioned barriers were: 360

• Lack of staff time, 361

• Lack of institutional mandate, buy-in and acknowledgement, 362

• Poor staff and learner digital literacy, 363

• Existing digital systems are limited (i.e., poorly performing and lack resilience), 364

• Academics' unwillingness to change, adapt, and be flexible, 365

• Lack of a school-level plan for improving resilience and responding to disruption, 366

• Lack of cohesive and nurturing institutional community, 367

• Lack of incentives to encourage resilience initiatives, and 368

• Academics lack of resourcefulness in finding solutions during disasters. 369

The most common barrier mentioned is supported by research indicating that academics 370 report feeling overworked with fewer resources (Adams, 1998 ) and sustaining high 371 workloads. Therefore, academics will require ongoing support and incentives to ensure 372 resilience-building initiatives are undertaken or maintained. 373

Post-earthquake research (Mackey et al., 2011) indicates that lack of digital literacy among 374 staff, as identified in Table 2 #18, hinders institutional efforts to respond during and after 375 disruption. Notably, the adoption of digital learning relies on the integration of new and 376 existing tools, staff, and protocols (Blin & Munro, 2008) so to develop resilience, ongoing 377 efforts are needed. Also, some teachers moving to flexible or distance learning strategies may 378 express fears about losing ownership over their educational materials ( Table 2 #8 : 'Academic 379 freedom'; #10: 'Distrustful of digital technology') and a potential lack of 'real' contact with 380 students (Errington, 2004) . 381

At the individual-level, several barriers are likely to be difficult to change (e.g. Table 2 #6 382 'Unwillingness to adapt, change, or be flexible', #11 'Inability to work with others' and #13 383 'Limited pedagogical perspectives') and could be considered 'innate traits'. 384

Academics prioritising family and personal life over work responsibilities ( Table 2 # Interestingly, discipline-based barriers to resilience were much less reported than initially 406 anticipated by the researchers. This result reflects Krause's (2014) contention that when it 407 comes to L&T, there may be less of a "hard line" between disciplines than for research. 408

We also asked participants to identify potential incentives to reduce barriers and enhance 410 resilience across the institution. Table 3 provides readers with 28 strategies that an institution 411 can use to promote and engage staff in resilience-building. Many have cost or time 412

implications, but often address multiple barriers. The number of barriers addressed by each 413 incentive are also listed in the right-hand column of Table 3 . 414 [Insert Table 3 here]. 415

The top three incentives were: 416

• providing one-on-one staff academic/educational development support, 417

• providing buy-out for academics to explicitly focus on these initiatives, and 418

• university leadership demonstrating support and endorsement for resilience initiatives 419

Barriers to resilience-building and the incentives to overcome these barriers are intrinsically 421 linked. The key themes identified, how each barrier can be addressed, and a detailed 422 description and consideration of how key incentives can be enacted are described in this 423 section. 424

Lack of community was a prominent barrier at the school-and institutional-level identified 425 by most participants ( Table 2 ). Communities of practice (Wenger, 1999 ) are known to 426 support changes in L&T practice (Vescio, Ross, & Adams, 2008) , enhance innovative and 427 flexible L&T learning approaches (Errington, 2004) , and are critical for supporting learners 428 in a crisis (Mackey et al., 2011) . Additionally, research has shown that in times of crisis, 429 community support (or the perception thereof) is strongly related to 'people feeling normal 430 again' (Bonanno et al., 2010; Gill, 2007) . We recommend that universities incorporate 431 community-building efforts into their resilience plan, such as shared activities in L&T 432 innovation (Cherrington et al., 2017) . 433

Five teaching-dominant participants in our research explicitly linked the lack of cooperative 434 institutional-level community and culture at our university to the lack of recognition and 435 importance of L&T. Similar to other academic institutions worldwide (Adams, 1998) , L&T 436 was not perceived as a key criterion for promotion. Also, many academics are overworked 437 with less resources (Adams, 1998) and L&T and research are seen, by some, as adversaries 438 (Roberts, 2015) . To support resilience initiatives, acknowledgement, endorsement, and 439 engagement from senior leaders is needed (Table 2c, 2d, 3b) . Senior leaders can support 440 resilience efforts in many ways, namely through reward and recognition schemes for L&T 441 community initiatives (supported by results here and also in Krause, 2014) . For example, our 442 university redesigned its promotion criteria to more explicitly acknowledge L&T in 2018. 443

Notably, while it may be appealing for leaders to use the results of this work to enact 'top-444 down' changes, a divide often exists between managerial approaches to higher education and 445 academics' desire for autonomy (Ramsden, 1998; Winter, 2009) Within a resilience quantification framework, targeting particular academic disciplines may 460 make the most sense, e.g. those with costly teaching laboratories. If such an approach is 461 followed, the incentives and barriers identified here may be helpful in working with academic 462 staff in target high risk areas to purposefully increase these strategic areas of resilience. 463

Resources and support are the backbone of L&T professional academic development. Staff 464 and learners need to be digitally capable, and infrastructure (virtual and physical) is needed to 465 support flexible learning. Lacking both these elements poses a significant barrier to 466 resilience. As learning management system technology evolves or courses move from 467 emergency remote teaching to a purposefully designed online model, staff and learners will 468 require continual training. Not all staff will want to engage in flexible learning with the 469 perceived loss of ownership over their courses a noteworthy fear among academics 470 (Errington, 2004) . Resilient higher education institutions should prioritise support services 471 integral to long-term, sustainable changes (Toohey, 1999) . Professional development in 472 resilience can be facilitated through workshops or one-on-one support with educational 473 technologists and academic developers (Table 3 A1 ), recognising that individual support is 474 less possible in the immediate aftermath of a disaster event due to high demand. We also 475 note, though, that professional development may not be appropriate for all perceptions and 476 behaviours as those behaviours may be linked to internal values and resistant to change 477 (Table 2) . 478

Implementing L&T in some disciplines requires more physical infrastructure than wholly 479 online, flexible, or distance learning approaches can provide. For example, conventional 480 geoscience education relies on face-to-face hands-on laboratory lessons. This infrastructure 481 could be damaged during major disruptions, or access could become limited. When 482 discussed, participants perceived this possibility as an opportunity to develop alternative 483 settings/modes of L&T rather than an insurmountable challenge. For example, a blended 484 learning approach can provide some physical access to materials and equipment, supported 485 with online learning opportunities. We recommend using the online learning environment to 486 foster unique approaches, to increase learner engagement and maintain a 'minimum online 487 presence' (i.e., contact information, unit outlines, weekly agenda, and communication 488 channels), which learners can use to communicate with peers and staff. We see this 489 meaningful online environment as the first 'goal post' towards resilience, even if the online 490 space is simply an alternative for face-to-face activities (e.g., delivery of lectures, readings 491 and assessments). 492 Furthermore, we assessed which incentives might be the most potent in reducing barriers to 493 resilience. The authors matched and summed the number of barriers addressed by each of 494 these incentives (Table 3 , right-hand column; Supplementary Information) . Some barriers 495 could be more 'easily' reduced or mitigated, others much less so, and 7 barriers had no 496 incentive provided by the participants (Table 3 #5 , 6, 11, 30, 33, 35, and 36) . However, most 497 of the incentives addressed more than ten barriers, with three "high impact" incentives of: 498

• Sharing L&T practice within schools to encourage capacity-building, 499

• Practicing crisis scenarios, and 500

• Buy-out: providing time from other duties to focus on resilience-building. also talked extensively about the incentive of buy-out (i.e., using funding to pay for teaching 507 or administrative support) providing academics much needed time to test resilience 508 initiatives. In New Zealand, buy-out is commonly used to focus on research. We propose that 509 'resilient teaching buy-out' could be a powerful experience, with far-reaching impacts. We 510 suggest institutions trial buy-out schemes to give staff with the greatest vulnerabilities to 511 disruption (i.e., highest student numbers, most dependent on physical infrastructure, caring 512 responsibilities, etc.) or early adopters (i.e., engaged staff likely to implement initiatives and 513 inspire others around them to build resilience) the time to creatively identify, plan, and 514 implement unique resilience-building solutions. 515

At a minimum, we recommend that a business continuity plan is co-created by academics, 516 senior leaders, and business continuity experts. To promote longevity and sustainability, 517 planning needs to take a long-term view (Kapucu & Khosa, 2013) and leaders need to buy-in 518 and promote resilience initiatives. For those seeking guidance in campus responses to 519 disasters, there is a range of existing research which discusses universities' preparedness and 520 overall response to crisis and organisation-level communication management and planning 521 (Beaven et al., 2014; Dabner, 2012; Fillmore et al., 2011; Palen, 2008; Seville et al., 2011; 522 Tanner & Doberstein, 2015) . A prepared and responsive university would help to ensure that 523 the institution continues to function as a business, that learners continue their studies, and 524 therefore staff and student wellbeing can be prioritised (Bates, 2013; SchWeber, 2013) . 525

This study explored academics' perceptions of resilience to disruption and documented the 527 perceived benefits, barriers, and incentives of resilience-building. Our approach contributes 528 to the literature through its multidisciplinary lens by including participants across a spectrum 529 of disciplines (e.g., including psychology, geography, biology, chemistry, and science 530 history), with different teaching commitments, and by including both academic and 531 professional staff. The collation of responses across different academic disciplines and the 532 similarity of those responses across our participants for L&T demonstrates the wider 533 applicability of our results. Our focus on academics enhances the understanding of impacts 534 on and incentives for resilience on individual practice. However, our focus on individual 535 practice and perceptions does not replace the need for an organisation-level assessment. 536

Further, higher education is infrequently explored and investigated within the resilience 537 literature. We applied a pragmatic approach and mixed methods methodology within a 538 university in New Zealand, though we propose that the results can be used to build resilience 539 in many other higher education contexts. 540

The key findings and discussion are grouped into the three distinct levels of the university 541 (individual, school, and institution), so that resilience-building strategies are readily 542 facilitated by individuals, schools, and institutional senior leaders. Tables 1-3 are presented in 543 a way that allows direct application to practice, reducing the burden for individuals, and 544 streamlining the call-to-action. The benefits of continued resilience to disruption for new and 545 existing staff outside of an immediate threat need to be understood to ensure resiliency gains 546 are built upon, rather than lost, in a return to business as usual. Awareness of the continued 547 benefits to resilient teaching may encourage the continued engagement of academics in 548 rethinking and reprioritising their course delivery and teaching practices to ensure continued 549 resilience. 550

Our research highlights the importance of support, community, and leadership at universities. 551

We also present dual-benefits of implementing flexible learning strategies, which can support 552 meaningful L&T even if face-to-face interactions cannot take place after a disruption. This 553 enables the framing of resilience initiatives as meaningful, regardless of whether an event 554 happens or not. However, if a disaster does occur, we should not dismiss face-to-face 555 interactions where possible (e.g., earthquake rather than pandemic scenario), because 556 stronger community connections are important for 'feeling normal again'. 557

Resilience to disruption will be experienced differently across educational and socio-political 558 contexts, where benefits, barriers, and incentives may be of greater/lesser importance. Every 559 university, therefore, needs to build a customised plan taking into consideration its unique 560 vulnerabilities and areas for improvement. Such quantification methods have been discussed 561 elsewhere. Universities comprise complex, tiered relationships between the levels 562 (individuals, schools, and leaders) that are negotiated through communities of practice with 563 unique decision-making and power distributions. It may be appealing for senior leaders to use 564 the results of this work to enact 'top-down' changes in response to these, commonly, 565 institution-based frameworks, but we recommend a blended strategy (top-down and bottom-566 up) that supports academics' autonomy and diversity in L&T. 567

In most contexts, we cannot change that disasters happen but we can change how vulnerable 568 our universities are to these events and we can reduce the impacts to our people and property. 569

By documenting barriers and highlighting benefits and incentives, we hope academics, senior 570 leadership, and professional staff can enact changes across all levels of their institution, 571 ultimately leading to reduced risk from disaster and long-term disruptions. 572

We hope that this research will benefit those currently working to continue L&T despite the 573 overwhelming challenges presented by the COVID-19 pandemic and other disasters. 574 575   Table 1 . Benefits to Improving Resilience 576 Before a disruption, (22) During (34) be focussed on learning outcomes rather than course logistics. (9) be flexible to changing delivery modes. (4) be able to continue, during the disruption. 

There is no financial interest or benefit that has arisen from the direct applications of this research. There are no competing interests to declare. The researchers declare there are no personal relationships with people or organisations that could inappropriately influence this research.

Lack of institutional mandate, buy-in, and acknowledgement (15/18)

Unwillingness to adapt, change, or be flexible (11/18) 7. Not

Overwhelmed by the resilience-building task (7/18) 10. Distrustful of digital technology (7/18) 11

Family/personal life prioritised over work responsibilities (6/18) 13. Limited pedagogical perspectives (5/18) 14. Research prioritised over teaching (4/18) 15. Resilience professional development sounds boring (4/18) 16. Lack of recognition and motivation for L&T changes (3/18) 17

Competencies (14/18) 18. Poor staff digital literacy (12/18) 19. Lack of resourcefulness

Community & Culture (16/18) 21. Lack of shared L&T culture

15/18) 25. Lack of school-level resilience planning (11/18) 26

Inter-dependency of staff (5/18) 31. Lack of L&T tools, equipment, lecture materials (5/18) 32. Uneven workload (4/18) 33. Inter-dependency of courses (3/18) 34. Lack of staff cover for teaching (3/18) 35. High staff turnover (3/18) 36. Over reliance on tutors and teaching assistants (3/18) Discipline-specific (7/18) 37. Different disciplinary approaches to resilience (5/18) 38

Poor institutional leadership (6/18) 40. Lack of longevity in resilience vision and efforts (6/18) 41. Resilience as a standalone initiative; not embedded (5/18) 42. Bureaucracy (3/18) Infrastructure (14/18) 43. Limited digital systems, structures and processes (12/18) 44. Limited physical systems

Lack of rewards to encourage resilience initiatives (10/18) 47. Lack of staff expertise, resources and training (8/18) 48. Lack of funding and material resources

Lack of pastoral care model for learners (8/18) 50. Learners are not digitally literate (5/18) 51. Learners have diverse needs (4/18) 52. Learners are expected to be ready-to-learn

Communication (9/18) 55. Ineffective communication between senior leaders and community (7/18) 56

Examining the fabric of academic life: An analysis of three decades of research on the 581 perceptions of Australian academics about their roles. Higher Education

Institutional continuity and distance learning: A symbiotic relationship

Risk and resilience 588 factors reported by a New Zealand tertiary student population after the 4th

Research engagement after disasters: 593 Research coordination before, during, and after the 2011-2012 Canterbury Earthquake sequence

Why hasn't technology disrupted academics' teaching practices? Understanding 597 resistance to change through the lens of activity theory

Weighing the costs of disaster

Consequences, risks, and resilience in individuals, families, and communities. Psychological Science in the 602 Public Interest

Developing a pan-605 university professional learning community

Research methods in education

Breaking Ground' in the use of social media: A case study of a university earthquake 611 response to inform educational design with Facebook. The Internet and Higher Education

Training in crisis communication 615 and volcanic eruption forecasting: Design and evaluation of an authentic role-play simulation

Using role-play to improve students' 619 confidence and perceptions of communication in a simulated volcanic crisis

Observing the volcano world

Preventing 624 School Failure: Alternative Education for Children and Youth

Agility in higher education: Planning for business continuity in the face 627 of an H1N1 pandemic: Quarterly journal

The impact of teacher beliefs on flexible learning innovation: Some practices and 630 possibilities for academic developers

Doing mixed methods research pragmatically: Implications for the rediscovery of 634 pragmatism as a research paradigm

After the waters 638 receded: A qualitative study of university official's disaster experiences during the Great Iowa Flood of 639

The Oxford 642 Handbook of Women and Economy

Secondary trauma or secondary disaster? Insights from Hurricane Katrina

A multilevel approach to building and leading learning organizations. The 648 Leadership Quarterly

The 2010 and 2011 Canterbury earthquakes and organisational learning at the University of 651

Canterbury: Does practice make perfect

The difference between emergency remote 655 teaching and online learning

The experiences of faculty at academic institutions preparing themselves for academic 659 continuity after a disaster in Florida in the last decade: A phenomenological study

Community 663 understanding of, and preparedness for, earthquake and tsunami risk in Wellington

Disaster resiliency and culture of preparedness for university and college 667 campuses

Sleeplessness, stress, 670 cognitive disruption and academic performance following the

A resilience framework: Perspectives for educators

Challenging perspectives on learning and teaching in the disciplines: the academic 677 voice

What is 'social resilience'? Perspectives 680 of disaster researchers, emergency management practitioners, and policymakers in New Zealand

Tiered approach to resilience assessment

The science and practice of resilience

Riding the seismic waves: re-blending 689 teacher education in response to changing demands

Facing Hazards and Disasters: Understanding Human Dimensions

The restructuring of tertiary education in New Zealand: governmentailty, neo-liberalism, 696 democracy

Online social media in crisis events

Canterbury earthquakes

Higher education curriculum orientations and the implications for institutional curriculum 706 change. Teaching in Higher Education

Learning to Lead in Higher Education

How effects from teacher-training of academic teachers propagate into the 711 meso level and beyond

Determined to learn: Accessing education despite life-threatening disasters

Survival lessons: Academic continuity, business continuity, and technology

Facilitating learning in the 21st Century: Leading through technology, diversity and authenticity

Shaken but not stirred. A university's resilience in the face of 722 adversity

Organisational 725 resilience: Researching the reality of New Zealand organisations

Emergency preparedness amongst university students

Designing courses for higher education

The impacts of climate change on the risk of natural disasters

Teaching resilience: A narrative inquiry into the importance of 737 teacher resilience. Pastoral Care in Education

A review of research on the impact of professional learning 740 communities on teaching practice and student learning. Teaching and Teacher Education

Communities of practice: Learning, meaning, and identity

Academic manager or managed academic? Academic identity schisms in higher 747 education

Quantifying and mapping resilience within large 751 organizations