Research Article

Resilience as a coping strategy for reducing departure intentions of accounting students

Published: 2020-1-2

Journal: Accounting Education

DOI: 10.1080/09639284.2019.1700140

Abstract

## ABSTRACT This study evaluates the influence of resilience as a potential coping strategy to help reduce student departure from the accounting major. We collected data from 443 accounting majors at four geographically disbursed U.S. universities using a battery of psychometric instruments. With these data, we analyzed the relations between role stressors, psychological health, burnout, and departure intentions, and assessed the extent to which individual resilience levels served as a positive influence by enhancing health, and diminishing burnout and departure intentions. We found sources of role stress to have significant negative associations with psychological health, and significant positive associations with academic burnout (direct), and departure intentions (indirect). However, resilience counteracted those associations through its direct positive association with psychological health, and direct negative association with burnout. Resilience also had a significant indirect negative association with departure intentions through its direct associations with psychological health (positive) and burnout (negative). ### KEYWORDS: role stress, academic burnout, departure intentions, resilience ### Disclosure statement No potential conflict of interest was reported by the authors. ### Notes 1. Excessive turnover in public accounting has long been a topic of academic and practitioner interest. See, e.g. Leathers (1971), Hermanson et al. (1995), Brundage and Koziel (2010), and Chong and Monroe (2015). 2. http://www.insidepublicaccounting.com/wp-content/uploads/2018/09/The-2018-IPA-National-Benchmarking-Report.pdf. 3. In 1994, the number of test takers was 140,232, and by 2017, that number had dropped to 95,645 (Ng, 2019). 4. In a subsequent study, Byrne and Flood (2005) found among a sample of 129 first-year accounting students at an Irish university, that 84% expected (hoped) that their self-esteem would increase as a result of matriculating in the program. 5. Technically, Brown et al. (2013) found that accounting majors had a lower level of narcissism than other undergraduate and graduate business students, as well as a general population of college-age students. The authors citing Morrison (1997, p. 375) stated that ‘…in adults, a reasonable amount of healthy narcissism allows the individual's perception of his needs to be balanced in relation to others’. They also quoted DuBrin (2012, ii) who stated ‘A healthy dose of narcissism can facilitate career success, because reasonable concern with the self helps a person think of achieving important goals and being admired as a leader’. 6. Related to the concept of resilience are the concepts of ‘grit’ (e.g., Duckworth, 2016) and ‘growth mindset’ (Dweck, 2006). Prior research finds that accounting students with higher levels of grit have greater intensions to become professionally qualified accountants (Aziz et al., 2017). Meanwhile, accounting students with growth mindsets are more likely to aspire to work for a Big Four accounting firm (Liu et al., 2018) and are more capable of improving performance on exams (Ravenscroft et al., 2012). 7. The three items that comprise the role ambiguity scale (denoted as RA in Appendix A) are reverse-scored. 8. The MBI-GS[S] is a derivative of the Maslach Burnout Inventory (MBI), the latter instrument arguably the leading measure of burnout in organizational settings as evidenced by the 15,050 citations (as of November 2019) to the publication that described its development (Maslach et al., 1986). 9. We also conducted an exploratory factor analysis of the 16 MBI-GS[S] items and found that they all loaded on their respective dimensions as prescribed by Maslach et al. (2016). 10. A series of analyses of variance revealed no significant differences in any of the primary study constructs by school. Thus, the data from all four schools were combined in all of the ensuing analyses. 11. These composite indicators appear in Table 2 below as RC1, RC2, RA1, RA2, RO1, RO2, Health1, Health2, and TI1, and TI2. 12. The maximum likelihood (ML) method CFA output indicated significant multivariate kurtosis with a Mardia’s Coefficient of 48.27 and a normalized estimate of 23.62. Bentler (2006) notes that that normalized estimate values greater than 3 are indicative of nontrivial positive kurtosis, but cautions that modelling statistics may not be affected until the values exceed 5–6 or higher. We incorporated the elliptical estimation procedure to address the kurtosis issue. 13. The Wald test is not theory driven, i.e., it is a sample-specific, post-hoc procedure. Therefore, to determine if the associations identified in this investigation hold, replication with another sample is necessary. 14. Chong and Monroe (2015) state that as long as the conceptual integrity of a model is maintained, dropping non-significant parameters from a saturated structural model is an acceptable practice. 15. Hooper et al. (2008) note that the relative/normed chi-square (χ2/df) minimizes the impact of sample size on the model chi-square while acknowledging that there is no consensus regarding an acceptable ratio for this statistic, with recommendations ranging from a high of 5.0 (Wheaton et al., 1977) to a low of 2.0 (Tabachnick & Fidell, 2007). 16. For the multi-sample analysis by GPA, we split the sample into two groups: students who reported a GPA below 3.4 (n=205) and students who reported a GPA of 3.4 or above (n=234). 17. Chong and Monroe (2015), quoting Kerlinger and Pedhazur (1973, p. 19) state that this technique ‘can be used to determine whether the pattern of correlations for a set of observations is consistent with a specific theoretical formulation’. 18. Our model has its roots in Fogarty et al.’s (2000) Beyond the Role Stress Model which was drawn from the occupational health, applied psychology, and related literature. One important extension of the Fogarty et al. (2000) model that we made was to include psychological health as a burnout mediator, which follows the medical student burnout literature (Dahlin et al., 2011; Backović et al., 2012). 19. As noted in Appendix A, all three role stressors are measured on a five-point score ranging from A=strongly disagree, to E=strongly agree. 20. The relationship between hardiness and resilience can be gleaned from the authors’ (p. 566) statement that ‘hardiness has emerged as a pattern of attitudes and skills that is a pathway to resilience under stressful circumstances’. 21. Within the U.S., our findings also have noteworthy implications for public universities in states with performance-based funding policies. These policies use performance metrics such as freshman retention rates (i.e., the percentage of first-year students that continue into their second year) and six-year graduation rates to determine the level of funding provided to public universities. 22. While these admonitions are related to studies of the accounting work environment, they appear equally valid in academic accounting settings.

Faculty Members

  • Bob G. Wood - Dean of the Mitchell College of Business, University of South Alabama, USA
  • David J. Emerson - Department of Accounting & Legal Studies, Salisbury University, Salisbury, MD, USA
  • Kenneth J. Smith - Department of Accounting & Legal Studies, Salisbury University, Salisbury, MD, USA
  • Shawn Mauldin - Department of Accounting & Legal Studies, Mississippi State University, Starkville, MS, USA
  • Timothy D. Haight - Accounting, Loyola Marymount University, Los Angeles, CA, USA

Themes

  • Influence of psychological health on academic outcomes
  • Student retention in accounting
  • Academic burnout
  • Coping strategies for students
  • Resilience in education
  • Role stressors and their impact on psychological health

Categories

Download Article