Abstract
Teaching music can be very stressful. The music education literature includes investigations of sources of stress and suggestions for stress relief, although it lacks a consistent underlying conceptual framework for stress. Role stress—a framework for understanding stress related to workplace dynamics—has the potential to provide structure while accounting for the nuance and complexity of organizations. The purpose of this study was to adapt and develop psychometrically sound questionnaires and then collect data to test a model to determine relationships among role stressors, role stress, and other variables of interest. Participants (N = 1,576) responded to items related to six hypothesized role stressors, specialization and teaching loads, holding multiple jobs, years taught, job satisfaction, organizational commitment, and intent to leave the current job. I analyzed the data using descriptive statistics, factor analysis, and structural equation modeling. Key findings included (a) strong relationships among role stress and negative occupational outcomes, (b) confirmation of four known role stressors but nonseparation of role conflict and role ambiguity in statistical models, and (c) weak relationships between role stress and model covariates (years teaching, within-specialization teaching, and having multiple jobs). These results have implications for current and prospective secondary music teachers, music teacher educators, and school administrators.
Keywords
Being a music teacher can be very stressful. Music education scholars have identified common sources of stress for music teachers, including overwork (Sandene, 1995); problematic interactions with others in the school (Gordon, 2000; Stringham & Snell, 2019); emphasis on accountability (Shaw, 2016); ideological conflict with school administrators (Scheib, 2006); insufficient support—financial and otherwise (Gordon, 2000); and limitations of preservice preparation (Conway, 2002; Gordon, 2000). As a response to this stress, several authors have offered suggestions for stress reduction. Some suggestions—such as improving classroom management (Stern & Cox, 1993) and approaching performances in innovative and less stress-inducing ways (Allsup, 2005)—have addressed external sources. Others have focused internally, with recommendations of relaxed breathing exercises (Hamann & Gordon, 2000) and mindfulness (Varona, 2019). While prior study of music teacher stress and suggestions for its relief have been beneficial, there remains a need for a more consistent conceptual organization of stressors and responses to stress.
Many of the stressors identified in previous work are consistent with role stress as described by Kahn et al. (1964). Role stress is brought about by dynamics related to an individual’s position within an organization. One’s office (i.e., occupational position, such as music teacher) exists within an organization and is associated with certain roles (i.e., tasks or responsibilities, such as concert planning). Each office is related to others in its role set by common goals and resources; a music teacher’s role set includes students, parents, other teachers, administrators, and many others. Interactions among the office holder, their role set members, the definitions and scope of various roles, and the resources and needs of the organization can all cause role stress.
The construct of role stress is well established in industrial and organizational psychology literature. Rizzo et al. (1970) constructed a 14-item scale designed to measure role conflict and role ambiguity—two aspects of role stress, or role stressors. Despite some concerns about validity (Tracy & Johnson, 1981), the scale’s items and their derivatives have been used frequently and recently (Acker, 2004; Bacharach, Bamberger, & Conley, 1990; Richards et al., 2016; Schwab & Iwanicki, 1982). Such studies have typically involved investigating relationships between role stressors and other workplace variables. Researchers have consistently found role stressors to be predictors of burnout, job satisfaction, commitment, and turnover intentions (Acker, 2004; Richards et al., 2016; Schwab & Iwanicki, 1982).
This line of research has extended into schools. Reyes and Imber (1992) found that teachers’ perceptions of the fairness of their workloads (pertaining to role overload, another role stressor) were related to their morale, commitment to the school, and job satisfaction. Conley and Woosley (2000) found that role ambiguity and role conflict were related to job satisfaction and organizational commitment (“a bond or linking of the individual to the organization,” p. 183). They also found that role overload predicted lower organizational commitment for some teachers. Conley and You (2009) found a set of predictive paths from role ambiguity and role conflict—through job satisfaction and organizational commitment—to intent to leave the held position. A follow-up study provided further confirmation of these relationships, adding a statistically significant path from role overload to organizational commitment (Conley & You, 2014).
Role stressors have been discussed in music education (e.g., Pohland, 1995) but often without reference to the original framework established by Kahn et al. (1964; e.g., Hamann, 1985, 1989; Sandene, 1995). A notable exception was Scheib (2003), who conducted a multiple case study to understand music teachers’ role stress. He considered six role stressors:
role conflict, in which someone receives conflicting messages about their roles;
role ambiguity, in which someone has insufficient clarity about what their roles are;
role overload, in which someone has too much to do or too many roles to fulfill them all well;
underutilization of skills, in which someone has relevant skills that are not used in their role(s);
resource inadequacy, in which someone does not have what is needed to fulfill a role; and
nonparticipation, in which someone has little control over what is expected of their role.
Role conflict, role overload, underutilization of skills, and resource inadequacy were common for the experienced music teachers studied; role ambiguity and nonparticipation were not.
Role stress—and the underlying dynamics of organizations—might serve as a valuable framework for understanding and interpreting the stress music teachers experience. It provides structure beyond the subjective experience of stress while accounting for the nuance and complexity inherent in organizations. The framework has not been applied to its full potential in music education; there is need for an investigation of music teachers’ role stress that (a) is large in scale, (b) is conducted using psychometrically sound measures, and (c) includes potentially related psychological and occupational variables.
The purpose of this study was to fulfill that need by adapting and developing questionnaires and then collecting data to test a model to determine relationships among role stressors, role stress, and other variables of interest. This inquiry was guided by the following research questions: (1) Is secondary music teachers’ role stress related to negative occupational outcomes, including job satisfaction, organizational commitment, and intent to leave? (2) Does secondary music teachers’ role stress comprise role conflict, role ambiguity, role overload, underutilization of skills, resource inadequacy, and nonparticipation? and (3) Are years taught, amount of within-specialization teaching, and multiple jobholding related to role stress and its outcomes for secondary music teachers?
Other Variables of Interest
Negative Occupational Outcomes
In a pair of studies, Russell (2008, 2012) reported that about half of music teachers surveyed intended to leave their positions within 5 years. Annual rates of leaving (or intending to leave) positions—attrition and migration combined—have been reported between approximately a fifth and a quarter of music teachers (Hancock, 2009; Russell, 2008, 2012). Outside of music education, Conley and You (2009, 2014) found intent to leave to be an outcome of role stress, mediated by job satisfaction and organizational commitment. This progression of outcomes, culminating in intent to leave at the end of the school year, serves as the core of the hypothesized model in the present study.
Years Taught
The more years a teacher has taught, the more time they have had to develop stress management strategies. Hedden (2005) conducted a longitudinal study of 62 midwestern music educators to determine if stress changed over the course of a music teaching career. She measured these teachers’ stress in 1996 and then again 7 years later in 2003, finding a significant overall effect of time on stress. Specific decreases were related to time management, work-related stressors, professional distress, and discipline and motivation. Similarly, Bernhard (2016) found that more experienced teachers exhibited lower levels of burnout (a phenomenon related but not identical to stress) compared to beginning teachers.
Amount of Within-Specialization Teaching
Many music educators are tasked with teaching outside their primary areas of expertise (e.g., band, choir, orchestra, or general; Gillespie & Hamann, 1998; Groulx, 2016; National String Project Consortium, 2010), with many jobs involving teaching in more than one specialization area (Austin & Kellerman, 2005; Bowles, 2003; Give A Note Foundation, 2017). Nonspecialist teaching often presents difficulties because insufficient content knowledge limits one’s ability to teach in less-familiar areas (Allard, 1992; Grieser, 2014; Newbrey, 2008; Sckipp, 2010). Because preservice education prepares teachers to fill specific roles, it is unsurprising that there are difficulties when teachers must take on roles other than the ones for which they were most specifically prepared.
Multiple Jobholding
Music teachers’ stress may be affected by having additional jobs. In both 2006 and 2016, music education study participants indicated working an average of 5 hours per week in such jobs (Bernhard, 2016). One major question in the emerging organizational behavior literature on the topic of holding multiple jobs is “whether MJH [multiple jobholding] is personally depleting or enriching” (Campion et al., 2020, p. 175). This seems likely to depend heavily on the context; holding multiple jobs could generally increase music teacher stress (i.e., be depleting) because the additional responsibilities contribute to role overload, or it might decrease stress (i.e., be enriching) if underutilization of skills is alleviated when underutilized skills are used in a second job.
Method
Participants
Using the Research Survey Assistance program established by the National Association for Music Education (NAfME; 2019), I sent the study questionnaire in Spring 2020 to all 29,153 NAfME members who indicated teaching music at the secondary (middle and/or high school) level at the time. Data collection began on Monday, February 10, and lasted until Friday, February 28; potential participants received an identical invitation on Monday, February 24, as a reminder. All data collection took place prior to shutdowns related to the COVID-19 pandemic. The final usable sample size with no missing data was 1,576 participants, constituting 5.41% of the population.
The sample was diverse in terms of the collected demographic data. Participants represented all six NAfME regions within the United States (Southern: n = 443, 28.1%; North Central: n = 355, 22.5%; Eastern: n = 333, 21.1%; Southwestern: n = 181, 11.5%; Western: n = 144, 9.1%; Northwest: n = 113, 7.2%). Regarding levels, 578 (36.7%) taught only high school, 481 (30.5%) taught only middle school, and 517 (32.8%) taught both. All three traditional large ensemble areas were represented (band: n = 885, 56.2%; choir: n = 668, 43.7%; strings/orchestra: n = 344, 21.8%). Most respondents taught in at least one of these areas, and many had responsibilities in other areas such as secondary general (n = 748, 47.5%), guitar (n = 286, 18.1%), piano (n = 263, 16.7%), and/or mariachi (n = 14, 0.9%). Participants indicated specializations in band (n = 824, 52.3%), choir (n = 519, 32.9%), orchestra (n = 194, 12.3%), secondary general (n = 29, 1.8%), and elementary general (n = 10, 0.6%). Most had at least some of their workload within their specializations. Some participants indicated teaching elementary general music (n = 39, 2.5%) despite previously indicating that this was not a regular part of their responsibilities. I retained these participants and assumed that they only occasionally taught elementary general music (e.g., as part of recruitment efforts).
Measures
Music Teacher Role Stress Inventory
To develop an instrument to measure each of the six hypothesized role stressors, I began by considering those used in early studies of role stress (Beehr, 1974; Beehr et al., 1976; Quinn & Shepard, 1974; Rizzo et al., 1970) and their later revisions (Bacharach, Bamberger, & Conley, 1990; Conley & You, 2009). Extensive adaptations were necessary due to the age, noneducational focus, and—in some cases—poor reliability of the original scales (for a full discussion of the adaptation process, see Taft, 2020). I reversed several items to minimize response bias, adjusted dated and confusing language, shortened needlessly long items, and removed or replaced frequency and magnitude words from the item stems for clarity. Additional items were also necessary due to poor reliability and small numbers of extant items. I created items based on Scheib’s (2003) definitions of role stressors as related to music teaching. The initial item pool contained 48 items (eight per hypothesized role stressor), all configured to use a Likert-type agreement scale from 1 (strongly disagree) to 6 (strongly agree).
To gather preliminary information about scale reliability and the questionnaire experience, I conducted a pilot test with graduate students who had K–12 music teaching experience (n = 8). Additionally, to help establish validity, seven experts in social psychology and music education research categorized the items based on a provided definition of each role stressor. Of the 48 items, 43 (89.6%) were correctly classified by a majority of the experts. The most apparent pattern of incorrect responses involved classification of role overload items, especially those related to time, as resource inadequacy items. I removed items from the pool based on their detraction from pilot α reliability estimates and low correct expert classification rates until there were six items per stressor, except for resource inadequacy; here, I left seven items because reliability suffered upon removing more than one item.
Having established a degree of reliability and validity of the Music Teacher Role Stress Inventory (MTRSI) before distribution, I conducted an exploratory factor analysis (EFA) using data from the main study to ensure items adequately and distinctly represented the desired stressors. Because Likert-type scales are not always considered to be continuous, I used the asymptotically distribution-free weighted least squares (WLS) estimation method. An oblimin rotation allowed the factors to correlate in their representation of role stress as a superordinate construct. I used parallel analysis to determine the number of factors. After examining the initial table of factor loadings, I removed items that did not load well (<.400) onto the factors representing their intended constructs. I removed such items one at a time, selecting (with each new solution) whichever loaded most strongly on an unintended factor. Role conflict and role ambiguity items initially loaded onto the same factor, so I treated that factor as intended for both sets of items. Twelve poorly loading items were removed. The final solution included only five factors; role conflict and role ambiguity items did not separate. See Table 1 for factor loadings and Table 2 for factor correlations and characteristics.
Role Stress Exploratory Factor Analysis Factor Loadings.
Note. Loadings below .400 are suppressed. No items shown here cross-loaded above .300.
Role Stress Exploratory Factor Analysis Factor Correlations and Characteristics.
I then used confirmatory factor analysis (CFA) to test the factors derived with the EFA and determine whether they contributed to a second-order role stress factor. Use of diagonally weighted least squares (DWLS) estimation accounted for ordinality of item response options. The model was identified using factor variance scaling. Results of the CFA indicated good fit for the five-factor model, χ2(270) = 772.20, p < .001, comparative fit index (CFI) = .989, root mean square error of approximation (RMSEA) = .034.
Finally, I conducted invariance testing with reference to school level (middle, high, or both), whether the job involved teaching in multiple areas (e.g., choir only [no] or both choir and orchestra [yes]), and teaching experience (Mdn = 15 years, split into groups of 1–14 and 15+). To test invariance, I examined the CFI of increasingly stringent invariance models: configural, metric, scalar, and strict. The null hypothesis of invariance is to be rejected if ΔCFI—the CFI of the more constrained model minus that of a less constrained model—is smaller than −.01 (i.e., if the CFI drops by more than .01; Cheung & Rensvold, 2002). This criterion was met for all three variables tested. The final MTRSI included 25 items pertaining to five subscales with three to seven items per subscale. Table S1 in the supplemental material includes—for the MTRSI and the two subsequently described measures—the items tested, their codes, and whether and when they were removed.
Revised Job Satisfaction Scale
Four job satisfaction items were adapted by Conley et al. (1989) from Bacharach and Mitchell (1982) and used reliably with teachers by Conley and You (2009, 2014). I adapted these and developed four original items (e.g., “My work provides me with a high degree of satisfaction” [SATIS4]). I used pilot results to reduce the number of items from eight to six, removing the items detracting most from reliability.
I also tested the Revised Job Satisfaction Scale (RJSS) for invariance for the same three variables as the MTRSI. Each ΔCFI was above −.01. The final RJSS included six items, each using a 6-point Likert-type agreement scale.
Organizational Commitment Questionnaire for Teachers
To measure organizational commitment, I began with two items adapted from those used by Conley and You (2009; e.g., “I am proud to tell others I am part of this school” [COMM1]), which they adapted from Mowday et al. (1982). I then selected four more of the original items based on their logical applicability to the work of secondary music teachers, simplified their language where needed, and recontextualized them from the generic “organization” to the context-specific “school” (e.g., “I am willing to go above and beyond to help this school be successful” [COMM2]). Pilot reliability estimates provoked the removal of one item.
As with the MTRSI and RJSS, I tested the Organizational Commitment Questionnaire for Teachers (OCQ-T) for invariance across levels of three variables. For teaching experience, ΔCFI was −.016 from the configural step (CFI = .994) to the metric step (CFI = .978). At the configural step, one item (“I would be just as happy doing the same job in a different school” [COMM5]) loaded at .486 for those with 15 or more years of experience but at .299 for those with 14 or fewer. I removed this item, after which each ΔCFI was acceptable. The final OCQ-T included four items, each using a 6-point Likert-type agreement scale.
Other model variables
Intent to leave, years taught, within-specialization teaching percentage, and multiple jobholding were each reported using a single item. Participants indicated their intent to leave (“What are your intentions regarding staying in or leaving your current position after this school year?”) on a scale from 1 (I will definitely stay) to 6 (I will definitely leave). Years taught (“Including this year, how many years have you been teaching music?”) were entered manually by participants. Within-specialization teaching percentage (“What percentage of your teaching responsibilities [curricular and cocurricular] falls within the specialty area you identified above?”) was reported using a slider immediately after participants designated their single area of specialization (“Regardless of what you currently teach, which one of these areas do you consider to be your main specialty area as a music teacher?”; options included choir, orchestra, band, elementary general, and secondary general). The multiple jobholding item was dichotomous (“Do you have an additional job [or jobs] beyond your secondary music teaching job?”). To contextualize results, participants selected reasons for taking their primary and secondary (if applicable) jobs from a list adapted from Campion et al. (2020). Those with multiple jobs answered two additional questions about the impact of having multiple jobs on their stress (also using a 6-point Likert-type agreement scale).
Screening items and demographics
Participants responded to three screening items to help ensure they were full-time music teachers at the secondary level whose jobs did not involve any regular teaching of general music at the elementary level. They also provided their locations (state in which they taught), teaching levels, and teaching areas. Teaching level was limited to middle and high school, so the only response options were middle school, high school, or both. Options for teaching areas were based on the categories set by NAfME for members to report when joining the organization.
Analysis
To establish fit of the full measurement model, I conducted a CFA including the items that were retained after previous analyses of the MTRSI (five factors contributing to a second-order factor), RJSS, and OCQ-T. I then examined the full structural equation model (SEM) for fit and for significant estimates of measurement and structural parameters. Structural paths estimated were from role stress to job satisfaction, job satisfaction to organizational commitment and intent to leave, and organizational commitment to intent to leave. I also included years taught, percentage of within-specialization teaching, and multiple jobholding as covariates. As with the CFAs, model parameters were estimated using the DWLS method. Errors for the final CFA and full SEM were calculated using 1,000 bootstrap samples. Factors were scaled using factor loadings, with the (nonstandardized) parameter estimate of the first item in each scale fixed to 1.000. I assessed model fit using the CFI and RMSEA with the conventional cutoff values suggested by Hu and Bentler (1999): CFI > .95 and RMSEA < .06. Standardized parameter estimates were examined for interpretation. I conducted all analyses using JASP (JASP Team, 2020), which includes modeling functionality based on the R package lavaan (Rosseel, 2012).
Results
Participants’ teaching experience ranged from 1 to 50 years, with an average of 15.87 years (SD = 10.28). The average within-specialization teaching load was 77.68% (SD = 24.39%). Almost one third of participants (n = 497) taught entirely within their specializations, and very few taught exclusively outside their primary areas (see Table S2 in the supplemental material for detailed counts of teaching areas by specialization). Over half of respondents had at least one other job (n = 854, 54.2%), which they generally agreed added to their stress (M = 4.53, SD = 1.30). They responded more moderately regarding whether having multiple jobs made it difficult to do them well (M = 3.49, SD = 1.52). The top reasons for taking music teaching jobs included enjoying the work (65.9%), meeting regular expenses (65.5%), and job security (62.3%); for additional jobs, top reasons were having an opportunity for extra paid work (41.8%), enjoying the work (33.2%), and paying off debts (27.2%).
Means for the five role stressors ranged from 2.45 to 4.18; the mean for role overload was the only one to fall above the scale midpoint of 3.50. On average, participants were somewhat satisfied with their jobs. They generally exhibited a high degree of organizational commitment, with 293 (18.6%) reporting the maximum possible score of 6 on the organizational commitment scale. Most participants (n = 810, 51.4%) definitely intended to stay in their current jobs, with only 287 (18.2%) responses in the upper half of the intent to leave scale. Descriptive statistics for main study variables are shown in Table 3.
Descriptive Statistics for Main Variables.
Note. Scale statistics presented here include retained items only and were calculated as means of scale items, not factor scores. Role stress scores were calculated as means of the five subscale scores. Items related to all variables listed—except years taught and within-specialization teaching percentage—were measured on a scale from 1 to 6.
McDonald’s ω, a measure of reliability used to avoid various known issues with the more common Cronbach’s α (Hayes & Coutts, 2020; McNeish, 2018).
Measurement Model
The measurement model for role stress and its factors, job satisfaction, and organizational commitment exhibited good fit, χ2(552) = 2,379.87, p < .001, CFI = .979, RMSEA = .046. All factor loadings were significant, and standardized estimates were all above .400. Table S3 in the supplemental material shows parameter estimates for the CFA.
Structural Model
The full SEM fit well, χ2(682) = 2,661.36, p < .001, CFI = .978, RMSEA = .043. Figure 1 shows the model graphically. Structural paths were all significant. Relationships between each of the three covariates and role stress were statistically significant but small in magnitude (multiple jobs: ϕ = .074, p = .008; within-specialization teaching: ϕ = −.160, p < .001; years taught: ϕ = −.165, p < .001). Multiple jobs also significantly predicted greater organizational commitment, again with small magnitude (β = 0.056, p = .010).
Structural equation model of role stress and occupational outcomes.
Job satisfaction had a direct effect of β = −0.193 (p < .001) on intent to leave. There was also an indirect effect of β = −0.284 (p < .001) through organizational commitment, for a total effect of β = −0.477 (p < .001). Role stress had an indirect effect of β = 0.374 (p < .001) on intent to leave. The direct effect of role stress on intent to leave was not part of the model; however, when added, the path was not significant (β = 0.012, p = .809). Table S4 in the supplemental material shows parameter estimates for the full SEM.
Discussion
Results of the present study suggest a connection between role stress and negative occupational outcomes including job satisfaction, organizational commitment, and intent to leave. As has been consistently determined for workers in other fields (e.g., Acker, 2004; Conley et al., 1989; Conley & Woosley, 2000; Conley & You, 2009, 2014; Reyes & Imber, 1992; Richards et al., 2016; Schwab & Iwanicki, 1982), the role stress of secondary music teachers in this sample was a meaningful predictor of negative occupational outcomes. These outcomes terminated with intent to leave the present position, the prevalence of which was consistent with previous reports at about one fifth of respondents (Hancock, 2009; Russell, 2008, 2012).
For those wishing to decrease role stress in hopes of improving occupational outcomes, the precise definition of “role stress” is critical. In the model, role stress comprised five factors examined by Scheib (2003) and studied extensively in various contexts: Role Conflict/Role Ambiguity, Role Overload, Underutilization of Skills, Resource Inadequacy, and Nonparticipation. Although Role Overload (e.g., Sandene, 1995) and Resource Inadequacy (e.g., Gordon, 2000) are often cited as major stressors, they did not contribute to overall role stress more than the other three factors; in fact, Role Overload contributed the least to overall role stress. All five factors ought to be considered when contemplating possibilities for role stress reduction.
The five factors of role stress were further defined by the items that were retained and omitted. Role conflict and role ambiguity—although they have been determined to be separate in other contexts (Rizzo et al., 1970)—functioned as one factor related mostly to administrative guidance (including rules and policies), with uncertainty about core responsibilities absent entirely. These stressors warrant further research with focus on the part that administrators play in inducing or alleviating them. Role Overload was characterized by having too much to do, including insufficient time and the possibility of limited quality of work given the quantity. Underutilization of Skills included items related to concrete “knowledge” and “skills” but not the less-specific “challenge” and “expertise.” Resource Inadequacy items considering other people (including administrators) as resources were removed, leaving only budget- and equipment-specific items and general items referencing “resources” broadly. Finally, remaining Nonparticipation items were related to participants’ own agency rather than the ways in which other people might limit that agency.
An important quality of role stress is that it is related to concrete job characteristics. This means that while it may not be easy, minimization of role stress is not an abstract endeavor. Music teachers can examine the list of items included in the final model and consider specific ways in which their jobs may be causing them stress. Importantly, many, if not most, of the job characteristics described by the items are—at least theoretically—directly changeable. Music teachers likely hold some of the power to change their occupational situations, such as by keeping the number of extracurricular activities at a reasonable level. In many cases, however, school and district administrators hold the majority of this power in the form of policies, expectations, course loads, and resource allocation. In some cases, music teacher may simply need to bring a problem (e.g., “incompatible policies and procedures” [RC1]) respectfully to the attention of the right administrator to provoke change. Solutions that are difficult or expensive (e.g., hiring additional personnel) are less likely to be implemented without greater efforts.
When a given stressor cannot be meaningfully reduced, it may be possible to capitalize on the interactions among stressors to reduce total role stress. Administrators might insist that ensembles regularly perform at specific functions, but they might in exchange be willing to locate additional funding for needs related to those performances. Role overload would not be reduced, but diminished resource inadequacy could result in more manageable role stress overall. However, the interactions among role stressors may not always be constructive; reducing one role stressor may escalate another. Increased fundraising efforts, for example, may decrease resource inadequacy but increase role overload. Depending on the context, this might be a worthwhile balancing strategy or a debilitating side effect.
Even when no role stressors can be meaningfully reduced, attending to job satisfaction and organizational commitment might help keep secondary music teachers in their positions. For the current sample, the effect of role stress on intent to leave was entirely indirect. Rather than causing teachers to leave their jobs, role stress may have degraded the satisfaction and commitment that helped keep them there. Administrators interested in high retention rates might consider context-authentic ways to improve job satisfaction and organizational commitment when they are not empowered to alleviate role stressors. Organizational commitment, which appears to have acted as a buffer between dissatisfaction and intent to leave, may be especially important and is worthy of further study. Most participants were quite committed to their schools, and while much commitment was related to their job satisfaction, some was derived from unknown factors outside the model and may have helped hold them in their positions even when they were dissatisfied.
Music teacher educators can help prepare preservice teachers to contend with role stressors, which could help prolong the careers of their graduates. Methods for doing so might include teaching about role stress, providing opportunities for students to practice advocating for themselves, modeling healthy work habits, and emphasizing the importance of psychological well-being for a long and productive career. It may also be useful to introduce preservice teachers to NAfME’s (2020) National Opportunity to Learn Standards, which include guidelines for acceptable workloads and necessary resources. Preparing in these ways might help early career music teachers to see and deal with role stressors and perhaps even avoid taking positions that seem likely to be unreasonably stressful.
Limitations
This study was delimited to full-time secondary music teachers who were members of NAfME. I omitted elementary music teachers because of likely differences in the mechanics of role stress for secondary and elementary music teachers. Bacharach, Bamberger, and Mitchell (1990) found that determinants of role stress differed between elementary and secondary teachers; these differences seem likely to hold or even be exacerbated in music based on the stark differences in responsibilities between elementary and secondary music teaching positions. Both sets of responsibilities have the potential to be stressful at times, but the ways in which that potential manifests seem likely to be distinct.
Even for full-time secondary music teachers who are members of NAfME, interpretation of these results is limited by several elements of the study. First, although the sample size was substantial and more than adequate for the statistical analyses conducted, over 90% of potential participants did not respond. These teachers (and those whose responses were removed) may have answered questions differently than those who responded fully. The results and related discussion are most applicable to those whose responses were analyzed and those whose contexts closely resemble theirs. Furthermore, the measures used in this study were novel. This is both a contribution and a limitation: Much of the preliminary work of crafting and documenting reliable and valid measures is done, but researchers using these measures should continue to refine them.
A word of caution related to causation is warranted. SEM results are no more causally certain than other correlation-based results. While one might reasonably suspect that role stress causes the occupational outcomes observed in this study, all that can be said from the data alone is that role stress is related to those outcomes. The relationships in the model may be directly causal, or there may be common causes or other unexplored causal elements.
Finally, it should be noted that the final model included role stress specifically, not other kinds of stress or the subjective experience of any kind of stress. Stress not defined by concrete job characteristics may interact with model variables differently. The model, therefore, has no implications related to non-role stress.
Opportunities for Future Research
To determine the structure and relationships of role stress for music teachers not represented in this sample, the final model could be tested with data from elementary music teachers and other omitted populations such as part-time music teachers and those who are not members of NAfME. The model resulting from the present study is not necessarily applicable for these teachers, but it can serve as a starting point for modification when considering how role stress may function for members of these groups. Researchers wishing to undertake this task and other extensions of this work may wish to initially include both role conflict and role ambiguity, perhaps deemphasizing the administrative elements included in the items used in the present study.
Fewer years taught, less within-specialization teaching, and having multiple jobs were all weakly related to more role stress for participants in the present study. Because nonspecialist teaching (e.g., Allard, 1992; Grieser, 2014; Newbrey, 2008; Sckipp, 2010) and inexperience (e.g., Hedden, 2005) have been found to cause difficulties, it may be worth exploring relationships between these variables and other outcomes such as student achievement or non-role stress. Multiple jobholding has not been studied substantially in music education, but the prevalence of multiple jobs among secondary music teachers—found in the present study and in previous research (Bernhard, 2016)—suggests the need for further study of the enriching and depleting qualities of those jobs and the main reasons music teachers take them. The nature of the additional jobs may also be of interest. Some could have been related to the secondary music teaching jobs, such as playing with a local orchestra or tutoring math, and others entirely unrelated, such as evening and weekend retail shifts. It seems likely that these different kinds of additional jobs may have importantly different impacts on the lives of music teachers.
Practical study of approaches for systematically alleviating role stressors could also be extraordinarily valuable. In conducting this research, scholars should not neglect the psychologically healthiest individuals; case studies of those best able to minimize role stress may be particularly helpful for those with difficulty doing so. Furthermore, role stressors beyond those studied thus far may exist; future qualitative study grounded in the original framework of role stress (Kahn et al., 1964; for a partial summary, see Taft, 2020) might reveal additional categories of job characteristics with a tendency to cause role stress.
Conclusion
Role stress is a useful framework in part because it is objectively rooted in changeable organizational dynamics. Much scholarly and popular discourse about stress reduction has emphasized internal responses to the subjective experience of stress. Certainly, mindfulness and other techniques are important to help endure difficulties and dismantle internal structures that induce stress (e.g., fear of being seen as inferior for having a less-active program). However, such approaches can leave external stressors—including role stressors—intact. For long-term relief, stressful situations may require action to address the external structures that induce stress. Such action may be necessary for the health and longevity of the music teaching workforce.
Supplemental Material
sj-docx-1-jrm-10.1177_00224294221134538 – Supplemental material for Music Teacher Role Stress: A Structural Equation Model
Supplemental material, sj-docx-1-jrm-10.1177_00224294221134538 for Music Teacher Role Stress: A Structural Equation Model by Seth A. Taft in Journal of Research in Music Education
Footnotes
Declaration of Conflicting Interests
The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author received no financial support for the research, authorship, and/or publication of this article.
Author Biography
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
