Thesis – Early-Career Teacher Retention and the Role of Induction

Abstract

Attrition among Australian teachers concentrates in the first five years of practice, yet the specific working conditions that convert early enthusiasm into departure remain contested. This study examined how the quality of formal induction shapes the retention intentions of early-career teachers, and whether teacher self-efficacy operates as the mechanism linking induction to the decision to stay. A convergent mixed methods design combined a national survey of 400 teachers within their first five years with 15 semi-structured interviews. Induction quality, mentoring quality, self-efficacy, workload manageability and school leadership support were measured alongside retention intention. Almost one in three respondents reported weak intention to remain in teaching. Hierarchical regression showed that self-efficacy, workload manageability and leadership support were the strongest predictors of retention intention, with induction and mentoring quality contributing additional variance. Interviews distinguished induction experienced as compliance from induction experienced as classroom practice. Implications for national and state induction frameworks are discussed.

Introduction

The stability of the Australian teaching workforce depends heavily on what happens in the years immediately after graduation. National workforce data indicate that the profession loses a disproportionate share of its members early, before the return on their initial teacher education has been realised (Australian Institute for Teaching and School Leadership [AITSL], 2023). Estimating the exact rate is difficult, because Australian datasets define and measure attrition inconsistently, but the balance of evidence places early-career losses well above those of the mid-career workforce (Weldon, 2018). Against a backdrop of sustained workforce pressure documented by the Australian Bureau of Statistics ([ABS], 2023), each avoidable departure represents lost public investment and disrupted learning for students.

Induction has become the principal policy response to this problem. The national guidelines set out in Graduate to Proficient position structured induction as the bridge between initial teacher education and proficient practice (AITSL, 2016), and the Quality Initial Teacher Education Review reaffirmed high-quality induction and mentoring as levers for retention (Department of Education, 2022). State induction frameworks, such as the New South Wales Strong Start, Great Teachers program, translate these expectations into school-level provision. Yet provision varies widely in intensity and quality, and it is not well understood how, and through what psychological mechanism, induction influences whether early-career teachers intend to remain.

This thesis addresses that gap through a national mixed methods study. It positions teacher self-efficacy as the candidate mechanism connecting induction to retention. Three research questions guided the work:

1. What is the strength of retention intention among early-career teachers, and how does it relate to the quality of induction, mentoring, workload and leadership support?
2. Does teacher self-efficacy predict retention intention once these conditions are accounted for?
3. How do early-career teachers themselves describe the features of induction that sustain or erode their intention to stay?

Literature Review

Early-career attrition in Australia

Australian research frames early-career attrition as a systemic rather than an individual phenomenon. Mason and Poyatos Matas (2015) argued that retention is best understood through the interaction of personal, school and system factors, rather than as a deficit in the resilience of individual graduates. Internationally, Kelchtermans (2017) similarly reframed the question from “why do teachers leave” to “under what conditions do they stay”, locating the answer in the everyday working environment. The most comprehensive quantitative synthesis to date found that induction, mentoring, administrative support and manageable workload are consistently among the strongest correlates of turnover (Nguyen et al., 2020). These conclusions justify examining conditions rather than dispositions.

Induction and mentoring

Induction and mentoring are frequently treated as a single intervention, but the evidence distinguishes their quality from their mere presence. Ingersoll et al. (2018) demonstrated that the number of induction components a beginning teacher receives, and their intensity, predicts commitment more reliably than nominal participation. In the Australian setting, the national guidelines specify that effective induction is sustained across the first two years, is classroom-focused, and pairs beginning teachers with a trained mentor (AITSL, 2016). Evidence collated by the Australian Education Research Organisation ([AERO], 2023) indicates, however, that many early-career teachers experience induction as an orientation event rather than an ongoing practice, and that access to protected mentoring time is uneven. Quality, not availability, appears to be the operative variable.

Self-efficacy as a mechanism

Teacher self-efficacy, a teacher’s belief in their capacity to bring about student learning, offers a plausible mechanism through which supportive induction translates into retention. A synthesis of four decades of research links higher self-efficacy to better classroom processes, greater wellbeing and lower burnout (Zee & Koomen, 2016). Because efficacy beliefs are formed substantially through mastery experiences and credible support during the early years, high-quality induction and mentoring are theoretically well placed to build them. The present study therefore treats self-efficacy not as a control variable but as the hypothesised bridge between induction quality and the intention to remain.

Methodology

Design and conceptual framework

A convergent mixed methods design was adopted (Creswell & Plano Clark, 2018), in which quantitative and qualitative strands were collected in parallel and integrated at interpretation. The conceptual framework guiding the study is shown in Figure 1. It positions self-efficacy as the mechanism through which induction quality, mentoring quality and leadership support influence retention intention, while workload demands exert a direct countervailing pressure.

Participants

The survey sample comprised 400 teachers in their first five years of practice, recruited nationally through professional associations and school networks. Mean time since graduation was 2.7 years (SD = 1.3). Reflecting the composition of the workforce, 296 respondents (74%) were women. Employment spanned the government (268, 67%), Catholic (74, 18.5%) and independent (58, 14.5%) sectors, and primary (214, 53.5%) and secondary (186, 46.5%) settings. Using the ABS remoteness structure (ABS, 2023), 300 taught in major cities (75%), 88 in inner and outer regional areas (22%) and 12 in remote or very remote locations (3%). The 15 interview participants were purposively sampled for maximum variation across sector, school level and remoteness.

Measures and analysis

Induction quality, mentoring quality, workload manageability and leadership support were measured with multi-item scales scored from 1 to 5, with higher scores indicating more favourable conditions. Self-efficacy was assessed with a 12-item scale adapted from established teachers’ sense of efficacy instruments (see Zee & Koomen, 2016). Retention intention was measured on a 1 to 7 scale, with higher scores denoting stronger intention to remain in teaching. Internal consistency was acceptable to high for all scales (Cronbach’s alpha .82 to .91). Quantitative analysis comprised descriptive statistics, Pearson correlations and hierarchical multiple regression. Interviews of 45 to 60 minutes were analysed using reflexive thematic analysis (Braun & Clarke, 2022).

Ethical considerations

The study was approved by the administering university’s Human Research Ethics Committee and conducted in accordance with the National Statement on Ethical Conduct in Human Research (National Health and Medical Research Council [NHMRC], 2023). Participation was voluntary and confidential, informed consent was obtained, and interviewees are identified only by code and non-identifying descriptors.

Results

Descriptive and correlational findings

Retention intention was moderate overall (M = 4.57, SD = 1.48 on the 1 to 7 scale), but 126 respondents (31.5%) scored at or below the scale midpoint, indicating that almost one in three early-career teachers held weak intention to remain. Workload manageability recorded the least favourable mean of the study variables (M = 2.68), consistent with interview accounts of intensified early-career loads. As shown in Table 1, retention intention correlated positively with all five conditions, most strongly with leadership support (r = .49), self-efficacy (r = .46) and workload manageability (r = .44).

Table 1: Means, standard deviations and correlations among study variables (N = 400)

Note. All correlations are significant at p < .001. Retention intention is scored 1 to 7; all other variables are scored 1 to 5. Cronbach’s alpha values are induction quality .89, mentoring quality .91, self-efficacy .86, workload manageability .82, leadership support .90 and retention intention .88.

Predicting retention intention

A hierarchical regression predicted retention intention. Background controls entered at Step 1 (years since graduation, sector, school level and ABS remoteness) explained little variance (R² = .06). Adding the five conditions at Step 2 raised the model substantially (ΔR² = .36, p < .001), giving a final model that accounted for 42% of the variance in retention intention, F(11, 388) = 25.6, p < .001. As Table 2 reports, leadership support (β = .24), workload manageability (β = .23) and self-efficacy (β = .21) were the strongest unique predictors, while induction quality (β = .19) and mentoring quality (β = .11) contributed further variance. The attenuation of the induction and mentoring coefficients relative to their bivariate correlations reflects their shared variance with leadership support and with self-efficacy, consistent with the framework in Figure 1.

Table 2: Hierarchical regression predicting retention intention (final model, N = 400)

Note. Outcome is retention intention (1 to 7). Step 1 controls explained R² = .06; focal predictors were added at Step 2 (ΔR² = .36). Final model R² = .42, adjusted R² = .40. Control coefficients are omitted for concision; none reached significance at Step 2.

Interview themes

Reflexive thematic analysis produced four themes, summarised in Table 3. The most pervasive distinction was between induction experienced as compliance and induction experienced as practice. As one participant put it, “the induction was a folder and a half-day; the real learning happened when a colleague let me watch her teach” (P4, secondary, government, major city). Mentoring was valued only where it was relational and timetabled: “my mentor was assigned but we were never released together, so it never happened; in my second school we had release every fortnight and it changed everything” (P9, primary, Catholic, inner regional). Workload intensification was near-universal, with several early-career teachers describing full loads and out-of-field allocations, and leadership emerged as the condition that set the tone for the rest.

Table 3: Themes from reflexive thematic analysis of interviews (n = 15)

Discussion

The findings answer the research questions in a mutually reinforcing way. Quantitatively, retention intention rose with every supportive condition, and self-efficacy, workload manageability and leadership support carried the largest unique weights. Qualitatively, early-career teachers described the same conditions, and their accounts explain why the quality of induction, rather than its presence, is decisive. Induction reduced to a folder and an orientation day did little; induction embedded in observed classroom practice and sustained mentoring did a great deal. This distinction aligns with evidence that intensity and classroom focus, not nominal participation, drive commitment (Ingersoll et al., 2018), and with the sustained, practice-based model set out in the national guidelines (AITSL, 2016).

The prominence of self-efficacy supports the study’s central proposition that induction influences retention partly by building teachers’ belief in their own effectiveness, a belief repeatedly linked to wellbeing and persistence (Zee & Koomen, 2016). Mentoring mattered, but only when it was relational and protected by timetabled release; a nominal pairing without shared time produced little benefit, echoing the uneven access documented nationally (AERO, 2023). Workload manageability was both statistically influential and, in the interviews, the most common grievance, consistent with meta-analytic evidence on turnover correlates (Nguyen et al., 2020) and with the reframing of retention as a function of conditions rather than individual resilience (Kelchtermans, 2017; Mason & Poyatos Matas, 2015). Leadership support functioned as the enabling condition that determined whether induction and mentoring were resourced at all.

Several limitations qualify these conclusions. The design is cross-sectional, so the mediational ordering implied by Figure 1 cannot be confirmed causally. All measures are self-reported, and retention intention is an imperfect proxy for actual attrition, which remains difficult to quantify in Australia (Weldon, 2018). The remote subsample was small, limiting inference for the highest-need settings.

Conclusion

Early-career teacher retention in Australia is shaped less by whether induction occurs than by how well it is delivered. When induction is classroom-focused and sustained, mentoring is relational and protected, workload is graduated, and leadership actively resources early-career development, teachers build the self-efficacy that underpins their intention to stay. The policy implication is that national and state induction frameworks should be specified and monitored for quality and intensity, not merely for coverage, in line with the direction set by the Quality Initial Teacher Education Review (Department of Education, 2022) and the national workforce evidence base (AITSL, 2023). Investing in the conditions that make the first five years survivable is likely to be a more efficient retention strategy than recruiting to replace those who leave.

References

Australian Bureau of Statistics. (2023). Schools, Australia. ABS.

Australian Education Research Organisation. (2023). Beginning teachers’ experiences of induction and early-career support. AERO.

Australian Institute for Teaching and School Leadership. (2016). Graduate to proficient: Australian guidelines for teacher induction into the profession. AITSL.

Australian Institute for Teaching and School Leadership. (2023). Australian teacher workforce data: National teacher workforce characteristics report. AITSL.

Braun, V., & Clarke, V. (2022). Thematic analysis: A practical guide. Sage.

Creswell, J. W., & Plano Clark, V. L. (2018). Designing and conducting mixed methods research (3rd ed.). Sage.

Department of Education. (2022). Next steps: Report of the Quality Initial Teacher Education Review. Australian Government.

Ingersoll, R. M., Merrill, E., Stuckey, D., & Collins, G. (2018). Seven trends: The transformation of the teaching force (Updated ed.). Consortium for Policy Research in Education.

Kelchtermans, G. (2017). “Should I stay or should I go?”: Unpacking teacher attrition/retention as an educational issue. Teachers and Teaching, 23(8), 961-977.

Mason, S., & Poyatos Matas, C. (2015). Teacher attrition and retention research in Australia: Towards a new theoretical framework. Australian Journal of Teacher Education, 40(11), 45-66.

National Health and Medical Research Council. (2023). National statement on ethical conduct in human research. NHMRC.

Nguyen, T. D., Pham, L. D., Crouch, M., & Springer, M. G. (2020). The correlates of teacher turnover: An updated and expanded meta-analysis of the literature. Educational Research Review, 31, 100355.

Weldon, P. R. (2018). Early career teacher attrition in Australia: Evidence, definition, classification and measurement. Australian Journal of Education, 62(1), 61-78.

Zee, M., & Koomen, H. M. Y. (2016). Teacher self-efficacy and its effects on classroom processes, student academic adjustment, and teacher well-being: A synthesis of 40 years of research. Review of Educational Research, 86(4), 981-1015.