Abstract
This study aims to develop and validate the Future-Ready School Leadership Instrument (FRSLI), a psychometrically robust instrument that measures the multifaceted dimensions of effective school leadership in an era of rapid change and uncertainty. Three independent studies were conducted to evaluate the FRSLI's psychometric properties: Study 1 used exploratory factor analysis (EFA) to identify the underlying factor structure; Study 2 employed confirmatory factor analysis (CFA) and structural equation modeling (SEM) to assess construct, convergent, and discriminant validity, measurement invariance, and nomological validity; and Study 3 tested criterion-related and incremental validity by examining the FRSLI's predictive power for organizational learning culture both independently and after controlling for transformational leadership. The FRSLI demonstrated sound psychometric properties, identifying four key components of future-ready school leadership: strategic foresight, addressing difference and diversity, technology leadership, and systems thinking. The final 15-item instrument demonstrated high internal consistency and explained a substantial proportion of the variance in the construct. CFA confirmed the four-factor structure, and SEM endorsed its nomological network in relation to transformational leadership. Notably, after statistically controlling for the transformational leadership construct, the FRSLI uniquely predicted teachers’ perceptions of organizational learning culture, underscoring its value as a salient measurement construct.
Keywords
Introduction
While the concept of “future” carries numerous positive connotations for educational leaders, it also evokes a range of ambiguous meanings stemming from educational policies and multiple, ever-shifting contexts laden with both uncertainty and opportunity (Drysdale and Gurr, 2017; Fink and Brayman, 2006). Drawing from L.P. Hartley's opening passage in The Go-Between (1953), Hargreaves and Shirley (2012) metaphorically describe “the past as a foreign country,” from which school leaders extract stories and ideas. However, as they look and work “ahead” proactively, their schools and broader contexts are increasingly dynamic, undergoing “change in ways that cannot always be predicted” (Day and Leithwood, 2007: 184).
Such forward-looking leadership has become essential as today's school leaders confront unprecedented, often unknowable challenges while aiming to prepare future-ready students. These challenges include globally inspired policy surges and initiatives aimed at enhancing domestic economic resilience and international competitiveness (Gopinath et al., 2025), addressing widening social inequity (Gümüş et al., 2026b; Khalifa, 2018), demographic and ethnic shifts within student populations (Arar and Örücü, 2024; Dimmock and Walker, 2005), and the integration of technological advancements in educational settings (Fullan et al., 2024; Harris and Jones, 2020).
These intersecting demands create a dense web of internal and external complexities, increasingly placing school leaders at the center of competing expectations (Richard and Ahlström, 2025). The simultaneity and unpredictability of these demands amplify uncertainty in school leadership, rendering reactive or compliance-driven approaches insufficient (Sum, 2022). Consequently, school leaders are compelled to act proactively by exercising strategic influence over school change and inspiring shared future-oriented goals (Dimmock and Walker, 2000; Tang et al., 2014), navigating trade-offs across multiple system levels (Qian et al., 2025), and mobilizing organizational learning (OL) cultures that enable schools to adapt continuously rather than respond episodically to external pressures (Senge et al., 2012).
In response to these escalating complexities and drawing on the proactive behavior model (Grant and Ashford, 2008; Parker et al., 2010), we conceptualized Future-Ready School Leadership (FRSL) for school leaders operating in complex, uncertain, and accountability-driven environments. Leader proactivity could be defined as self-initiated, context-sensitive, and future-oriented action through which school leaders anticipate, strategically plan for, and actively shape their organizational environments amid increasing ambiguity, external accountability pressures, and constrained professional autonomy. From this complexity-driven perspective, leaders’ work motivation extends beyond adaptive responsiveness to encompass intentional sense-making, opportunity recognition, and the strategic mobilization of organizational capacities ahead of emerging demands (Parker et al., 2010).
Future-ready school leaders with a proactive mindset should be able to develop and leverage their capacities for innovation, inclusivity, adaptability, and systems thinking (Fullan, 2023; Ng and Wong, 2020; Senge et al., 2012). They aspire to build future-ready schools that prepare future-ready students to succeed in multiple, unknown futures. Intentionally embracing a strategic mindset is crucial for achieving this, as it not only addresses the complexity of leading schools in today's turbulent environment, but also fosters a culture of innovation that empowers stakeholders to explore “new ways of doing things and for pedagogical innovations, as they seek to position their school so that they can take advantage of them as they unfold” (Quong and Walker, 2010: 28). By effectively providing resources, including new technology, recognition, delegation, and consultation, as well as genuinely considering the diverse backgrounds and needs, these leaders foster the desired cultural and structural changes that enhance the school's overall capacity for future-readiness (Gil et al., 2018).
Overall, future-ready school leaders intentionally develop their mindsets and capabilities to anticipate, adapt to, and innovate amid continual, often unsettling changes. As such, they focus on shaping and managing the present while simultaneously and strategically preparing for uncertainty ahead (OECD, 2021). To that end, this study aims to develop and validate a new measurement instrument, the Future-Ready School Leadership Instrument (FRSLI), grounded in a rigorous conceptual framework and the methodological guidelines proposed by key sources. To achieve this objective, the study poses two research questions (RQs):
What core components define future-ready school leadership?
How reliable and valid is the developed FRSLI in assessing these components?
Conceptual framework
The proposed FRSL framework is grounded in four key components: strategic foresight, addressing difference and diversity, technology leadership, and systems thinking. The framework offers a comprehensive approach for school leaders to develop the capabilities needed to meet the demands of today's educational landscape. Adopting this framework is not merely a reaction to current challenges; it is a proactive strategy to cultivate an inclusive, innovative, and resilient educational environment that benefits all stakeholders.
Strategic foresight
Innovative practices and policies serve as a touchstone for distinguishing between administration/management and leadership, as noted by Davies (2011), “how we distinguish between administration/management and leadership is how individuals both operate in the ‘now’ of the current school year, while challenging and creating new ways of doing things for the future” (18–19). Strategic foresight, therefore, is not concerned with prediction per se, but with the systematic exploration of plausible futures and the enhancement of school leadership capacity to anticipate, imagine, and respond to complex futures by recognizing trends and uncertainties (Kaytu, 2025). Accordingly, the strategic foresight component underlines the need for school leaders to cultivate their capacity for innovativeness by integrating both hindsight and foresight perspectives (Balser and Tafuro, 2025).
Beyond its future orientation in temporal planning, strategic foresight requires school leaders to adopt an outward-looking proactive stance that transcends the school's physical boundaries (Hughes and Davis, 2025; Quong and Walker, 2010). Leaders with well-developed strategic foresight are expected to engage in “strategic scanning,” extending their attention beyond the immediate organizational context to proactively assess the alignment between existing organizational capacities and emerging demands within the broader environment (Bindl and Parker, 2011). Evidence from Parker and Collins's (2010) study indicates that opportunity seeking and innovation are closely intertwined, insofar as future- and change-oriented behaviors function as critical enabling conditions for the enactment of innovation
In recognition of this forward-looking perspective, future-ready school leaders may be characterized by a proactive mindset that fosters innovation and the creation of professional knowledge throughout the school (Godfrey, 2016). This approach also enables school stakeholders to identify potential opportunities and threats from both inside and outside perspectives, positioning their schools for long-term success (Gurr and Drysdale, 2020). By nurturing a culture of innovation and forward-thinking, school leaders can effectively navigate the uncertainties that characterize contemporary education (Fullan, 2023).
Addressing difference and diversity
Building capacity for inclusiveness is no longer a choice but a necessity, given the increasing diversity of student populations across almost all education systems (Bhowmik et al., 2023; Dimmock and Walker, 2005). This component has become the cornerstone of effective school leadership since schools are increasingly challenged to manage heterogeneity equitably and effectively (Tan and Dimmock, 2019). Leaders must not only understand the unique needs of diverse student groups but also take initiative to anticipate how these needs may evolve and actively consider these needs across various aspects of their practice (Chilah Abdelkader et al., 2024). Consequently, such an inclusive leadership perspective has the potential to enhance relational trust and proactive work behavior among organizational members (Rogozińska-Pawełczyk and Sudolska, 2024).
To successfully embrace the demographic, cognitive, and value-oriented shifts, school leaders are urged to recognize multiple intersections among diverse school actors and the community values within the complex school ecosystem (Diaz-Gibson and Daly, 2020). This requires thoughtful navigation across all subsystems and participants in education systems, extending beyond the individual classrooms (Senge et al., 2012). Such an approach calls for a shift in mindset and practices to ensure that every student has access to quality education, while also providing tailored support to all school stakeholders based on their individual circumstances (Kuyurtar and Korumaz, 2023).
This approach also emphasizes the importance of balancing recognition and appreciation of individuals’ unique backgrounds and experiences with the pursuit of coherent school improvement (Hargreaves and Fink, 2006; Óskarsdóttir et al., 2020). School leaders can advocate for whole-school improvement by deliberately initiating inclusive actions that align individual needs with collective goals, ensuring that all school stakeholders receive the necessary individualized support and a sense of collective responsibility that should not be overlooked (Gunnþórsdóttir et al., 2024).
Technology leadership
Technology leadership is a critical component of FRSL, referring to the type of leaders who are “prepared to lead changes in schooling, as catalyzed by technology and its ubiquitous presence” (Mcleod and Richardson, 2011: 216). It has already become an integral part of school leadership literature, with the argument that the learning needs of tomorrow's learners are unlikely to be met by traditional leadership and learning models (Richardson et al., 2024). This aspect underscores the dynamic role of technology in establishing an evidence-based vision for future-ready schools (Kurland et al., 2010), in which school leaders intentionally exercise technology leadership to promote an OL culture rather than merely reacting to technological change (Banoğlu et al., 2023).
As technological landscapes evolve with new tools and upgrades (OECD, 2024), school leaders must benchmark educational technologies and strategically anticipate the challenges and opportunities associated with school community needs (Quong and Walker, 2010; Zhao, 2025). Such a proactive orientation enables leaders to act in advance of external pressures to adopt technology, deliberately integrating bottom-up pedagogical needs with top-down strategic processes (Dexter, 2018). This boundary-spanning approach supports ethically driven technology integration initiatives that are guided by well-considered plans aligned with the school's strategic intent, thereby promoting equitable access for all students (Liu et al., 2024).
Systems thinking
The school functions as a dynamic system, and its adaptability to the environment hinges on its leaders’ ability not only to thrive within that context but also to replicate successful practices and ultimately influence the broader ecosystem (Dimmock and Walker, 2004; Fullan, 2023). Systems thinking underscores the necessity of a multidimensional view, enabling school leaders to switch cognitively among different perspectives to address complex issues (Shaked and Schechter, 2020). This capacity supports leaders in moving beyond short-term problem solving toward proactive action that shapes system-level conditions over time. Consequently, systems thinker school leaders “proactively and naturally take into account large portions of the educational system because they know that context matters, for better or for worse, and that part of their work involves changing the context, which can only be accomplished by taking action in the broader sense” (Shaked and Schechter, 2020: 108).
This holistic perspective enables school leaders to make informed decisions that account for the emotional needs of students, staff, and the wider community (Gonzales, 2020). In this way, school leaders advocate for a far-sighted systemic understanding of school leadership that embraces uncertainty and tensions, viewing them as opportunities for “co-creating the future” rather than setbacks. As noted by Senge et al. (2015), “Change often starts with conditions that are undesirable, but artful system leaders help people move beyond just reacting to these problems to building positive visions for the future” (29).
Future-ready school leadership in relation to other relevant constructs
FRSL is conceptualized in this study as a proactive, future-oriented leadership capacity through which school principals intentionally anticipate emerging challenges, strategically navigate uncertainty, and shape enabling conditions for continuous organizational learning and school improvement. It reflects a genuine leadership orientation explicitly oriented toward future conditions—social, technological, and organizational—requiring leaders to act ahead of change, align individual and collective learning processes, and mobilize schools as adaptive systems. In this sense, FRSL is positioned not as a variant of existing leadership models, but as a distinct construct that captures a configuration of leadership competencies specifically attuned to uncertainty, complexity, and long-term educational sustainability.
School leadership research has long recognized that no single model can fully capture the breadth of behaviors required to lead contemporary schools effectively (Bellibaş et al., 2021; Neumerski et al., 2025; Printy et al., 2009). Accordingly, FRSL is theoretically grounded in and extends established accounts of leadership models, including inclusive, strategic, technology, and transformational leadership. More precisely, the addressing difference and diversity dimension aligns with inclusive leadership scholarship, which emphasizes equity, recognition of difference, and responsiveness to diverse school communities (Adams et al., 2023), while extending this work by positioning inclusivity as part of the overall proactive and outward-looking school leadership orientation (Bhowmik et al., 2023; Bryant and Walker, 2024). Secondly, the technology leadership dimension relates to established models of technology leadership that focus on guiding technology integration to enhance teaching, learning, and school improvement (Mcleod and Richardson, 2011), yet also reframes technology leadership within an anticipatory and innovation logic (Hashim and Kearney, 2025). The strategic foresight dimension is primarily rooted in strategic leadership literature highlighting present and future environmental influences (Carvalho et al., 2021), long-term planning (Davies, 2011), and navigating the dynamic complexity of school reforms (Chen, 2008), with further advancements by conceptualizing foresight as a proactive capacity for innovativeness to anticipate change and mobilize OL culture ahead of emerging pressures (Quong and Walker, 2010). Finally, the systems thinking dimension in particular, as well as the overall structure of FRSL, are aligned with transformational leadership theory, particularly its emphasis on whole-school change, shared vision, and collective capacity building (Gümüş et al., 2026a; Leithwood and Jantzi, 2006; Mulford and Silins, 2009); while transformational leadership primarily captures relational and motivational processes that enable change, FRSL preserves this systemic orientation and extends it by explicitly foregrounding school leaders’ engagement with system dynamics (Shaked and Schechter, 2020), as well as uncertainty and future conditions (Gurr and Drysdale, 2020), thereby proposing an integrated, future-ready architecture for systemic transformation.
Among the leadership models mentioned above, transformational leadership is selected for investigating the nomological validity of the FRSIL, given its significant role in predicting teacher outcomes and the overall relevance to the FRSL. Transformational leadership comprises five dimensions: idealized influence, inspirational motivation, intellectual stimulation, individualized consideration, and contingent reward. While these dimensions differ conceptually from the four dimensions underpinning FRSL, both leadership approaches are grounded in a shared change-oriented logic. Specifically, each foregrounds leaders’ roles in shaping enabling conditions for OL culture. This convergence in distal outcomes suggests a significant association between the FRSL framework and transformational leadership within the same school leadership nomological network (Wang and Ahn, 2025).
Complementarily, the relationship between FRSL and OL culture elucidates the criterion-related validity of the developed measurement tool. In contemplating the continuum that bridges an organization's past and future, Senge (1990) defines OL culture as the capacity to continually enhance its adaptive and generative learning capabilities, thereby shaping its future. Therefore, it is not surprising that school leaders who embody FRSL skills are expected to cultivate an environment conducive to organizational learning—one characterized by collectively accountable change directed toward the future (Watkins and Marsick, 1997). Consequently, the capacity of FRSL to forecast outcomes related to OL culture provides compelling evidence of the instrument's criterion-related validity. By elucidating this relationship, the developed measurement instrument captures not only the essence of future-ready leadership but also its broader implications for educational practices and outcomes.
Hong Kong context
This research was conducted in Hong Kong, China. The educational system in Hong Kong, characterized by high accountability and rapid policy shifts, mirrors many significant educational movements worldwide over the past two decades (Cheng, 2019). Hong Kong has consistently garnered international recognition as a high-performing education system (Gopinathan and Lee, 2018; Lee and Manzon, 2014). This reputation is bolstered by waves of educational reforms initiated by the Education Bureau since 2001, which aimed to shift the educational focus from rote learning and exam-oriented approaches toward fostering critical thinking, creativity, and lifelong learning skills among students. The transformation is intended to prepare students to thrive in an uncertain and rapidly evolving global economy. Consequently, although the Hong Kong context is shaped by multicultural influences and a unique socio-political environment, it provides a salient and relevant context for investigating the FRSL construct.
Method
This study develops and evaluates the psychometric properties of the FRSLI, a teacher-reported measure. As the baseline analytical framework, the steps advocated by Hinkin (1998) were followed in the development and validation of the FRSLI. Additionally, the American Educational Research Association (AERA), American Psychological Association (APA), and National Council on Measurement in Education (NCME) standards (AERA, APA & NCME, 2014) informed the systematic assessment of the instrument's psychometric properties. We began by ensuring the content and face validity, comprising two steps: item generation and cognitive interviewing. For psychometric evaluation, internal reliability, construct validity, nomological, criterion-related, and incremental validity, as well as measurement invariance across school stage and teacher gender, were assessed across three studies. Collectively, these procedures ensured that the FRSLI was developed and validated in accordance with established best practices in educational and psychological measurement, providing a robust foundation for both theoretical and applied research (Camara, 2014).
Instrument development
To operationalize the newly developed measurement construct effectively, a comprehensive literature review was conducted. By screening various scholarly databases, including Web of Science, Scopus, Education Resources Information Center (ERIC), ProQuest, and Google Scholar, we iteratively identified journal articles, books, and book chapters for inclusion. Additionally, we reviewed previously developed measurement tools to deepen our understanding of the proposed components before creating a generic item pool. Although no item wording or content was directly adopted from existing scales, the conceptual emphases of the pooled items were guided by sensitizing concepts drawn from prior measures of diversity and inclusion (Al-Atwi and Al-Hassani, 2021; Moya et al., 2020), technology leadership competencies/strategies (Banoğlu, 2012; Weng and Tang, 2014), and leadership and trust (Daly and Chrispeels, 2008), which were reinterpreted through the proactive and anticipatory leadership lens guiding this study.
Drawing on the literature review and reflexive discussion, we first identified the four key components of the proposed FRSL framework, as discussed above. Using the same literature set, 50 items were developed for the generic item pool. Subsequently, the generated items were presented to a panel of ten experts for review. This panel included five professors of educational leadership from four universities across three countries, three methodology experts, and two practicing school leaders. Their diverse expertise provided a robust foundation for conducting an expert review to validate the items.
The expert review process, as described by DeVellis and Thorpe (2021), involved categorizing each item into three groups: “essential,” “useful but not essential,” or “not necessary.” Additionally, panel members’ comments and suggestions were obtained through open-ended questions. To ensure a high level of agreement, items considered “essential” by at least 80% of the experts were included in the final instrument. Conversely, items that did not meet the agreement threshold were excluded, in accordance with Lynn’s (1986) criteria. As a result, a draft instrument with 32 remaining items was created.
As a final stage of item refinement, we examined how potential respondents and school leaders interpreted the items through cognitive interviews. We conducted several rounds of interviews with two teachers, one teacher leader, and two vice principals to elicit insights by asking respondents to explain their understanding of each item and the rationale for their responses (Willis, 2015). As a result of the interviews, we refined the wording of three items to represent the relevant components more accurately.
Psychometric evaluation of the instrument
As reported in previous sections, an expanded review of the literature in the item generation process and the subsequent expert panel endorsement of the revised draft items support the content and face validity of this instrument. In addition, since this study represents the first comprehensive examination of the FRSLI's psychometric properties, we aim to assess its internal reliability (the degree to which all items measure the same construct), construct validity (the extent to which the instrument accurately reflects the theoretical construct), nomological validity (the degree to which the FRSLI correlates with the transformational leadership construct as perceived by teachers), criterion-related validity (the ability of the FRSLI to predict the organizational learning construct as perceived by teachers), incremental validity (the extent to which the FRSLI explains additional variance in organizational learning beyond transformational leadership after controlling for its effect), and measurement invariance across school stage and teacher gender (the extent to which the FRSLI measures the same constructs consistently across different groups).
To evaluate the psychometric properties of the FRSLI, three independent studies were conducted. Details of these studies are presented in Table 1. Study 1 employed exploratory factor analysis (EFA) to reduce items and identify the underlying factor structure (n₁ = 209). Study 2 used confirmatory factor analysis (CFA) and structural equation modeling (SEM) to assess construct, convergent, and discriminant validity, measurement invariance, and the instrument's nomological relationship with transformational leadership (n2 = 269). Study 3 examined criterion-related validity by testing the FRSLI's predictive power for teachers’ perceptions of OL culture, both independently and after controlling for the effects of transformational leadership (n3 = 244).
Summary of studies, procedures, and data/sample characteristics.
Nomological and Criterion-Related Validity
“Nomological validity determines whether the scale demonstrates the relationships shown to exist based on theory or prior research” (Hair et al., 2019: 162). To that end, the MLQ-5X (Multifactor Leadership Questionnaire-5X; Bass and Avolio, 2004) was used to assess the nomological validity of the FRSLI by examining its association with transformational leadership. The MLQ-5X is a 20-item scale measuring five dimensions of transformational leadership. It demonstrates high internal consistency and has been validated through factor analysis (Antonakis et al., 2003; Hemsworth et al., 2013).
For criterion-related validity, the DLOQ (Dimensions of Learning Organisation Questionnaire; Watkins and Marsick, 1997) was used to assess whether FRSLI scores predict perceptions of OL culture. The DLOQ consists of 21 items across seven factors representing aspects of organizational learning. It is a psychometrically sound measurement instrument with high reliability scores across diverse national contexts (Watkins and Dirani, 2013).
Data collection process
Data were collected as part of a research and development project aimed at the leadership development of school leaders in the Hong Kong SAR. Before data collection, ethical approval was obtained, and an informative letter outlining the project was sent to local schools.
The study comprised three independent datasets collected across two phases. In Phase 1 (Study 1), 261 teachers from 13 invited schools (7 secondary and 6 primary) across three regions of Hong Kong participated in an online survey. Schools were invited using maximum variation based on size, location, and stage. As the analyses in Study 1 did not require a nested data structure, teachers were not asked to provide school information. Data from 42 teachers who consented but did not respond were excluded, and five cases with item-level missing values were confirmed as missing completely at random (MCAR; χ2 = 55.91, df = 55, p > .05) and removed via listwise deletion, resulting in 214 usable responses. Study 1 employed exploratory factor analysis (EFA) to identify the instrument's factor structure and assess reliability. Phase 2 involved 587 teachers from 20 schools for Studies 2 and 3. Twelve cases with item-level missing values were MCAR (χ2 = 16.72, df = 10, p > .05) and removed, leaving 575 responses, which were randomly split into 310 for Study 2 (construct validity, CFA, measurement invariance, and nomological validity via SEM with MLQ-5X) and 265 for Study 3 (criterion-related validity predicting organizational learning via DLOQ and regression analyses).
Data cleaning
All datasets underwent additional screening for careless or inconsistent responses. Following recommendations by Zhang and Conrad (2014), respondents with extremely fast completion times or straight-lining behaviors were removed (5 cases in Study 1, 20 cases in Study 2, 7 cases in Study 3). Completion thresholds were set at 147 s for 57 items (Study 1) and 236 s for 68 items (Studies 2 and 3). Besides, multivariate outliers were identified using Mahalanobis distance (Type I error = .001) and removed (21 in Study 2, 14 in Study 3). The final sample sizes were n₁ = 209 for Study 1, n₂ = 269 for Study 2, and n₃ = 244 for Study 3. Teachers’ demographic characteristics in each sample (i.e., gender, teaching experience, and age for n₁, and additionally educational background and school stage for n₂ and n₃) are presented as frequencies and percentages in Table 1. To assess sample representativeness, the gender distribution of the study samples was compared with official statistics for the Hong Kong primary and secondary teacher population (i.e. 33% male, 67% female; Education Bureau, 2025). As shown in Table 1, the gender proportions across the three samples closely align with the official population distribution in Hong Kong.
Results
Study 1: Initial item reduction by EFA
In this section, we present the findings from EFA, which disclosed the underlying factor structure of the FRSLI. Utilizing a sample of 209 teachers, this study aimed to identify the factor structure of the FRSLI measurement construct. To that end, EFA serves as the initial empirical step toward obtaining psychometrically valid measurement (Comrey and Lee, 1992; Hair et al., 2019).
The FRSLI was rated on a 5-point Likert-type response format ranging from “never” to “always.” The EFA procedure was conducted using ML estimation and oblique rotation, as the complete data and extracted factors were assumed to be interrelated (Field, 2013). The draft instrument, comprising 32 items resulting from expert panel reviews, was analyzed using the standard (principal axis) factor analysis procedure. The seven rounds of the iterative EFA process yielded a four-factor, 17-item measurement construct, with all factor loadings exceeding .50 on their primary factors (Costello and Osborne, 2005), higher communalities above .40 (Hair et al., 2019), no cross-loadings above .20 (Howard, 2016), and each item contributing to internal consistency reliability, with Cronbach's α coefficients exceeding .70 for each factor (Field, 2013).
During the first iteration, two items with lower communalities tan .40 were eliminated, as they contributed limited shared variance to the overall factor solution. In the second iteration, three items with primary factor loadings below .50 across three different factors were removed, as they failed to demonstrate a minimally acceptable association with any latent construct. The third iteration identified four items that exhibited substantial cross-loadings on two factors, with loading differences smaller than the .20 cutoff, indicating insufficient discriminant alignment; these items were therefore excluded. In the fourth iteration, three items were removed due to weak primary loadings on their extracted factors, reflecting inadequate construct representation. The fifth and sixth iterations involved again cross-loading refinements for two and one items that respectively loaded under two different factors. The iterative process was terminated at the seventh iteration with the remaining 17 items, which met the predefined stopping criteria: a stable four-factor solution with no salient cross-loadings over .20, all retained items loading above .50 on their primary factors, satisfactory communalities over .40, and acceptable internal consistency over .70 for each factor. This final solution demonstrated both statistical adequacy and conceptual interpretability, supporting its retention for subsequent confirmatory analyses.
Table 2 presents the items, communalities, factor loadings, and internal consistency reliability estimates, sorted by factor loading magnitude.
EFA factor structure and internal consistency reliability.
The extracted four-factor, 17-item EFA solution accounts for 70.38% of the total variance with an adequate sample size, as indicated by a Kaiser–Meyer–Olkin (KMO) index value of .960 and the significance of Bartlett's test of sphericity (χ2 = 3056.22, df = 136, p < .01). The four factors of the FRSLI exhibit high internal consistency reliability, with Cronbach's α values ranging from .875 to .934.
Because the conceptualization of FRSL is grounded in a holistic, overarching theoretical framework, the FRSLI is intended to yield a total score representing a composite construct that can be measured on a higher-order factor. The use of a total or average score from a multi-item measurement scale violates a primary assumption of Cronbach's α estimation: that all items contribute equally to the overall construct (Hayes and Coutts, 2020). To tackle this issue, McDonald's ω coefficient was additionally calculated for each factor as well as for the overall construct, resulting in similarly high values ranging from .876 to .935, with the composite construct achieving an even higher coefficient of .963, identical to that of Cronbach's α.
Study 2: CFA and further validity analyses
Using an independent sample of 269 teachers, CFA was conducted to validate the factor structure derived from EFA in the earlier study. The results of the CFA model comparisons are presented in Table 3. As shown in the table, the initial EFA-informed four-factor 17-item measurement model yielded a chi-square value of 273.510, based on 153 distinct sample moments and estimated with 40 parameters, resulting in 113 degrees of freedom (p < .001). All regression paths and covariances were also statistically significant (p < .001).
CFA model comparisons.
CFA models are typically evaluated using several model-data fit criteria, such as a chi-square to degrees of freedom ratio (χ2/df) below 3 (Collier, 2020), comparative fit index (CFI) and Tucker-Lewis index (TLI) values of .90 or higher (Hair et al., 2019), and a root mean square error of approximation (RMSEA) value of .08 or lower (Brown, 2015). These criteria represent a comprehensive strategy for both absolute and relative fit indices (Hu and Bentler, 1999).
Based on these criteria alone, we could have concluded that the initial measurement model demonstrates an “acceptable fit” to the data (χ2 = 273.51, df = 113, χ2/df = 2.42, CFI = .95, TLI = .94, RMSEA = 0.073). However, upon examining the standardized residual covariance matrix, we identified a notably high covariance value of 2.90 between Item 08 and Item 29, which exceeds the threshold of 2.58 (Jöreskog and Sörbom, 1993). Upon reviewing the semantics of these items (i.e., Item 08: “My school principal is sensitive and responsive to the well-being of others”; Item 29: “My school principal fosters caring relationships during times of change”), we realized that both reflect an emotional-support aspect of leadership that was not explicitly incorporated into the relevant domains.
A closer comparison with the retained items within the dimensions of “addressing diversity” and “systems thinking” indicated that the core conceptual content of these domains remained fully represented. Specifically, the target school leadership behavior of inclusivity in the “addressing diversity” factor continued to be captured through items emphasizing recognition of diversity, equal treatment, and the active promotion of inclusive mindsets (Items 09–12), while a leader's systemic comprehension in the “systems thinking” factor was sufficiently represented by items focusing on responsiveness to external change, flexibility under uncertainty, relational leadership during transitions, and future-oriented planning (Items 30–32). In contrast, Items 08 and 29 primarily foreground affective care and interpersonal support, which—although important leadership attributes—might be conceived as extending beyond the intended construct boundaries of FRSL as operationalized in this scale. Their removal, therefore, enhanced construct distinctiveness without compromising theoretical coverage of the targeted dimensions.
Additionally, further analysis, as indicated by modification indices, suggested that removing these two items would significantly reduce the χ2 value. To address this issue, we developed an alternative four-factor, 15-item model for comparison with the original 17-item model.
As shown in Table 3, the modified 15-item model resulted in a significant reduction in the χ2 value (χ2 = 160.95, df = 84, Δχ2 = 11.558, Δdf = 29, p < .001). More importantly, the alternative 15-item model demonstrated a good fit to the data (χ2/df = 1.91, CFI = 0.97, TLI = 0.97, RMSEA = 0.058). Consequently, we used this improved measurement model to conduct a higher-order CFA, yielding a composite variable of FRSL. The CFA statistics and the factor structure of the Four-factor 15-item measurement model are illustrated in Figure 1.
Four-factor 15-item measurement model with standardized coefficients.
Nomological validity
As noted by Hair et al. (2019), adding a second-order layer to a saturated model with good fit indices alters the number of estimates and degrees of freedom, but is not expected to yield a statistically significant change in the χ2 value. This is evident in Table 3, which indicates that the first- and second-order CFA models produced nearly identical model-data fit indices, with a nonsignificant chi-square difference (χ2 = 164.66, df = 86, Δχ2 = 3.71, Δdf = 2, p = .16).
As a remedy, Hair et al. (2019) recommend verifying the nomological validity of the second-order measurement model in relation to its relevance to other constructs. Accordingly, the structural model illustrated below in Figure 2 was tested using SEM to determine whether the overarching/second-order FRSLI construct is significantly related to the transformational leadership construct.
SEM for nomological validation.
The SEM presented above simultaneously tests a comprehensive nomological network based on regression and covariance matrices, wherein the statistical assessment of nomological validity extends beyond examining simple correlations between newly developed and existing constructs (Hagger et al., 2017). The resulting fit indices for the nomological validation model (χ2 = 313.92, df = 165, χ2/df = 1.90, CFI = .97, TLI = .96, RMSEA = 0.058) indicate a good model fit, demonstrating that FRSL and transformational leadership are components of the same effective school leadership network and supporting the observed correlation of 0.85 (p < .001).
To better understand the independent and combined effect of FRSL with transformational leadership on school culture, we further examined the predictive and incremental validity of the FRSLI for OL culture, which serves as a criterion variable in Study 3, after controlling for transformational leadership (MLQ-5X). The results confirmed the independent significant explanatory power of FRSLI (see Study 3).
Convergent validity
CFA results have shown that all standardized factor loadings were statistically significant at the .001 level and well above the recommended threshold of .5. Additionally, estimated CR values ranged from .85 to .90, corroborating the previously reported Cronbach's α and McDonald's ω coefficients from Study 1. Regarding convergent validity, the estimated AVE values ranged from .60 to .75, collectively affirming the high construct and convergent validity of the proposed measurement model (Hair et al., 2019).
Discriminant validity
To evaluate discriminant validity, we followed Hair et al. (2019) and employed the Heterotrait-Monotrait Ratio (HTMT) criterion, which addresses limitations of conventional methods. For a stricter interpretation, the HTMT should ideally be below .85, as Henseler et al. (2015) suggested, or below .90 for a more lenient threshold proposed by Gold et al. (2001). In our study, although the HTMT ratios for the relationships between the systems thinking and strategic foresight dimensions, as well as between systems thinking and addressing diversity dimensions, were .86 and .87—slightly exceeding Henseler et al.'s (2015) threshold of .85—they remained below the .90 cutoff suggested by Gold et al. (2001). All other correlation pairs also supported the discriminant validity of the instrument, resulting in HTMT ratios between .80 and .85.
Measurement invariance
Measurement invariance ensures that the instrument consistently measures the same construct across various populations, thereby strengthening the validity of the results. The stability of the final measurement model across gender (male vs. female) and school level (primary vs. secondary) was assessed through multi-group confirmatory factor analysis (MGCFA). For this analysis, the sample size was reduced to 260, as nine respondents did not disclose their demographic information (nmale = 93, nfemale = 167; nprimary = 119, nsecondary = 141). Initially, a completely free configural invariance model was computed for both gender (χ2 = 267.82, df = 168, χ2/df = 1.60, CFI = .96, TLI = .95, RMSEA = 0.048) and school level (χ2 = 267.39, df = 168, χ2/df = 1.59, CFI = .96, TLI = .95, RMSEA = 0.048) as the baseline model. Given the resulting “good fit” indices, we found evidence that the measurement model was stable across male and female respondents.
Additionally, we specified equality constraints on the factor loadings to gauge metric invariance for gender (χ2 = 276.21, df = 179, χ2/df = 1.54, CFI = .96, TLI = .96, RMSEA = 0.046) and school level (χ2 = 276.07, df = 179, χ2/df = 1.54, CFI = .96, TLI = .96, RMSEA = 0.046). To compare the restricted metric invariance model with the baseline configural model, we used the χ2 difference statistic, along with the additional ΔCFI and ΔMcDonald's NCI indices, with cutoff values of .01 and .02, respectively (Cheung and Rensvold, 2002). The results from the metric invariance model provided strong statistical evidence that the fixed factor loadings fit the data with the same factor structure for both male and female respondents (Δ χ2 = 8.39, Δdf = 11, p = .68, ΔCFI = .001, ΔNCI = .004), as well as for primary and secondary school respondents (Δ χ2 = 8.68, Δdf = 11, 3jp = .65, ΔCFI = .001, ΔNCI = .004).
Study 3: Replication for criterion-related and incremental validity
In the last study, we extended our investigation by exploring the criterion-related validity of the FRSLI in predicting the organizational learning construct (DLOQ) as perceived by teachers. Using a sample of 244 teachers, this study employs regression analysis to assess the extent to which the FRSLI predicts key criterion outcomes associated with perceived organizational learning in schools. The results below illuminate the relationship between the proposed FRSLI components and teachers’ perceptions of organizational learning.
We used multiple regression analysis to set the four dimensions of the FRSLI as potential predictors of perceived organizational learning in schools. Table 4 displays the statistical results from the regression analysis.
Regression results from the OL culture prediction model.
A check for multicollinearity revealed that all VIF values were below 5 and Tolerance values above 0.10, indicating no concerns regarding multicollinearity among the FRSLI dimensions (Hair et al., 2019). We conclude that the four dimensions of the FRSLI can explain 59.50% of the total variance in perceived organizational learning (R2 = .60, F(4, 239) = 87.61, p < .001). Three components also demonstrated significant predictive power for perceived organizational learning in schools. Overall, the composite score from the FRSLI was strongly correlated with teachers’ perceptions of organizational learning in their schools (r = .77, p < .001).
To demonstrate that the overall FRSLI has unique explanatory power beyond transformational leadership as a test of incremental validity, we also conducted a stepwise multiple regression analysis. No multicollinearity concerns were detected between the FRSLI and MLQ-5X (VIF = 3.29, tolerance = 0.30), indicating that their correlation does not prevent the constructs from distinctively explaining teachers’ perceptions of OL culture. The FRSLI significantly predicted OL culture (β = .41, p < .001), even after accounting for the effect of the MLQ-5X (β = .42, p < .001), demonstrating that the FRSLI is a salient construct with a uniquely meaningful contribution to teachers’ perceptions of OL culture (ΔR2 = .05, F(1, 241) = 34.51, p < .001).
In interpreting incremental validity, prior empirical research has consistently treated a ΔR2 value of .05 as a small-to-medium effect (Cohen, 1988) and as a meaningful contribution when statistically significant (McGill et al., 2026). For example, Bing et al. (2004) reported an incremental validity value of ΔR2 = .05 (p < .01) as sufficient evidence supporting their personality scale. Similarly, Nelson and Canivez (2012) found that additional variance ranging from 1.9% to 7.6% constituted meaningful incremental validity in explaining intellectual abilities. Moreover, Mackay et al.'s (2017) meta-analytic investigation of employee engagement demonstrated that incremental contributions of 2% to 6% were both common and substantively meaningful for predicting employee effectiveness. On this basis, the present findings indicate that the observed incremental contribution of the FRSLI is consistent with established empirical standards and supports the conclusion that the FRSLI provides meaningful and theoretically relevant explanatory value beyond transformational leadership.
Discussion
Despite the increasing importance of FRSL as a unique form of school leadership and management practices, the topic has received little empirical attention (O’Brien and Robertson, 2009; Pietsch and Mah, 2025). Researchers have not attempted to conceptualize its factorial components or its relationships with related leadership constructs and outcomes. Addressing this gap, the current study contributes to the school leadership literature by developing and validating a new, integrative conceptualization of FRSL, offering a robust construct to advance both research and practice.
To lead a future-ready school, school leaders must calibrate their prevailing leadership practices to shape their schools for multiple imagined and unknown future scenarios. Although practices may shift and adapt across stages and contexts, future-ready leaders integrate key beliefs into their future-oriented thinking. Key among these beliefs is that while looking to the future, future-ready school leaders maintain a sense of social responsibility and are committed to creating equitable access to resources and opportunities for all students. Future-ready school leaders also continuously scan for and leverage emerging technologies to enhance educational ecosystems. They ensure that future-focused frameworks, cultures, and structures are in place, fostering innovative ways of working.
Based on the above overall perspective, we conceptualized FRSL as enacted, observable leadership practices and intentionally used teachers’ ratings. The instrument we developed through item generation, expert review, cognitive interviews, and three consecutive studies to measure FRSL reflects our conceptualization of the construct. FRSL presents a conceptually and psychometrically distinctive set of leadership practices that includes four components: strategic foresight, addressing difference and diversity, technology leadership, and systems thinking.
The unique contributions of each dimension of FRSL underscore the necessity for leaders to adopt a multifaceted approach in their practices. Strategic foresight enables leaders to anticipate future challenges and to encourage teachers to move beyond past-bound problem-solving, thereby cultivating innovative, future-oriented responses to emerging educational demands (Balser and Tafuro, 2025). Addressing differences and diversity foregrounds leaders’ heightened awareness and sustained problem-seeking capacity to recognize, value, and continually draw on heterogeneous student and staff perspectives to generate context-responsive and future-oriented solutions for school improvement (Marques, 2020). Additionally, the technology leadership domain underscores the critical importance of integrating digital resources into teaching practices to prepare students for a digital future (Richardson et al., 2024). Lastly, systems thinking encourages leaders to adopt a holistic perspective of the school ecosystem, facilitating collaboration among all educational stakeholders and enabling a comprehensive, adaptive approach to meet today's multifaceted challenges (Gurr and Drysdale, 2020; Shaked and Schechter, 2020).
The novelty of the research lies in the perspective that FRSL presents a timely, unique repertoire of leadership attributes and practices explicitly designed to navigate uncertainty, complexity, and long-term educational sustainability through a proactive strategic approach (Hargreaves and Fink, 2006; Quong and Walker, 2010; Shaked and Schechter, 2020). Existing leadership models have yet to foreground the simultaneous integration of technological foresight, equity responsiveness, systems-level thinking, and proactive strategic anticipation as a unified leadership orientation. In this sense, FRSL represents a configuration-based construct that captures how these dimensions operate collectively to shape organizational learning culture in preparation for multiple and uncertain futures. This integrative orientation is further supported by a series of empirical evidence presented in this study, including the replicated four-factor structure, measurement invariance across school stages and teacher gender, and discriminant validity relative to transformational leadership, which together provide a psychometrically coherent justification for the construct's multidimensional nature.
Our findings reveal that these dimensions collectively predict a higher-order construct, that is, OL culture, that reflects the relevance of effective school leadership to a positive school culture. Additionally, given that school leadership relies on a context-dependent, comprehensive set of target behaviors, the evidence of incremental validity for this newly developed, multifaceted construct—relative to the widely accepted framework of transformational leadership—is particularly noteworthy. After controlling for transformational leadership, the FRSLI still demonstrated unique predictive power for teachers’ perceptions of OL culture. This suggests that FRSL encompasses additional aspects crucial to fostering an adaptive and innovative educational environment.
Overall, the 15-item measurement demonstrated satisfactory psychometric qualities. Specifically, the instrument demonstrates high reliability, a replicable factorial structure, a reasonable relationship with transformational leadership, and additional predictive power for perceived organizational learning in schools. These results provide consistent and coherent evidence for the scale's validity and potential significance. The scale includes a reasonable number of items, ensuring a manageable cognitive load for most respondents. The estimated time required to complete the survey is 10 min or less, making it suitable for inclusion in future survey research and replication attempts.
Implications
From a practical standpoint, the FRSLI may serve as a valuable tool for principal professional development diagnostics, enabling the assessment of school leaders’ competencies and the identification of areas for growth aligned with the demands of a rapidly changing educational landscape. Using the four dimensions of the FRSLI as a framework for professional development, educational institutions can tailor training programs to enhance leaders’ ability to navigate complex challenges, implement inclusive practices, and integrate technology effectively in their schools. Such diagnostic information may also support the design of targeted leadership development pathways, mentoring programs, and leadership standards aligned with future-ready competencies. Additionally, the instrument can inform school improvement efforts by providing actionable insights that guide strategic decision-making and foster a culture of innovation.
Moreover, the FRSLI's applicability extends to policy evaluation, offering a means to assess the effectiveness of school leadership initiatives and educational reforms from the teacher perspective. At the system level, policymakers may utilize FRSLI data to monitor leadership capacity across schools, evaluate reform initiatives, and inform evidence-based leadership policies and resource allocation. Additionally, adapting the current scales for other educational stakeholders, such as parents and students, could be useful for evaluating the extent to which relevant school leadership practices are associated with desired reforms and policy outcomes from multiple perspectives. This ability to measure and analyze leadership effectiveness in relation to educational reforms and policy goals positions the FRSLI as a critical resource for advancing educational leadership research and practice, ultimately contributing to the development of future-ready schools equipped to meet tomorrow's challenges.
Beyond its practical utility, the present study makes several meaningful contributions to the theoretical landscape of educational leadership research. First and most fundamentally, it addresses a notable gap in the school leadership literature by offering the first empirically grounded construct definition and dimensional conceptualization of FRSL. While the demands of uncertainty, technological change, diversity, and systemic complexity have been discussed extensively in qualitative and normative leadership scholarship (Gurr and Drysdale, 2020; Pietsch and Mah, 2025), FRSL had not previously been operationalized as a measurable, multi-dimensional construct. The present study conceptualizes FRSL as a set of enacted, observable, teacher-perceived leadership behaviors organized around a proactive, future-oriented logic grounded in the proactive behavior model. This conceptualization advances the theoretical precision of future-oriented leadership discourse by translating an aspirational leadership ideal into a replicable, assessable construct.
Second, the identification and validation of four empirically distinct yet theoretically coherent dimensions contribute meaningfully to ongoing debates about the structure of school leadership. The school leadership literature has long grappled with the proliferation of overlapping models and the challenge of integrating them into coherent frameworks (Bush and Glover, 2014; Gümüş et al., 2018; Wang and Ahn, 2025). The FRSL framework advances this conversation by demonstrating that these four dimensions cohere into a unified higher-order construct underpinned by a shared proactive and anticipatory logic. Importantly, this integration is not merely theoretical: the results confirmed the parsimony and goodness-of-fit of the second-order measurement model, providing empirical warrant for treating FRSL as a coherent whole rather than a loose assemblage of independent competencies.
Finally, the study contributes to the broader literature on leadership measurement methodology by illustrating the value of a multistudy, sequentially cumulative validation design in educational contexts. The progression from EFA (Study 1) through CFA, SEM, convergent, discriminant, and measurement invariance testing (Study 2) to criterion-related and incremental validity (Study 3) reflects the full validity argument recommended by AERA, APA and NCME (2014) standards, and demonstrates how rigorous psychometric evaluation can simultaneously advance theoretical understanding and produce practically deployable instruments. This methodological contribution may serve as a useful model for future scale development efforts in educational leadership research.
Methodological limitations and future research
The first limitation concerns the nomological validity of the FRSLI. We have tested this validity only against transformational leadership. As different educational systems yield unique opportunities, we suggest that future research further examine the construct validity of this construct in relation to other school leadership constructs. While a relatively high correlation between FRSL and transformational leadership could also be interpreted as a limitation, it is not entirely surprising, given that both constructs share a change-oriented, future-focused logic. In addition, as van Knippenberg and Sitkin (2013) have argued, a high correlation between transformational leadership and another leadership construct is not inherently indicative of empirical redundancy, provided that the constructs are theoretically distinct, the dimensional content is clearly differentiated, and incremental validity over the criterion variable is demonstrated. Indeed, the leadership literature is replete with examples of well-established constructs that correlate highly with transformational leadership yet retain their independent theoretical standing, including authentic leadership (r = .82; Fuller et al., 2022; r = .75; Hoch et al., 2018), charismatic leadership (r = .88; Rowold and Heinitz, 2007), ethical leadership (r = .83; Brown et al., 2005; r = .70; Hoch et al., 2018), participative leadership (r = .73, Conger & Kanungo), servant leadership (r = .77; Fuller et al., 2022), and transactional leadership (r = .80; Judge and Piccolo, 2004).
The second limitation of the study was that it included only one outcome variable when testing the criterion-related validity. Our findings demonstrated that FRSL is a strong predictor of perceived organizational learning. FRSL may also influence the future readiness of students, teachers, and schools. Future research could examine these dynamics based on the readily available FRSL instrument developed in this study. A third limitation is that, except for the expert review, our survey data were collected from a single educational system, Hong Kong, which limits the scale's external validity. That said, our samples were diverse and representative of the school and teacher populations in Hong Kong. Future examination of FRSLI in other education systems is warranted.
Third, the issue of dimension discriminability warrants attention, as indicated by elevated HTMT values between the systems thinking and strategic foresight dimensions, and between the systems thinking and addressing diversity dimensions. Although these HTMT ratios were below the more lenient 0.90 cutoff, they exceeded the stricter 0.85 threshold. Future research should explore these dimensions in greater depth, potentially leading to item adjustments that enhance their discriminability while preserving the integrity of the overarching future-ready leadership construct.
Conclusion
This article reports on the development and validation of a new measurement instrument, the Future-Ready School Leadership Instrument (FRSLI; see Appendix). It aims to promote dialogue about effective school leadership by designing and testing a multifaceted framework that may inform leaders facing multiple unknown futures. Future-ready leaders enhance their awareness and capacities as system thinkers with strategic foresight, intentionally address differences and diversity, and understand the power of new technologies. Although the FRSLI was validated for these four components, we do not consider them definitive. Instead, this initial conceptualization may serve as a point of reference for further understanding of FRSL. The development and integration of new dimensions and components relevant to our conceptualization is warranted and encouraged, as the framework, like school leadership itself, continues to evolve.
Footnotes
Funding
This research is funded by the Hong Kong Jockey Club Charities Trust through the Trust-initiated JC InnoPower Education Fellowship program (S/N: 2024-0052-001).
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Supplemental material
Supplemental material for this article is available online.