Transformational and transactional leadership styles are part of the Full Range Leadership Model proposed by Bass and Avolio (1994). In educational contexts, transformational leadership has garnered increasing attention due to its capacity to foster emotionally positive, collaborative learning environments centered on students' holistic development. This leadership style is structured around four core dimensions: idealized influence, inspirational motivation, intellectual stimulation, and individualized consideration (Bass and Riggio, 2006).

First, idealized influence refers to the leader serving as a role model, guided by a clear vision, strong values, and self-confidence. This dimension fosters admiration, respect, and commitment among team members (Lan et al., 2019). Second, inspirational motivation pertains to the leader's ability to communicate shared goals in an enthusiastic and persuasive manner, generating optimism, group cohesion, and a sense of purpose among both teachers and students (Budur and Poturak, 2021). Through clear and purposeful communication, transformational leaders mobilize their teams toward challenging goals, overcoming obstacles and adapting with resilience and flexibility to changing contexts, thereby achieving substantial academic improvement (Berkovich, 2020).

Third, intellectual stimulation encourages a climate of critical thinking and innovation, promoting educators and learners to question assumptions, propose creative solutions, and make autonomous decisions. This dimension has proven particularly effective in catalyzing pedagogical innovation and adapting to increasingly complex and dynamic educational settings (Martínez, 2014).

Finally, individualized consideration is expressed through the leader's personalized attention to each group member, recognizing their unique characteristics, strengths, and developmental needs (Gorgulu, 2019). This practice promotes inclusive and cooperative learning environments, fostering intrinsic motivation and contributing positively to students' emotional wellbeing and academic performance (Shen et al., 2020).

A growing body of research has demonstrated that transformational leadership is associated with greater professional satisfaction, increased teacher commitment, reduced burnout, and improvement of organizational culture in schools (Avci, 2015; Berkovich and Eyal, 2021). Additionally, this leadership style positively influences the development of emotional intelligence in both teachers and students (Flores, 2023). By promoting empathy, active listening, and emotional validation, transformational leaders help create positive affective climates in which individuals feel understood, safe, and valued (Rojas-León et al., 2023; Valdés and Pérez, 2023). All of this fosters an educational setting conducive to meaningful learning, genuine collaboration, and socio-emotional development for all members of the educational community.

In contrast, transactional leadership is grounded in a logic of exchange, in which the relationship between leader and subordinates is based on conditional rewards and oversight of rule and goal compliance (Bass, 1985). This style is characterized by active supervision, correction of deviations, and performance driven by a system of rewards and punishments (Bass and Avolio, 1994). Specifically, it is structured around two key dimensions: goal attainment and management-by-exception. The transactional leader establishes expectations and targets without providing ongoing guidance and may deliver critical feedback based on failure to meet predefined agreements (Lan et al., 2019).

In educational contexts, transactional leadership can be useful in ensuring organizational efficiency and the completion of specific academic tasks (Sheena et al., 2025). Indeed, some studies have shown that contingent reward, that is, rewards explicitly conditioned on the achievement of previously established objectives, constitutes a central dimension of this style and can be effective in contexts involving structured tasks and clear performance criteria, fostering accountability and goal attainment (Jung and Avolio, 2000; Podsakoff et al., 2006). However, when applied in isolation, this leadership style can limit autonomy, hinder pedagogical innovation, and reduce students' intrinsic motivation (Judge and Piccolo, 2004; Lowe et al., 1996). Moreover, it has been associated with school climates characterized by emotional tension, low affective engagement, and a general sense of demotivation, particularly in contexts requiring collaboration, creativity, and shared problem-solving (Leithwood and Jantzi, 2008).

From an integrative perspective, scholars have argued that transactional leadership can complement transformational leadership by providing normative clarity and ensuring the achievement of objectives, particularly in routine or administrative tasks within schools (Abbas and Ali, 2023). In contrast, transformational leadership is more suitable for fostering pedagogical innovation, collaboration, and the renewal of school culture. In this sense, the effectiveness of each style depends on contextual characteristics, reinforcing the need for a hybrid approach that strategically combines structure with motivation, control with inspiration, and efficiency with emotional wellbeing (Hallinger and Wang, 2015).

The study of school leadership in relation to emotional climate remains an emerging field of research (Goetz et al., 2024). Yet advancing in this area is crucial, as leadership practices not only shape organizational functioning but also influence the daily emotional experiences of teachers and students. These experiences, in turn, play a decisive role in the quality of the school climate, in students' readiness to learn, and in the overall wellbeing of the educational community.

This study employed a quasi-experimental design to examine students' emotional experiences and performance during a collaborative task under two classroom leadership conditions: transactional and transformational leadership. The central activity involved constructing figures using LEGO bricks, providing a practical and cooperative context to observe group dynamics and assess how each leadership style influences emotional climate and team performance.

Two experimental conditions were established:

Before the experimental sessions began, all facilitators completed a standardized training program designed to ensure the accurate and consistent implementation of both leadership conditions. To ensure consistency in the implementation of both experimental conditions, facilitators received prior standardized training that included detailed scripts specifying the verbal and non-verbal behaviors characteristic of each leadership style. They also took part in supervised role-playing sessions in which they practiced enacting both styles and received immediate feedback from the research team. This preparation ensured a consistent and accurate representation of the leadership conditions across groups and minimized potential variability in facilitator behavior. The same facilitator remained with each group across both conditions to control for potential external variables related to individual differences.

To reduce potential biases and strengthen the internal validity of the study, the following methodological decisions were implemented:

Distinct tasks for each condition: To avoid learning effects, familiarity, or direct transfer between conditions, two different LEGO figures were used, one for the transactional condition and another for the transformational condition. Both figures were selected to be equivalent in difficulty in terms of the number of pieces, structural complexity, and estimated construction time. This ensured that any observed differences could be attributed to the leadership style rather than the nature of the task. The order of presentation was fixed: the transactional condition was conducted first, followed by the transformational condition.

Fixed order of conditions. The order of presentation was fixed: the transactional condition was always conducted first, followed by the transformational condition. This decision was based on methodological considerations. Beginning with the transactional condition, where facilitators only displayed the image of the final model and adopted a directive, outcome-oriented approach, allowed participants to organize and coordinate within their teams autonomously to achieve a common product. Conducting the transformational condition second ensured that this subsequent experience, characterized by guided instruction, positive feedback, and motivational support, did not influence performance in the initial condition. This sequence also helped minimize potential carryover or contamination effects between tasks, thereby enhancing methodological consistency and experimental control.

Within-subjects design: The same student teams participated in both experimental conditions, allowing for direct comparison of their responses to different leadership styles without the influence of intergroup variability.

Time allotted for the LEGO construction task: Pilot testing indicated that 30 min was an adequate duration for groups of four to complete models of approximately 130 pieces, ensuring the feasibility of the task while also simulating realistic time pressure conditions that facilitated the observation of leadership dynamics. These were low-difficulty models that a single individual could complete in about 30–35 min; however, in groups of four the process was considerably accelerated through task division (e.g., sorting, searching for pieces, assembling). This time limit not only guaranteed the feasibility of the exercise but also enhanced the manifestation of differences in how teams managed the activity under the two leadership styles.

These methodological choices enabled a controlled assessment of how teams respond to distinct leadership approaches while performing comparable tasks (i.e., different but equivalent LEGO figures), minimizing the risk of memory or repetition effects.

This design enabled a controlled analysis of the impact of two leadership styles on emotional variables and task performance in a collaborative educational setting.

The study sample comprised 93 undergraduate students enrolled in the Teacher Education program at a public Spanish university. The average age was 22 years (SD = 3.12). The majority were female (89%), which reflects the typical gender distribution in teacher education programs in Spain and other European countries (OECD, 2019).

For the implementation of the experimental conditions, students were organized into 23 working groups, each composed of four members. This size, frequently used in cooperative learning and group dynamics research (Forsyth, 2019), was considered particularly suitable for LEGO construction tasks, as it balanced diversity of input with efficient coordination and ensured active participation from all members within the allotted time. Group composition remained constant across both experimental conditions (i.e., first the transactional condition and second the transformational leadership condition). In addition, 23 trained facilitators (one per group) assumed the assigned leadership role. To ensure consistency, facilitators received prior training that guaranteed the homogeneous implementation of the leadership conditions. Although their role was essential to the experimental design, facilitator performance data were not analyzed, as the study focused on students' emotional experiences and task performance.

All participants were informed in advance about the study's objectives, the nature of the tasks, and their voluntary participation. Confidentiality was guaranteed, and participants were allowed to withdraw at any point. Informed consent was obtained from every participant, in line with ethical standards for research involving human subjects in educational settings.

The experimental design was implemented across four sessions, each organized and supervised by the research team. Approximately five to six groups participated per session, and each group completed both experimental conditions sequentially within a single 90-min session. This structure enabled efficient time and resource management, while ensuring consistent supervision and procedural control across all groups.

At the beginning of each session, participants were given a general overview of the activity's goals, emphasizing its collaborative nature and connection to teamwork and leadership processes. Specific details about the experimental conditions were withheld to minimize expectation-related biases. Groups were seated at tables arranged to facilitate interaction, cooperation, and coordination among members, thereby fostering an environment conducive to teamwork.

During each session, groups completed two LEGO construction tasks, each aligned with a different leadership style. Both tasks were limited to 30 min. Although the LEGO figures differed across conditions, they were matched in terms of the number of pieces and level of difficulty, which helped control for familiarity, memory effects, or transfer of strategies between tasks. A brief break was provided between the two tasks to reduce fatigue and minimize carryover effects from the first task to the second.

Regarding the administration of instruments, procedures were identical across both leadership conditions. Students completed the Emotional Climate Scale at three time points: before beginning the task (T1), midway through the task (after 15 min of group work; T2), and immediately upon completion (T3). After each condition, students completed the remaining items in the questionnaire, as outlined in the instruments section. In addition to the self-report measures, the researchers independently recorded each group's task performance throughout the activity. Using a structured observation sheet, they noted whether each team had successfully completed, partially completed, or failed to complete the LEGO figure. These observations were carried out in real time and verified immediately after task completion, providing an objective cross-check of the groups' self-reported performance.

All sessions were conducted by the same research team, ensuring uniformity in protocol implementation and rigorous control over experimental conditions. All groups participated under similar physical and environmental settings.

An ad hoc questionnaire was developed, combining previously validated scales with newly constructed items, with the aim of collecting both quantitative and qualitative information regarding participants' experiences. Initially, basic sociodemographic data were included, such as gender, in order to characterize the sample and allow for exploratory analysis of potential individual differences.

Next, a series of descriptive items was designed to assess participants' perceptions of the time allocated for the task and their performance in building the LEGO figure. These items were developed ad hoc by the authors to capture participants' immediate evaluation of these specific and directly observable constructs. An open-ended question was also included to capture participants' qualitative evaluations of their experience during the LEGO activity. Although single-item indicators may raise concerns about reliability, prior research suggests that they are appropriate when constructs are narrow, concrete, and clearly defined (Allen et al., 2022). Given the specificity and direct observability of these constructs (e.g., adequacy of the allocated time; correspondence between the LEGO model and the reference figure), the use of single-item measures was deemed suitable in this study.

The specific items were as follows:

Finally, the Emotional Climate Scale (ECE) developed by Páez et al. (1997) was administered. This instrument comprises six items that assess two dimensions: positive emotional climate (happiness, surprise, and calm) and negative emotional climate (anger, sadness, and fear). Responses were recorded on a five-point Likert scale (1 = not at all, 5 = very much). In the current sample, the scale demonstrated acceptable internal consistency (α = 0.71).

The statistical analysis was conducted using IBM SPSS Statistics version 29.0. Initially, descriptive statistics, including mean, standard deviation, skewness, and kurtosis, were calculated for all study variables to assess data distribution and suitability for further inferential analyses. The normality of the data was examined through the Kolmogorov-Smirnov and Shapiro-Wilk tests, which indicated significant deviations from a normal distribution (p < 0.05). Consequently, non-parametric procedures were applied. Specifically, Spearman's rank-order correlations (ρ) were computed among the main study variables (i.e., emotional climate, time perception, task performance, and time invested) to explore their interrelations.

From a descriptive standpoint, the influence of the two experimental conditions (i.e., transactional vs. transformational leadership) was examined across three specific indicators related to task execution: perceived adequacy of the time allocated, performance in building the LEGO figure, and actual time spent completing the task.

To examine the impact of leadership style on students' emotional climate during the cooperative task, both positive and negative emotional climate were assessed at three time points: at the beginning, during, and at the end of the LEGO activity. Non-parametric Mann-Whitney U tests were used to compare differences between leadership conditions at each time point (between-group analysis), while Wilcoxon signed-rank tests were applied to examine changes over time within each experimental condition (within-group analysis). Because multiple pairwise comparisons were conducted (3 time points × 6 emotions = 18 Wilcoxon signed-rank tests), a global Bonferroni correction was applied to control for Type I error inflation. The adjusted significance level was set at α = 0.0028 (0.05/18). Effect sizes (r) were calculated for each Wilcoxon signed-rank test, values around 0.10 indicate small, around 0.30 medium, and ≥ 0.50 large effects (Cohen, 1988).

Finally, a qualitative content analysis was performed on the open-ended responses regarding the participants' experience with the task. An inductive open coding procedure was carried out with the support of ATLAS.ti (version 25), which facilitated the systematic organization, segmentation, and management of the data. Two researchers coded the data independently, allowing the assessment of intersubjective consistency. Intercoder reliability was calculated using Cohen's κ, which indicated substantial agreement (κ = 0.82). Discrepancies were resolved through discussion and consensus. Preliminary codes were subsequently refined and grouped into broader categories through a process of constant comparison, which were considered final once conceptual saturation was reached. Finally, the frequencies of the categories were calculated to estimate their relative weight and provide a quantitative complement to the qualitative analysis.

Before conducting inferential analyses, the assumption of normality was assessed for the study's main variables (i.e., emotional climate, performance, and time perception) using the Kolmogorov-Smirnov and Shapiro-Wilk tests. In all cases, the results indicated significant deviations from normal distribution (p < 0.05), thereby justifying the use of non-parametric tests in subsequent analyses. Descriptive statistics (i.e., means and standard deviations) and Spearman correlations for each variable are presented in Table 1. Correlations above the diagonal correspond to the transformational leadership condition, and those below the diagonal correspond to the transactional leadership condition.

Under transformational leadership, the total emotional climate was strongly and positively correlated with the positive emotional climate (ρ = 0.89, p < 0.001) and moderately related to the negative emotional climate (ρ = 0.41, p < 0.01). As expected, positive and negative emotional climates were negatively associated (ρ = −0.22, p < 0.05). A higher positive emotional climate was also related to better task performance (ρ = 0.26, p < 0.05), while higher negative emotional climate was associated with poorer performance (ρ = −0.47, p < 0.01).

In the transactional leadership condition, a similarly strong positive correlation was found between total emotional climate and positive emotional climate (ρ = 0.77, p < 0.001), along with a weaker positive association with the negative emotional climate (ρ =.30, p < 0.01). As in the previous condition, positive and negative emotional climates were negatively related (ρ = −0.36, p < 0.01). Additionally, the negative emotional climate showed a small but significant inverse correlation with task performance (ρ = −0.25, p = 0.02), indicating that higher emotional negativity was related to slightly poorer performance outcomes.

Figure 1 displays the results regarding the accuracy of LEGO figure construction, revealing notable differences in group performance based on the type of leadership to which participants were exposed. Under the transformational leadership condition, 78.5% of teams achieved a perfect construction of the projected figure, 19% completed a partial construction, and only 2.5% failed to construct the figure.

In contrast, in the transactional leadership condition, no team achieved a perfect construction. The majority (69.2%) completed a partial construction, while 30.8% were unable to build the figure.

These findings suggest that transformational leadership positively influenced task performance by facilitating clearer understanding of the objective and enhancing team coordination toward a successful outcome.

Figure 2 shows the differences in perceived time availability during the task according to leadership style. In the transactional leadership condition, the vast majority of groups (98.9%) reported that they ran out of time to complete the task, only 1.1% indicated that the time was sufficient, and no group reported having extra time.

In contrast, under the transformational leadership condition, the results were more balanced: 31.6% of the groups stated they had extra time, 41.8% considered the time allocated was sufficient, and 26.6% reported they ran out of time to complete the task.

These results suggest that transactional leadership may foster a task organization more focused on temporal efficiency, whereas transformational leadership, by encouraging greater reflection, cooperation, and elaboration, could be associated with a greater perception of time pressure.

Table 2 presents the results of the non-parametric Mann-Whitney U tests comparing emotional climate across experimental conditions.

Regarding the positive emotional climate (i.e., happiness, surprise, and calm), the group under the transactional leadership condition reported significantly higher levels at the beginning of the task (U = 2,666.50, p = 0.003, r = 0.30), representing a medium effect size, which indicates a slight initial advantage in positive emotions for this group. However, no significant differences were found between groups during the activity (U = 3,507.50, p = 0.78). By the end of the task, the transformational leadership group showed significantly higher levels of positive emotional climate (U = 2,537.50, p < 0.001, r = 0.35), also a medium effect, suggesting that this leadership style was more effective in sustaining and enhancing positive emotions over time.

When examining specific positive emotions, no significant differences were found in happiness at the beginning of the activity (U = 3,177.50, p = 0.17). During the task, however, participants exposed to transformational leadership reported higher levels of happiness (U = 2,663.00, p = 0.003, r = 0.31), and this pattern became even stronger by the end of the activity (U = 2,222.50, p < 0.001, r = 0.46). The effect sizes, medium during the task and large at the end, indicate that transformational leadership had an increasingly strong impact on happiness as the activity progressed.

For the emotion of surprise, significant differences were observed at the beginning of the activity (U = 2,172.00, p < 0.001, r = 0.47), with a large effect and higher levels in the transactional leadership group. However, no significant differences were found during the task (U = 3,200.50, p = 0.20), and only a marginal difference was observed at the end (U = 2,918.50, p = 0.03, r = 0.23), corresponding to a small effect, suggesting that the initial heightened surprise under transactional leadership gradually diminished over time.

In the case of calmness, no significant differences were observed between groups at any point in the experiment (T1: U = 3,228.50, p = 0.23; T2: U = 3,439.00, p = 0.58; T3: U = 3,497.50, p = 0.73), indicating comparable levels of emotional tranquility across leadership conditions.

In contrast, the negative emotional climate (i.e., sadness, anger, and fear) showed clear differences based on leadership style as the task progressed. At the beginning of the activity, no significant differences were observed between conditions (U = 3,444.50, p = 0.58). During the task, however, participants under transactional leadership reported significantly higher levels of negative emotional climate (U = 2,247.50, p < 0.001, r = 0.46), and this difference became even more pronounced by the end (U = 2,195.00, p < 0.001, r = 0.48). In both cases, the large effect sizes indicate a substantial impact of transactional leadership on increased emotional distress, whereas transformational leadership acted as a protective factor against negative affect in cooperative contexts.

When analyzing specific negative emotions, no significant differences were found in sadness at the beginning (U = 3,505.50, p = 0.64) or during the task (U = 3,290.00, p = 0.16). By the end of the task, however, participants in the transactional group reported significantly higher levels of sadness (U = 2,777.50, p = 0.002, r = 0.32), representing a medium effect size, confirming the regulatory role of transformational leadership in mitigating this emotion toward the end of the experience.

For anger, no significant differences were observed at the beginning (U = 3,446.50, p = 0.46). However, significant differences emerged during and after the task, with lower anger levels reported under transformational leadership (during: U = 2,293.00, p < 0.001, r = 0.46; end: U = 2,619.00, p < 0.001, r = 0.35). The large effect during the task and medium effect at the end indicate that transformational leadership consistently reduced frustration and tension during cooperative work.

Finally, no significant differences were found in fear between conditions at any of the three time points (T1: U = 3,446.50, p = 0.46; T2: U = 3,365.00, p = 0.21; T3: U = 3,588.50, p = 0.96), indicating that this emotion remained stable and largely unaffected by leadership style throughout the activity.

As shown in Figure 3, the emotional climate evolved differently depending on the leadership style, with transformational leadership associated with more stable or improved positive emotions over time, and transactional leadership with a progressive decline.

In order to examine the evolution of the emotional climate based on the type of leadership (within-group analysis), non-parametric Wilcoxon tests were applied to compare the three time points of the LEGO task assessment (T1 = beginning, T2 = during, T3 = end) within each experimental condition (transformational leadership and transactional leadership). The corresponding results are presented in Table 3.

In the transformational leadership condition, the analysis of positive emotional climate revealed no significant variations between T1 and T2 (Z = 1.04, p = 0.30, r = 0.11) or between T1 and T3 (Z = 1.42, p = 0.15, r = 0.15). However, a marginally significant increase was observed between T2 and T3 (Z = 2.37, p = 0.02, r = 0.25), although this result did not remain significant after Bonferroni correction (α = 0.0028). These findings indicate that the positive emotional climate remained stable at the beginning of the task and showed a slight upward trend toward the end, suggesting a potential emerging benefit of transformational leadership on students' positive emotional state. The overall magnitude of effects was small to moderate.

A more detailed analysis of individual positive emotions showed that happiness did not exhibit significant changes across the three time points (T1–T2: Z = 0.60, p = 0.95, r = 0.05; T1–T3: Z = 1.45, p = 0.14, r = 0.15). However, the change from T2 to T3 approached the threshold of significance (Z = 1.76, p = 0.07, r = 0.18). The pattern suggests a slight, non-significant upward tendency, with small effect sizes overall.

In contrast, surprise showed small to moderate increases throughout the task (T1–T3: Z = 1.91, p = 0.05, r = 0.20; T2–T3: Z = 2.05, p = 0.04, r = 0.21), indicating that the sense of novelty or engagement tended to rise over time, although this effect did not reach significance after Bonferroni adjustment.

Calmness increased significantly during the initial phase, from T1 to T2 (Z = 2.82, p = 0.005, r = 0.29), and then stabilized (T1–T3: Z = 1.65, p = 0.09, r = 0.17; T2–T3: Z = 0.61, p = 0.54, r = 0.06). Although this result did not surpass the Bonferroni threshold, the overall effects were moderate in magnitude, suggesting an early improvement in students' emotional self-regulation that was maintained throughout the activity.

Regarding the negative emotional climate, no significant differences were observed across any interval (T1–T2: Z = 0.75, p = 0.45, r = 0.08; T1–T3: Z = 0.93, p = 0.35, r = 0.10; T2–T3: Z = 0.47, p = 0.63, r = 0.05), indicating low and stable levels of negative affect throughout the task. Similarly, individual negative emotions (i.e., sadness, anger, and fear) showed no relevant variations, with small effect sizes (r ≤ 0.17).

In contrast, under the transactional leadership condition, a progressive deterioration of the positive emotional climate was observed, with robust differences across all intervals: T1–T2 (Z = 5.45, p < 0.001, r = 0.57), T1–T3 (Z = 6.35, p < 0.001, r = 0.66), and T2–T3 (Z = 3.11, p = 0.002, r = 0.32). These results indicate a sustained and substantial decrease in positive emotions throughout the cooperative task, with effects ranging from medium to large in magnitude.

The analysis of individual positive emotions showed that happiness gradually declined over time (T1–T2: Z = 2.80, p = 0.005, r = 0.29; T1–T3: Z = 3.10, p = 0.002, r = 0.32; T2–T3: Z = 1.89, p = 0.05, r = 0.20). After Bonferroni correction (α = 0.0028), only the difference between T1 and T3 remained significant. This pattern suggests a consistent but moderate decrease in happiness as the activity progressed.

Surprise also declined sharply across all intervals (T1–T2: Z = 4.48, p < 0.001, r = 0.46; T1–T3: Z = 6.02, p < 0.001 r = 0.62; T2–T3: Z = 3.57, p < 0.001, r = 0.37), showing medium-to-large effects overall and reflecting a pronounced loss of novelty and motivational engagement over time. Calmness decreased significantly from T1 to T2 (Z = 2.75, p = 0.001, r = 0.29) and from T1 to T3 (Z = 2.54, p = 0.001, r = 0.26), remaining stable thereafter (T2–T3: Z = 0.09, p = 0.92, r = 0.01). These findings point to an early reduction in emotional tranquility, of moderate magnitude, that was not recovered at the end of the task.

The negative emotional climate increased significantly from the beginning to the midpoint of the task (Z = 4.92, p < 0.001, r = 0.51) and from the beginning to the end (Z = 4.65, p < 0.001, r = 0.48), remaining stable during the final stage (T2–T3: Z = 0.31, p = 0.75, r = 0.03). The magnitude of these effects was medium to large, indicating a strong initial escalation of emotional discomfort that plateaued toward the end. Among specific negative emotions, sadness increased significantly during the task (T1–T3: Z = 3.33 p < 0.001, r = 0.35; T2–T3: Z = 2.83, p = 0.005, r = 0.29), and anger also showed notable rises (T1–T2: Z = 5.38, p < 0.001, r = 0.56; T1–T3: Z = 4.26, p < 0.001, r = 0.44). These results reveal sustained medium-to-large increases in frustration and emotional tension throughout the activity. Finally, fear slightly decreased in the final stage (T2–T3: Z = 2.31, p = 0.02, r = 0.24), suggesting a small-to-moderate effect, possibly related to a mild adaptation or habituation process toward the end of the task.

A thematic content analysis was conducted on the open-ended responses provided by participants regarding their experience during LEGO construction activities, carried out under two distinct experimental conditions (i.e., transformational leadership and transactional leadership).

With respect to the transformational leadership condition, seven categories were identified (see Table 4). The most frequent was collaboration and group cohesion (31.3%). Participants emphasized the importance of coordination, active participation, and equitable role distribution:

This cohesion was perceived as key to collective success.

Positive emotional engagement (24.8%) was reflected in expressions of enjoyment, enthusiasm, and motivation:

These responses suggest that both the activity and the transformational leadership style contributed to a safe and motivating emotional climate.

The category personal development and meaningful learning (15%) highlighted the formative value of the experience:

Task facilitation through structured guidance (11.7%) included positive evaluations of having clear instructions:

Some responses included reflective task comparisons (9.7%), noting that while instructions facilitated task completion, they also reduced creativity:

Managed emotional tension (8.4%) encompassed mentions of stress or frustration that were successfully overcome thanks to group coordination:

In contrast, under the transaccional leadership condition, eight categories were identified (see Table 5). The most prevalent was technical and organizational difficulty (22.7%), in which participants described the task as particularly complex, attributing this perception primarily to the absence of clear instructions. The lack of guidance generated a widespread sense of confusion and disorganization, as illustrated by the following responses:

Additionally, some participants noted the added challenge of not being able to visualize the internal structure of the figure, which limited their ability to plan effectively:

Others reported that they were unable to identify or locate the necessary pieces, which reduced their level of engagement in the activity:

These experiences reflect a lack of structural support, possibly exacerbated by a leadership style more focused on rule enforcement than on process facilitation.

The second most frequent category was frustration (19.1%), expressed through feelings of helplessness, emotional blockage, and dissatisfaction with personal performance:

In many cases, frustration accumulated progressively, resulting from unmet expectations of enjoyment:

Loss of control over the task was also reported as a contributing factor:

The category enjoyment (16.4%) indicated that, despite the challenges, many participants perceived the activity as positive, creative, and playful:

However, some responses reflected ambivalent emotions, where enjoyment was mixed with discomfort:

In the category recognition of cooperation (12.5%), participants acknowledged teamwork as essential to task completion:

Some responses highlighted the need to synchronize individual efforts in the absence of external guidance:

Stress (11.8%) was primarily linked to time pressure, task complexity, and lack of instructions or coordination: ‘

At times, stress was coupled with frustration:

Satisfaction (9.2%) appeared in positive evaluations of the experience, even if the final goal was not achieved:

Some responses indicated a blend of enjoyment and partial frustration:

Unequal participation (4.4%) reflected experiences of exclusion:

This suggests that an uneven distribution of roles may have negatively impacted emotional involvement and self-efficacy.

Lastly, learning and personal growth (3.9%) included narratives of improvement, skill discovery, or changes in expectations:

Finally, the qualitative evidence complements and contextualizes the quantitative findings, offering a more comprehensive understanding of how leadership styles shaped students' emotional climate during the LEGO tasks. Quantitative analyses indicated that leadership style significantly affected both the positive and negative dimensions of emotional climate (Tables 2, 3). Under transformational leadership, participants exhibited a gradual enhancement of positive emotional climate toward the end of the activity and maintained consistently low levels of negative affect. In contrast, transactional leadership was associated with a marked decline in positive emotions and a concurrent increase in negative ones.

These statistical patterns were mirrored in the qualitative data. Within the transformational condition, the most prevalent categories were collaboration and group cohesion and positive emotional engagement (Table 4). Participants described teamwork, mutual support, and shared motivation [e.g., “Thanks to teamwork, we were able to reach our goals”, (P12)], which aligned with the sustained positive emotional tone and increases in happiness and calm identified quantitatively. Mentions of structured guidance and managed emotional tension further supported the stability of affective regulation observed in the statistical analyses, suggesting that transformational leadership promoted both emotional balance and collective efficacy during the cooperative process.

Conversely, in the transactional leadership condition, the dominant qualitative categories, technical and organizational difficulty, frustration, and stress (Table 5), closely reflected the higher levels of negative emotional climate and the elevation of specific negative emotions such as sadness and anger. Narratives of confusion, pressure, and emotional overload [e.g., “It was frustrating because we didn't have any instructions”, (P45)] help explain the deterioration of affective experience under this leadership style. Although some participants mentioned enjoyment or recognition of cooperation, these appeared alongside reports of tension and disorganization, indicating that positive emotions were fragile and often undermined by the task's structural challenges.

Taken together, the integration of quantitative and qualitative evidence reveals that transformational leadership fosters a cohesive, emotionally regulated, and motivating group atmosphere, whereas transactional leadership amplifies emotional strain and disorganization, resulting in a more negative emotional climate and lower affective engagement. This convergence highlights the emotional mechanisms through which leadership influences cooperative learning experiences.

The findings of this study provide important theoretical and practical implications for the design and facilitation of cooperative learning tasks in educational contexts. Promoting transformational leadership among teachers appears to enhance the classroom's emotional climate, strengthen students' sense of belonging, foster group engagement, and improve academic performance (Day et al., 2016; Hosseingholizadeh et al., 2020). Therefore, incorporating the development of transformational leadership within both pre-service teacher education and continuous professional development programs is essential for cultivating emotionally supportive and pedagogically sustainable learning environments. In this regard, collaborative tasks using LEGO-based methodologies represent powerful tools for practicing and observing leadership behaviors, emotional regulation, and team coordination in action.

Furthermore, structuring cooperative tasks that balance clear organizational guidance with student autonomy can enhance both process effectiveness and creative self-efficacy (Jonassen, 1999). This balance is particularly important, as excessive control may inhibit initiative and active participation, whereas excessive freedom may result in dispersion or lack of focus. In this sense, promoting cooperative learning experiences that combine structured direction with genuine spaces for exploration and shared decision-making helps foster a positive emotional climate, intrinsic motivation, and shared responsibility for learning outcomes.

In addition, the integration of complementary leadership styles may offer significant advantages in contexts characterized by overlapping social, economic, and political challenges, allowing for more flexible, inclusive, and resilient educational responses (Du Plessis and Keyter, 2020). These authors use the term complementary leadership to describe the strategic combination of different leadership approaches, such as transformational, transactional, or distributed leadership, applied flexibly according to contextual needs. From this perspective, leadership styles are not mutually exclusive but rather mutually reinforcing, leveraging the strengths of each to compensate for their limitations.

For instance, while transformational leadership promotes vision, motivation, and innovation, transactional leadership provides the structure, norms, and coordination mechanisms necessary to ensure organizational effectiveness. Likewise, distributed or participative leadership strengthens collaboration and shared responsibility among teaching staff. Integrating these styles in a complementary way enables educational leaders to maintain institutional stability and clarity of objectives without sacrificing teacher creativity, autonomy, or emotional engagement.

In this regard, Printy (2014) also highlights that many schools fail to reach their full potential due to the lack of articulation between different leadership styles (e.g., transformational and transactional). Her integrated leadership school model proposes that effective leadership combines structured direction with collaborative participation, fostering environments where strategic planning and pedagogical innovation coexist productively.

Taken together, these implications underscore the need to promote integrated and adaptive educational leadership approaches that not only foster collaboration and emotional wellbeing but also strengthen adaptability, creativity, and innovation in response to the complex systemic demands of contemporary education.

This study presents several limitations that should be considered when interpreting the findings. First, the sample consisted of a limited number of participants from a single educational context, university students from one institution, which restricts the generalizability of the results to other educational levels and settings. The high proportion of female participants reflects the typical composition of teacher education programs in Spain, providing ecological validity to the sample. Although all participants came from the same university, this ensured cultural and institutional homogeneity, which was advantageous for maintaining experimental control. Nevertheless, the generalization of the findings should be interpreted with caution, and future studies are encouraged to replicate the design in more diverse contexts (e.g., different educational levels, organizations, cultures, or with greater gender diversity).

Second, although the fixed order of leadership conditions was methodologically justified, it may have introduced emotional carryover effects between conditions. Participants' emotional responses or engagement during the second (i.e., transformational) task could have been influenced by their prior experience in the transactional condition. To minimize potential sequence-related biases, future research should consider counterbalancing the order of conditions, increasing the temporal interval between tasks, or assessing participants' emotional baselines prior to each condition.

Third, self-report questionnaires present inherent limitations arising from the tendency to respond in socially desirable ways or from biases related to the subjective interpretation of processes, such as participants' perceptions of their own performance. To mitigate these limitations, task performance was evaluated through a dual-check procedure (i.e., combining participants' self-reports with researchers' real-time observations) in order to enhance objectivity and consistency. In both cases, this assessment was based on a single item evaluating three performance levels (i.e., complete, partial, or incomplete construction). Future studies could benefit from incorporating more detailed observer-based evaluations, such as multi-item performance rubrics assessing a broader range of aspects or video analyses of group interactions, which would allow for more precise insights into task execution and collaborative dynamics. The integration of such complementary methods would strengthen both the reliability and validity of performance assessments in cooperative learning contexts.

Furthermore, the use of single-item measures to assess task performance and time perception may have reduced measurement reliability. However, since these constructs were specific, narrow, and directly observable, the use of single-item indicators was considered methodologically appropriate in this context (Allen et al., 2022). Even so, future research could employ multi-item scales to increase measurement precision and allow for more robust reliability analyses.

Fourth, the assessment of emotional climate through self-reports may have been influenced by social desirability biases or individual differences in emotional expression. To advance the understanding of group affective processes, it would be advisable to complement self-reports with physiological or behavioral indicators of emotion, such as heart rate, facial expression, or voice tone, to obtain a richer and more objective understanding of group emotional dynamics.

In addition, future research should expand samples to include participants from different educational levels, sociocultural contexts, and with greater gender diversity, in order to strengthen the external validity of the findings. From a methodological perspective, future studies could examine the mediating role of emotional climate between leadership style and group performance, as well as evaluate the impact of transformational leadership on other relevant outcomes, such as self-efficacy, deep learning, and group cohesion.

Finally, although this study focused on transformational and transactional leadership, it would be valuable for future research to explore hybrid or distributed leadership approaches, which could provide a more nuanced and valid understanding of how different leadership strategies are combined in real collaborative contexts. Likewise, future studies could employ longitudinal designs to better understand how leadership behaviors evolve over time and how they influence the sustainability of group performance and emotional climate. Furthermore, incorporating experimental settings that use construction block activities (e.g., LEGO) or other interactive simulations could offer a methodologically valid means of observing leadership processes as they emerge, enabling researchers to link behavioral indicators, affective dynamics, and performance outcomes within authentic group interactions.

The present study examined the differential effects of transformational and transactional leadership styles on emotional climate, performance in a cooperative task, and time perception, as well as the evolution of emotional climate throughout a collaborative LEGO-based activity in a university educational context.

The results, obtained through a mixed-methods approach, provide a robust understanding of the regulatory role played by different leadership styles in shaping group emotional dynamics and the effectiveness of structured cooperative experiences. Overall, the findings demonstrate that transformational leadership is associated with greater emotional stability and a progressive increase in positive emotional climate throughout the task, along with a sustained reduction in negative emotions. This leadership style fostered working environments marked by greater emotional engagement, efficient coordination, positive perception of available time, and improved task performance.

In contrast, transactional leadership exhibited the opposite pattern, a gradual decline in positive emotional climate and an increase in frustration, stress, and emotional discomfort, especially toward the end of the activity. These results suggest that a leadership approach focused exclusively on rule enforcement and behavioral supervision may undermine the emotional quality of group interactions, negatively affecting both the experience and the outcomes of the task.

In this regard, the data underscore the pedagogical value of transformational leadership as a key tool for fostering emotionally safe, motivating, and cohesive climates that support both individual wellbeing and collective effectiveness. Furthermore, the findings highlight that cooperative tasks require not only structural organization but also emotional support, a sense of belonging, and a culture of mutual support to reach their full educational potential.

Finally, this study contributes to the educational leadership literature by empirically demonstrating, through methodological triangulation, how leadership styles influence not only academic performance but also the emotional experiences of students. Promoting transformational leadership practices in educational settings may represent a strategic pathway for enhancing both group effectiveness and students' socio-emotional wellbeing.