Cognitive fatigue induced by spaced and massed practice: effects on idiom accuracy across L2 proficiency levels

2.1 Idioms

Idioms are expressions whose meanings cannot be inferred from the meanings of the individual words, making them a particularly challenging aspect of L2 learning. Examples like kick the bucket or break the ice require learners to understand not only the individual words but also cultural nuances and metaphorical meanings (Liontas 2017). The ability to comprehend and use idiomatic expressions is often considered a marker of advanced L2 proficiency (Liontas 2017). However, learners must overcome significant cognitive challenges to comprehend idioms effectively, as idiom comprehension involves complex cognitive processes. L2 learners often struggle with these non-literal meanings, which might interfere with their fluency and overall communication in the target language. Previous studies have highlighted that idiom comprehension is particularly taxing for learners due to the need for both conceptual and linguistic processing (Cacciari et al. 2018). Cognitive load theory, similarly, suggests that complex tasks like idiom comprehension require learners to manage cognitive resources effectively, but the demands of such tasks can lead to cognitive fatigue (Sweller et al. 2011a, 2011b). Cacciari et al. (2018) emphasized the relationship between cognitive functions and idiom comprehension, suggesting that successful idiom processing may require advanced cognitive skills that develop later in language acquisition (Sprenger et al. 2019). This connection implies that idiom comprehension is not merely a function of exposure but also reflects underlying cognitive capacities that vary among learners. Such individual differences can affect the stability and reliability of idiom knowledge, particularly among younger learners who may not have developed a robust idiomatic vocabulary (Sprenger et al. 2019).

In general, the acquisition of idiomatic expressions in L2 is intricately tied to cognitive processes similar to those governing other linguistic behaviors and learning paradigms. Fundamental principles of long-term memory formation, such as noticing, encoding, storage, and retrieval, play a crucial role in learning multi-word figurative expressions, paralleling the learning of individual word meanings (Vasiljevic 2015). Research has indicated that idiom comprehension and acquisition depend significantly on individual differences, including cognitive abilities and personality traits, which can influence both the learning process and the eventual mastery of idioms (Sprenger et al. 2019).

2.2 Cognitive fatigue and its dimensions

Cognitive fatigue is a multifaceted construct that affects various domains, from workplace productivity to learning (Karim et al. 2024). Researchers have identified five core dimensions of cognitive fatigue: cognitive, emotional, physical, behavioral, and motivational fatigue. These dimensions capture the complex and interconnected nature of fatigue and its impact across different domains of functioning. Cognitive fatigue is characterized by a decline in mental efficiency, attention, and decision-making abilities after prolonged mental exertion. Neurocognitive models suggest that cognitive fatigue arises from resource depletion in the prefrontal cortex, which governs higher-order cognitive functions such as working memory and executive control (Baumeister and Vohs 2016). Empirical studies have also shown that sustained cognitive tasks, such as problem-solving or learning, lead to decreased accuracy and slower response times (Ackerman 2011). This dimension is particularly relevant in academic and professional contexts where mental stamina is critical. In language learning, cognitive fatigue can impair processing, retention, and recall, particularly in demanding tasks like idiom learning or complex grammar exercises.

Emotional fatigue refers to feelings of emotional depletion and reduced emotional resilience due to prolonged exposure to stress or emotionally taxing situations. Symptoms include irritability, apathy, and a reduced capacity to empathize with others (Maslach and Leiter 2016). Emotional fatigue is linked to burnout in high-stress environments, such as teaching, caregiving, or jobs requiring extensive interpersonal interactions. In SLA, emotional fatigue may affect learners’ attitudes and motivation, particularly when they face repeated setbacks or challenges in acquiring a new language.

Physical fatigue involves a decline in physical energy and endurance due to sustained physical effort or inadequate recovery. While primarily related to bodily exertion, physical fatigue often interacts with cognitive and emotional fatigue, as seen in holistic tasks like teaching or sports (Kroemer and Grandjean 1997). Physiological markers include increased heart rate, muscular soreness, and reduced physical stamina. It can indirectly impair cognitive tasks by diminishing overall energy levels (Marcora et al. 2009). In SLA, physical fatigue may arise in immersive language programs where learners participate in extended classroom activities and extracurricular engagement.

Behavioral fatigue manifests as reduced initiative, decreased task engagement, and an increased likelihood of errors in performance. Behavioral fatigue often presents as procrastination, withdrawal from demanding tasks, or avoidance behaviors (Ackerman and Kanfer 2009). Behavioral changes may exacerbate fatigue, creating a cycle of reduced effort and poor outcomes. In SLA, this dimension is evident when learners disengage from practice, fail to complete assignments, or avoid communicative opportunities due to perceived effort.

Motivational fatigue is a decline in the drive or willingness to engage in tasks, often stemming from repeated failures, monotonous routines, or perceived lack of progress. Generally, motivational traits significantly influence L2 proficiency and, consequently, idiom acquisition. Many studies, such as Dunn and Iwaniec (2022) and Papi and Khajavy (2021), have highlighted the noticeable relationships between various motivational constructs and L2 learning across L2 proficiency levels. Accordingly, learners experiencing motivational fatigue often report feelings of boredom, frustration, or defeat. Motivational fatigue can lead to dropout or reduced practice, particularly when learners do not perceive clear progress or relevance in their studies (Schunk and Zimmerman 2007).

Of note, the five dimensions of cognitive fatigue are not isolated entities but rather deeply interconnected and overlapping. These dimensions form a dynamic system where the effects in one domain can cascade into others, amplifying the overall experience of fatigue. For instance, cognitive fatigue from prolonged mental effort can heighten emotional fatigue, leading to frustration or irritability. Similarly, physical fatigue can diminish cognitive efficiency and motivation, creating a cycle where learners struggle to focus, lose emotional resilience, and withdraw from tasks (behavioral fatigue). This interplay is particularly evident in high-demand settings like language acquisition, where sustained effort is required across all these domains. Recognizing this interconnectedness is crucial for designing effective interventions. Holistic approaches that address multiple dimensions – such as combining physical rest, cognitive breaks, emotional support, and motivational incentives – are more likely to mitigate fatigue and promote sustained performance and well-being.

In summary, cognitive fatigue should not be viewed as a singular issue but as a complex, multi-dimensional phenomenon where the boundaries between its components blur, influencing each other in meaningful ways.

2.3 Spaced versus massed practice and cognitive fatigue

Spaced practice, also known as distributed practice, involves spreading learning sessions over time, allowing for periods of rest between learning episodes. The spacing effect, first identified by Ebbinghaus (1885), indicates that learners retain information better when study sessions are spaced out rather than crammed into a short period. Studies in the domain of SLA have found that spaced practice improves retention and recall, particularly for vocabulary and grammar (Bahrick et al. 1993). In contrast, massed practice, or cramming, involves concentrated practice over a short period, often without sufficient breaks or rest. While massed practice may lead to short-term improvements in performance (e.g., recall during a single session), it can induce cognitive fatigue over time. Research by Chen and Kalyuga (2020) found that massed practice leads to higher levels of cognitive fatigue, as learners must sustain attention and effort over extended periods. The fatigue induced by massed practice is thought to decrease cognitive resources, making it harder for learners to process new information effectively (Chen and Kalyuga 2020). Despite its drawbacks, massed practice has been shown to provide temporary benefits for tasks that require quick, focused performance. However, these benefits often dissipate over time, and learners may experience a performance decline (Dunlosky et al. 2013).

The concepts of cognitive fatigue, particularly in the context of spaced versus massed practice, are essential to understanding idiom learning. Spaced practice (i.e., distributing learning over time) has been shown to facilitate deeper processing and better long-term retention of language material, including idioms, compared to massed practice, which tends to lead to cognitive fatigue (Vasiljevic 2015). More clearly, by spacing out learning sessions, learners are given time to consolidate information, making it easier to retrieve and apply in future contexts (Cepeda et al. 2006). This approach also alleviates cognitive fatigue by preventing prolonged periods of intensive cognitive effort, which can lead to exhaustion and performance declines.

Overall, the effects of spaced and massed practice on L2 learning have been well-documented. For example, Bahrick et al. (1993) demonstrated that spaced review of vocabulary words led to better retention in the long term compared to massed review. Similarly, studies on grammar learning have found that spaced practice enhances both short-term and long-term performance (Miles 2014). In exploring the effects of task repetition on fluency in L2 learning, Lambert et al. (2017) conducted a detailed analysis of Japanese EFL learners engaged in repeated speaking tasks. Their study found that fluency improvements were significant up to the fourth or fifth performance, highlighting the potential benefits of structured practice on idiom usage and production (Suzuki and Hanzawa 2022). These findings align with the understanding that repeated exposure and practice can alleviate cognitive fatigue and enhance fluency in language learners. However, less attention has been paid to how these practices influence cognitive fatigue during tasks that require substantial cognitive effort, such as idiom comprehension.

2.4 Cognitive fatigue and L2 learning

Research on cognitive fatigue in education suggests that learners often experience a decline in task performance, attention, and motivation when fatigued. Studies by de Lange et al. (2004) and Kok (2022) showed that cognitive fatigue impairs performance in tasks requiring sustained attention and working memory. In the context of L2 learning, cognitive fatigue may hinder learners’ ability to recall and apply idiomatic expressions, as the retrieval of non-literal meanings requires substantial cognitive effort (Schmitt 2010).

L2 proficiency, as an important component of SLA, plays a crucial role in how learners manage cognitive fatigue during language tasks. Research suggests that learners at different L2 proficiency levels experience cognitive load differently, with higher proficiency learners often better equipped to handle complex tasks (Sweller et al. 2011a, 2011b). For example, high proficiency learners may have more efficient mental representations of vocabulary and grammar, which reduces the cognitive effort required for tasks like idiom comprehension (Fitzpatrick and Barfield 2009). Low proficiency learners, on the other hand, are likely to experience greater cognitive fatigue because they lack the mental frameworks and processing strategies needed to handle the complexity of idioms. As a result, they may struggle to recall and apply idiomatic expressions effectively, especially under conditions of high cognitive load (Macis et al. 2021). In contrast, more advanced learners may have developed strategies to manage cognitive fatigue, such as chunking or relying on context to interpret idiomatic meanings. Studies have also shown that proficiency level affects both task performance and the ability to manage cognitive fatigue. For example, research by Ellis (2008) suggests that higher proficiency learners can handle more complex language tasks, such as idiom comprehension, with less cognitive strain. This implies that lower proficiency learners may experience more fatigue, leading to declines in accuracy and recall when engaging in tasks like idiom comprehension.

2.5 Measuring cognitive fatigue in SLA

Two main types of methods to measure cognitive fatigue, especially in educational contexts, are self-report and performance-based methods. Self-report measures are widely used to assess cognitive fatigue because they provide direct insights into learners’ subjective experiences. Among these, the VAS is commonly employed in educational studies (Nguyen and Fabrigar 2018). It allows learners to rate their perceived fatigue on a continuous scale, providing a simple yet reliable tool to capture fluctuations in mental effort during learning tasks. Previous studies in SLA have used VAS to track cognitive fatigue before and after learning tasks, with findings indicating that learners report increased fatigue after intense study sessions (Lee and Park 2023).

Performance-based measures, such as accuracy tracking, can also provide an objective way to assess cognitive fatigue. Studies have shown that cognitive fatigue often manifests in declining performance, such as slower reaction times or decreased accuracy (Eysenck et al. 2007). In the context of idiom comprehension, accuracy tracking through tasks like free recall can offer valuable insights into how cognitive fatigue impacts learners’ ability to recall and apply idiomatic expressions. Research by Kok (2022) showed that cognitive fatigue is associated with lower accuracy and longer response times in cognitive tasks, making it an effective indicator of mental exhaustion.

4.1 Research design

Quasi-experimental between-participant research design was used for the study as it offers several merits for this study. By assigning participants to different proficiency levels and practice schedule conditions (spaced vs. massed), this design helps isolate the effects of each independent variables (i.e., practice conditions and L2 proficiency level) on dependent variables (i.e., cognitive fatigue and idiom accuracy) without the risk of carryover effects that might occur in a within-subjects design. It allows for a clearer comparison of how different proficiency levels interact with varying practice schedules, providing distinct group-level insights. This design also reflects real-world scenarios where learners typically engage in either spaced or massed practice, rather than both, ensuring ecological validity.

4.2 Participants

The participants were 92 young adult English as a Foreign Language (EFL) learners with mean age of 21 years old belonging to six intact classes from three different L2 proficiency levels (two classes per L2 proficiency): low (n = 33), mid (n = 31), and high (n = 28). They were all Iranian, with Persian/Farsi as their first language. Their exposure to the target language, English, was primarily limited to academic settings, as English is not commonly used in daily situations in the country. The participants’ L2 proficiency level was specified by the placement test of the language school. More specifically, based on the CEFR standard, the low L2 proficiency learners were at A1, mid L2 proficiency learners were at B1 level, and the high L2 proficiency learners were at C1 level. The Updated Vocabulary Levels Test (UVLT) results also indicated that the low, mid, and high L2 proficiency groups had lexical knowledge of the first 2,000 words, 3,000 words, and 4,000 words, respectively.

All participants were informed about the study’s purpose, procedures, and their rights before participation. Informed consent was obtained in writing from all participants. As an incentive, participants received graded readers upon completion of the study.

4.3 Materials

Considering the class time constraints, 20 opaque idioms were selected for the study from which 12 idioms were the main target idioms, and eight ones were control items. The idioms all shared some features. First, they were restricted to three to four words to uphold consistency in complexity and guarantee that they remained cognitively manageable for learners across all proficiency levels. Second, they did not include any pronouns to prevent learners from needing to significantly alter the idiomatic expressions during the learning process (e.g., pull my leg, pull his leg, etc.). Third, the individual constituents of them were within the most frequent 3,000 word families (based on COCA), so that the participants knew the meanings of the individual words but not the meaning of the whole idiom. The identical sets of English idioms were employed across all three proficiency groups of participants, enabling a direct comparison of the impact of learner proficiency on idiom acquisition. Control items were chosen based on the same criteria applied to the target idioms. This alignment between the control and target items facilitated the researcher’s ability to isolate the effects of independent variables while reducing the influence of potential confounding variables.

To prepare the idioms, first, an idiom pool consisting of 40 idioms was compiled. Then, a norming task was administered online with 10 native speakers of English to check the transparency of the idioms. The native speakers were required to rate the transparency of the idioms (i.e., the degree to which the meaning of an idiom can be understood from the meaning of its individual components) on a rating scale from 1 to 7 (1 and 2 indicated high transparency while 6 and 7 indicated low transparency). In the next stage, only the idioms with mean transparency of four and above were kept (26 idioms). This threshold was chosen because it signifies the midpoint of the scale, differentiating between idioms that possess minimal to no compositional meaning and those that provide a certain level of transparency. Moreover, in the course of initial data gathering and analysis of responses from native speakers, it was observed that idioms rated above 4 were uniformly assessed as lacking a significant connection between their elements and their figurative interpretation.

Following this, the list was reviewed by the participants’ classroom teachers, who were asked to provide feedback on the idioms’ familiarity and potential recognition by learners at different proficiency levels. Based on their professional judgment, the teachers identified three idioms that might have been recognizable or guessable due to a similar idiom in their first language. These idioms were removed to ensure that all selected idioms were genuinely unknown to the participants across all proficiency levels.

The remaining idioms were then given to 15 advanced-level (IELTS level) L2 learners in another online norming task to ensure the meaning of the idioms was unknown to them. The main reason for selecting advanced-level L2 learners for the norming task was that if they did not know the idioms, it was highly likely that intermediate and low-level learners did not know them either. This stage also resulted in omission of three other idioms.

4.4 Procedure

4.5 Data analysis

For the first research question, asking how spacing schedule affects overall cognitive fatigue across L2 proficiency levels, two-way ANOVA was conducted in which spacing schedule and L2 proficiency were taken as independent variables and overall cognitive fatigue as the dependent variable. This statistical approach could help to determine if there were significant differences in cognitive fatigue across L2 proficiency groups and practice conditions. One important point to highlight is that given that Cohen’s d is particularly useful for comparing two groups and Partial η² is more appropriate for factors with multiple levels, in case of obtaining significant results, both indexes were calculated to be able to make more fine-tuned interpretations. For the second research question, asking about how L2 proficiency levels differ in terms of cognitive fatigue dimensions under spacing schedules, to compare cognitive fatigue levels (cognitive fatigue scale scores) across the practice schedule (spaced vs. massed) and proficiency groups (low, mid, high), MANOVA was used in which the five dimensions of cognitive fatigue were taken as dependent variables and L2 proficiency and spacing schedules as independent variables. This statistical technique could determine if there were statistically significant differences in cognitive fatigue across the L2 proficiency groups and practice types while controlling for type I error rates. Finally, to examine the third research question examining the relationship between cognitive fatigue and idiom comprehension across L2 proficiency levels, multiple regression analysis was employed as it could assess how well cognitive fatigue predicts idiom comprehension accuracy while controlling for L2 proficiency levels.

5.1 Spacing schedule and overall cognitive fatigue across L2 proficiency levels

Findings pertinent to the first research question underscored a clear gradient in the differential effects of practice type on cognitive fatigue, with the impact of massed practice intensifying as proficiency levels increase (Figure 1). To be more detailed, in the low L2 proficiency, massed practice resulted in slightly higher cognitive fatigue compared to spaced practice, indicating a marginal difference in fatigue levels between the two conditions. In contrast, mid L2 proficiency learners exhibited a notable disparity, with fatigue levels significantly higher under massed practice than under spaced practice. The trend was even more pronounced in high L2 proficiency learners, where massed practice caused markedly elevated cognitive fatigue compared to spaced practice.

Figure 1:

Mean overall fatigue across practice conditions and L2 proficiency levels.

The two-way ANOVA results (Table 1) indicate significant main effects of both L2 proficiency (F(2,86) = 74.408, p < 0.001) and spacing condition (F(1,86) = 832.744, p < 0.001), as well as a significant interaction between them (F(2,86) = 183.051, p < 0.001).

Statistical analyses further revealed substantial effect sizes across all factors examined (Table 2). The spacing condition emerged as the most influential factor, demonstrating an exceptionally large effect size (partial η² = 0.906, Cohen’s d = 2.468), accounting for 90.6 % of the variance in fatigue scores when controlling for other variables. This effect is particularly evident in the mean differences between massed (M = 18.32, SD = 2.65) and spaced practice conditions (M = 11.36, SD = 2.99).

L2 proficiency level also demonstrated a considerable effect (partial η² = 0.634), explaining 63.4 % of the variance in fatigue scores. Moreover, a robust interaction effect between spacing condition and L2 proficiency was observed (partial η² = 0.810), indicating that 81 % of the variance could be attributed to the combined influence of these factors. Following Cohen’s (1988) guidelines for effect size interpretation, where η² values of 0.01, 0.06, and 0.14 represent small, medium, and large effects respectively, all observed effects in this study can be classified as very large.

The magnitude of these effects, particularly the spacing condition’s impact (d = 2.468), substantially exceeds Cohen’s conventional threshold for a large effect (d = 0.80). This finding underscores the practical significance of implementing spaced practice in L2 learning contexts. The robust interaction effect further suggests that the impact of spacing conditions varies meaningfully across proficiency levels, highlighting the importance of considering both factors in pedagogical design. These findings provide strong support for the implementation of spaced practice in L2 learning environments.

5.2 L2 proficiency differences and cognitive fatigue dimensions under spacing schedules

A MANOVA was conducted to examine the effects of L2 Proficiency, Spacing Condition, and their interaction on the combined dimensions of cognitive fatigue (Table 3). Several results were observed. First, A significant multivariate effect was observed for L2 proficiency (Pillai’s Trace = 1.513, F(10,166) = 51.533, p < 0.001) indicating that L2 proficiency levels significantly influenced the combined dimensions of cognitive fatigue. Second, A significant multivariate effect was also found for spacing condition (Pillai’s Trace = 0.919, F(5,82) = 185.288, p < 0.001) suggesting that the practice condition (spaced vs. massed) had a substantial impact on cognitive fatigue dimensions. Third, the interaction between L2 Proficiency and spacing condition was significant (Pillai’s Trace = 1.349, F(10,166) = 34.425, p < 0.001) indicating that the effect of spacing condition on cognitive fatigue dimensions varies across L2 proficiency levels.

To further understand the magnitude of these effects, partial eta squared values were calculated for each dimension of cognitive fatigue (Table 4). The motivational dimension showed the strongest effect of spacing condition (η² = 0.838, η² = 0.838), while the emotional dimension exhibited the largest effect of L2 Proficiency (η² = 0.748, η² = 0.748). The physical dimension demonstrated the most substantial interaction effect (η² = 0.669, η² = 0.669), highlighting the interplay between L2 proficiency and spacing condition in influencing this dimension. These findings collectively suggest that both L2 proficiency and spacing condition, as well as their interaction, significantly influence various dimensions of cognitive fatigue, with varying magnitudes across dimensions.

To be more specific, Figure 2 clearly underscores the nuanced interplay between L2 proficiency and spacing conditions, with distinct patterns emerging across different dimensions of cognitive fatigue. In the cognitive dimension, spaced learning consistently results in lower fatigue scores compared to massed learning, with the largest differences observed in mid-proficiency learners. The emotional dimension shows a similar trend, though the gap between spacing conditions narrows for high-proficiency learners. The physical dimension reveals substantial variability, with mid-proficiency learners showing the most pronounced differences between spacing conditions. The behavioral dimension demonstrates relatively stable patterns, with spaced learning generally leading to lower fatigue scores across all proficiency levels. Finally, the motivational dimension exhibits a unique trend where the impact of spacing conditions diminishes as proficiency increases, with high-proficiency learners showing minimal differences between massed and spaced conditions.

Figure 2:

Cognitive fatigue dimensions across L2 proficiency levels and practice conditions.

5.3 Relationship between cognitive fatigue and idiom comprehension across L2 proficiency

Lastly, data associated with the third research question (i.e., the interrelationship of overall cognitive fatigue and idiom comprehension) revealed a consistent negative relationship between cognitive fatigue and idiom task performance across all L2 proficiency levels, indicating that higher fatigue is generally associated with lower task accuracy. For high L2 proficiency learners, spaced practice yielded higher idiom task accuracy (M = 9.86) compared to massed practice (M = 7.93), suggesting the benefits of distributed learning for this group. Similarly, mid L2 proficiency learners demonstrated improved accuracy under spaced practice (M = 8.67) relative to massed practice (M = 6.00), reinforcing the efficacy of spaced intervals for enhancing learning outcomes. Interestingly, for low L2 proficiency learners, the trend was reversed, with slightly better idiom scores observed under massed practice (M = 6.88) compared to spaced practice (M = 5.38).

More specifically, the multiple regression analysis (Table 5) examining the relationship between cognitive fatigue dimensions and idiom comprehension accuracy yielded a statistically significant model (F(5, 86) = 32.82, p < 0.001), with an adjusted R² of 0.636, indicating that approximately 63.6 % of the variance in idiom comprehension accuracy is explained by the five dimensions of cognitive fatigue. Three dimensions emerged as significant predictors: Cognitive fatigue showed a positive relationship (β = 0.634, p < 0.001), emotional fatigue showed a negative relationship (β = −0.240, p < 0.05), and physical fatigue showed a strong negative relationship (β = −1.048, p < 0.001). Behavioral and Motivational dimensions did not significantly predict idiom comprehension accuracy (p > 0.05).

The Variance Inflation Factors (VIF) range from 1.37 to 2.54, indicating acceptable levels of multicollinearity (all values < 3). The correlation matrix reveals particularly strong negative correlations between idiom task scores and physical fatigue (r = −0.718) and behavioral fatigue (r = −0.575). These findings suggest that while increased cognitive engagement may facilitate idiom comprehension, physical and emotional fatigue may impair performance. The strong negative correlation with physical fatigue particularly highlights the importance of managing physical fatigue in L2 learning contexts.