Working Memory Dimensions and Their Interactions: A Structural Equation Analysis in Saudi Higher Education

2.1 Working Memory: An Overview

Working memory, a cognitive system essential for human functioning, temporarily stores and manipulates information to support complex tasks such as reasoning, problem-solving, and language comprehension (Logie, Camos, & Cowan, 2020). Unlike short-term memory, which passively holds information, working memory integrates external stimuli with long-term memory, thus enabling adaptive and goal-directed actions (Cowan, 2020). Its pivotal role in academic performance, language processing, and executive functioning demonstrates its theoretical and practical significance (Ji & Guo, 2023; Li, 2023). Despite its importance, questions remain regarding how its dynamic processes adapt to varying cognitive demands, leaving critical gaps in the literature.

Neuroscientific advancements have reshaped our understanding of the neural mechanisms of working memory, moving beyond traditional models that reinforce sustained neural activity in the prefrontal cortex. Contemporary research shows ‘activity-silent’ processes involving transient neural activations and network-level interactions as central to maintaining flexibility during task demands (Curtis & Sprague, 2021; Rose, 2020). The dynamic processing model captures this adaptability, offering insights into how working memory balances transient and sustained processes (Murphy, Bertolero, Papadopoulos, Lydon-Staley, & Bassett, 2020; Myers, 2022). However, the relationship between these mechanisms, particularly in high-pressure academic contexts, remains unclear.

Baddeley’s multicomponent model remains influential for its practical utility in distinguishing the key components of working memory: the phonological loop, visuospatial sketchpad, central executive, and episodic buffer (Logie et al., 2020). In particular, the episodic buffer enhances the explanatory power of the model by integrating multimodal information with long-term memory, making it indispensable for tasks requiring cross-modal coordination (Spencer, 2020). However, its limited ability to account for individual differences and dynamic neural processes has led to calls for integration of this model with contemporary theories. Furthermore, working memory capacity is dynamic and develops throughout childhood and adolescence, driven by neural maturation. Strengthened white matter pathways and improved connectivity between brain regions enhance storage, attention, and executive function during these stages (Forsberg, Blume, & Cowan, 2021). Metacognitive accuracy, which refines an individual’s ability to monitor and regulate working memory processes, also significantly improves, contributing to greater cognitive efficiency (Forsberg et al., 2021).

Individual differences in working memory capacity arise from variations in multimodal network dynamics. Robust neural connectivity enhances functional efficiency, whereas structural or functional disconnections contribute to deficits (Hamza & Helal, 2021). These differences influence academic performance and reinforce the need for tailored interventions. Computational models offer promising strategies for enhancing working memory efficiency, particularly in at-risk populations (Cowan, 2020; Poole, Phillips, Stewart, Harris, & Lah, 2021).

In educational contexts, working memory strongly correlates with academic outcomes, including GPA, and supports critical skills, such as mathematical problem-solving, reading comprehension, and creative reasoning (Amzil, 2022; Ji & Guo, 2023). Recent research in Saudi higher education further shows that working memory efficiency interacts with learners’ motivation and self-efficacy through self-regulated learning mechanisms, shaping overall academic performance (Almayez, Al-Khresheh, Al-Qadri, Alkhateeb, & Alomaim, 2025). Task framing significantly influences working memory performance: opportunity-focused framing fosters engagement and reduces cognitive load, whereas risk-focused framing heightens threat appraisal and impairs performance (Chen & Qu, 2021). These findings reveal the importance of contextual factors in optimizing the efficiency of working memory.

Gender differences also play a role in working memory performance. Research indicates that female students frequently perform better than male students in memory recall tasks, such as word and picture retention, potentially reflecting evolutionary adaptations (Al-Baghdadi & Ashmawy, 2021). Emotional factors, including stress and depressive cognition, further modulate working memory capacity. These negative influences can be mitigated through cognitive flexibility and emotion regulation strategies, bringing to the forefront the interconnected nature of cognitive and emotional resilience (Diotaiuti et al., 2024).

2.2 Storage Dimension of Working Memory

The storage dimension of working memory is fundamental to cognitive functioning, enabling the temporary retention and manipulation of information essential for academic and everyday tasks (Shen et al., 2021). Unlike long-term memory, which encodes information for extended periods, storage memory holds data transiently, often for only a few seconds, unless actively rehearsed (Baddeley, 2020). Despite its brief capacity, storage memory is fundamental in tasks such as reading comprehension and mathematical problem-solving, where rapid information integration and processing are critical (Lenartowicz et al., 2021).

The contribution of storage memory to academic performance is well documented. Verbal storage supports retaining linguistic elements, allowing students to maintain sentence coherence and extract meaning across paragraphs (Shin, 2020). Likewise, spatial storage facilitates mathematical reasoning by enabling mental manipulation of abstract numerical and geometric relationships (Li, Sun, Zhou, & Wang, 2023). Together, these components form the foundation of cognitive strategies for navigating complex academic tasks (Luchini, Wang, Kenett, & Beaty, 2024). However, the impacts of verbal and spatial storage vary across academic activities. Verbal storage is more effective for simple comprehension tasks, but less efficient under high cognitive load, where additional executive functions are required (Teng & Zhang, 2023). In contrast, spatial storage predominates in tasks such as counting, where visual organization and representation are paramount (Ambalova, 2021). This task-dependent variability points out the adaptability of storage memory in meeting diverse cognitive demands (Khan & Ghazala, 2023).

A defining characteristic of storage memory is its ability to adapt its organization and capacity in response to task requirements (Lorenc & Sreenivasan, 2021). For instance, Cao and Deouell (2023) proved that visual working memory adjusts its storage format, either separating features such as colour and shape or combining them into conjoined objects, depending on the task. Such flexibility ensures the prioritization of relevant information and enhances efficiency in demanding scenarios (Shen et al., 2021). Storage memory also interacts dynamically with processing functions, reallocating resources to balance immediate processing needs with retention. Barrouillet et al. (2024a) described this tradeoff as a sophisticated mechanism for optimizing performance under high cognitive demands. Thyer et al. (2022) also identified a content-independent pointer system that reallocates resources within storage memory, ensuring adaptability across diverse tasks. These findings revealed the active role of storage memory in cognitive functioning.

Insights into storage memory provide promising directions for educational interventions aimed at improving academic outcomes. Targeted exercises that enhance verbal storage, such as exposure to syntactically complex sentence structures, can strengthen reading comprehension by improving retention of linguistic patterns (Westby, 2020). Similarly, visualization tasks that develop spatial storage, such as mentally manipulating geometric or numerical data, can enhance mathematical reasoning (Jamaludin, 2022). Strategies such as chunking, in which information is grouped into meaningful units, have also proven effective in improving storage capacity. Suppawittaya and Yasri (2021) demonstrated that chunking significantly enhances information retention in high school students, demonstrating its practical value. Tailoring these interventions to specific academic demands ensures that students maximize the potential of their storage memory (Ambalova, 2021).

2.3 Attention Dimension of Working Memory

The attention dimension of the working memory is fundamental to cognitive efficiency. It enables individuals to focus on relevant stimuli while suppressing distractions (Draheim, Pak, Draheim, & Engle, 2021; Rustichini, Domenech, Civai, & DeYoung, 2023). By regulating what enters and remains in the working memory system, attention ensures that cognitive resources are directed toward priority tasks (Kotyusov et al., 2023). This capacity is particularly crucial in complex environments with competing demands, such as academic settings, where effective resource allocation significantly influences success (Van Ede & Nobre, 2023).

Attention control is widely recognized as a primary determinant of working memory capacity. Research indicates that individuals with strong attention control excel their focus on task-relevant information, allowing them to manage distractions and adapt to changing cognitive demands (Draheim et al., 2021). This distinguishes attention control from selective attention, which plays a minor role in determining working memory efficiency (Kotyusov et al., 2023; Van Moorselaar & Slagter, 2020). Mindfulness studies have further asserted the relationship between attention and working memory. Trait mindfulness, which supports sustained attentional focus, enhances working memory capacity, particularly under stressful conditions. Mindfulness-based attention mechanisms provide a protective buffer for cognitive performance by mitigating the disruptive effects of high-pressure situations (Li, Yang, Zhang, Xu, & Cai, 2021). Developmental research adds that attentional cues improve recall probability in children, although the quality of memory representations may remain unchanged, indicating the evolving role of attention across age groups (Shimi & Scerif, 2022).

In visual working memory, attention facilitates the prioritization of features within multiple objects. Bahle et al. (2020) reveal that these features coactively guide attention, ensuring effective utilization of working memory resources. This interaction exemplifies how attention regulates memory processes to optimize performance. Moreover, the relationship between attention and academic performance is multifaceted and well-documented. Effective attention control supports reading comprehension, problem-solving, and numerical reasoning (Unsworth, Miller, & Robison, 2020). Attentional deficits often result in heightened cognitive load and diminished performance, especially in high-pressure academic environments (Draheim et al. 2021).

Stress significantly challenges attentional control, disrupts working memory processes, and impairs academic outcomes (Hamza & Helal, 2021). Interventions such as mindfulness training have been shown to preserve attention and cognitive performance during stressful tasks and offer practical strategies for mitigating these adverse effects (Almarzouki, 2024). In addition, targeted training programs, including cognitive–emotional–social interventions and computerized working memory exercises, have demonstrated substantial benefits for students with attention-related challenges, such as ADHD (Sarshar, Emadian, & Hassanzadeh, 2024).

Attention mechanisms in working memory are inherently active and involve both automatic and strategic processes. Automatic attentional biases are influenced by salience or reward-based learning, complementing deliberate and goal-directed attention to prioritize information effectively (Ravizza & Conn, 2021). This dual mechanism confirms adaptability across diverse cognitive demands. Furthermore, attention is not static but rhythmic. Peters, Kaiser, Rahm, and Bledowski (2020) proposed that attentional selection follows theta rhythms, enabling cyclical prioritization aligned with external perceptual processes. Such rhythmic dynamics enhance resource allocation efficiency across tasks. Research on prioritization strategies indicates that item-based attention offers stronger recall benefits than dimension-based approaches, underscoring a system’s adaptability to task complexity (Hajonides, Van Ede, Stokes, & Nobre, 2020). Sustained attention, as Liu et al. (2023) accentuated, is critical for maintaining these benefits, particularly in environments with frequent interruptions. The interchange between attention and working memory introduces additional complexity to decision-making. Rustichini et al. (2023) illuminate how these processes collaborate to support effective choices, demonstrating the multifaceted nature of attention in real-world applications. Understanding these intricate mechanisms will facilitate the development of targeted interventions to optimize cognitive and academic performance.

2.4 Executive Control Dimension of Working Memory

The executive control dimension of working memory encompasses higher-order cognitive processes crucial for regulating, managing, and coordinating mental activities. This dimension enables individuals to prioritize tasks, adapt to novel situations, and allocate cognitive resources effectively, as Toba, Malkinson, Howells, Mackie, and Spagna (2020) have elucidated. Unlike the storage and attention dimensions, which focus on retaining and filtering information, executive control integrates these processes to ensure efficient task management. Vandierendonck (2020) and Christophel et al. (2024) pointed out that seamless coordination facilitates this integration, enabling individuals to handle complex cognitive demands effectively.

Executive control, as a core component of executive function, has profound implications for academic success. Spiegel, Goodrich, Morris, Osborne, and Lonigan (2021) showcase their predictive power across various academic domains, including reading, mathematics, and language arts. Mavrou (2020) noted that working memory, a fundamental aspect of executive control, is moderately associated with cognitive development and academic achievement. Privitera, Zhou, and Xie (2022) further corroborate its role in task monitoring and predicting subject-specific outcomes, particularly in complex areas such as mathematics.

Aubry et al. (2021) and Martí et al. (2023) posit that executive control surpasses emotional intelligence as a predictor of academic performance. For instance, inhibitory control, a subcomponent of executive functions, increases the processing of abstract concepts such as rational numbers, thereby improving problem-solving skills. Targeted interventions, such as game-based training programs, can strengthen executive control and yield measurable improvements in academic performance, particularly in reading and mathematics (Vandierendonck, 2021).

The relationship between executive control and academic outcomes is influenced by gender and cultural contexts. While Privitera et al. (2022) suggested that female students often excel in certain academic areas, working memory and interference control measures are consistent across genders. However, Deng et al. (2022) reinforced that cultural factors significantly influence the development of executive functions, mainly inhibitory control, which in turn affects learning outcomes. Alfonso and Lonigan (2021) affirm the importance of culturally sensitive educational strategies to address these variations effectively.

Executive control comprises of several interconnected functions that facilitate efficient cognitive processing. Christophel et al. (2024) claim that it operates across various brain regions, including the middle temporal gyrus and intraparietal sulcus, facilitating memory management independent of sensory input. Vandierendonck (2020) explained that working memory integrates domain-general modules, such as the episodic buffer, with domain-specific modules, including phonological and visuospatial buffers, to support goal-directed tasks. Zink et al. (2021) observed that these distributed processes enhance the flexibility and adaptability of the working memory in response to complex demands. Effective executive control also supports the transition from reactive to proactive cognitive strategies, enhancing sustained attention and task execution (Estefan et al., 2024; Toba et al., 2020). Moreover, it interacts dynamically with inhibitory control and cognitive flexibility to ensure efficient attention regulation, while minimizing distractions (Aubry et al., 2021; García et al., 2021).

Executive control develops significantly during childhood. Nelson, James, Nelson, Tomaso, and Espy (2022) and Ferguson et al. (2021) observe that its components become increasingly differentiated as children progress through early education. Vandierendonck (2020) and Zink et al. (2021) believe that while the central executive plays a pivotal role in managing high-control tasks, its contribution to the overall working memory performance remains unclear. The executive control dimension is also essential for planning, decision-making, and multitasking. Estefan et al. (2024) explored its role in maintaining task-related information and optimizing cognitive processes during complex tasks. García et al. (2021) speculated that executive control is particularly critical in high-demand environments, such as academic settings where multitasking is prevalent. Interventions targeting executive functions can enhance cognitive and academic outcomes, rendering them a valuable tool in educational contexts (Martí et al., 2023; Vandierendonck, 2021; Zink et al., 2021).

2.5 Research Gap

Working memory is widely acknowledged as a multidimensional construct encompassing storage, attention, and executive control. Although each of these components has been extensively examined in prior research, their interrelationships and combined influence on cognitive outcomes remain insufficiently understood. Existing evidence confirms the critical role of each dimension in shaping academic performance, problem-solving, and decision-making (Baddeley, 2020; Cowan, 2020; Vandierendonck, 2020). Nevertheless, the back-and-forth among these components, particularly in the context of cognitive complaints, is yet to be thoroughly investigated. Cognitive complaints, defined as self-reported difficulties in managing cognitive demands, are increasingly prevalent among university students facing heightened academic pressure (Curtis & Sprague, 2021; Ji & Guo, 2023). Understanding how these working memory dimensions interact to shape such complaints is a critical gap in the literature. In particular, prior work tends to isolate single components or uses task-based performance without modelling latent relations among storage, attention, and executive difficulties at the self-report level (e.g., Draheim et al., 2021; Harris et al., 2020; Hitch et al., 2020). Studies on gifted learners or developmental trajectories (Aubry et al., 2021; Ferguson et al., 2021; Forsberg et al., 2021) provide valuable insights at the component level, yet they do not test an integrated structural model of complaints or probe mediation among components. Evidence that attention may transmit the effects of storage constraints onto executive difficulties is suggested conceptually (e.g., Li et al., 2023; Rustichini et al., 2023) but remains under-tested with structural equation modeling (SEM) using validated self-report instruments in university populations.

Moreover, most empirical studies have been conducted in Western cultural and educational settings, limiting the generalizability of the findings to non-Western populations. In Saudi Arabia, the rapid pace of educational reforms and integration of advanced technologies has created unique academic challenges that necessitate localized investigation. Cultural expectations, educational frameworks, and societal stressors may significantly influence the dynamics of working memory and cognitive complaints (Alfonso & Lonigan, 2021; Deng et al., 2022). Work in adjacent higher-education domains within the region often targets broader outcomes, such as student satisfaction or workplace behavior, rather than cognitive complaints or the relationship of memory components (Aman-Ullah & Mehmood, 2023; Ikram, Kenayathulla, & Saleem, 2025a). Even neuroscience-leaning accounts that expand control beyond frontal loci (Christophel et al., 2024) or sleep/strain studies affecting attention/WM/EF (Diotaiuti et al., 2024; García et al., 2021) do not offer a Saudi-specific, student-level structural account linking storage, attention, and executive complaints. Consequently, there is a clear need for a context-sensitive SEM that tests direct and indirect paths among these complaints in Saudi higher education.

In addition to contextual limitations, theoretical models of working memory, such as Baddeley’s multicomponent and dynamic processing models, offer valuable insights into its mechanisms (Logie et al., 2020; Murphy et al., 2020). However, their applicability to real-world scenarios, particularly in understanding the interaction between transient and sustained neural processes under high cognitive load, remains underdeveloped (Myers, 2022; Rose, 2020). A model-based test that brings these frameworks into an applied university context, using a validated measure such as the working memory questionnaire (WMQ) (Aksoy, Saracoglu, & Akkurt, 2022; Guariglia, Giaimo, Palmiero, & Piccardi, 2020; Vallat-Azouvi, Pradat-Diehl, & Azouvi, 2012), can clarify whether attention functions as a conduit between storage and executive complaints and whether these relations vary across fields of study, gender, and academic level.

This study contributes to theory by (i) modelling working memory as an interconnected system of self-reported complaints rather than isolated components; and (ii) testing attention as a mediator between storage and executive difficulties within a structural framework. It contributes to practice by (i) generating Saudi-specific evidence on which student groups (e.g., non-STEM, female) report higher complaint profiles; and (ii) informing targeted supports (e.g., attention-strengthening strategies) that can reduce downstream executive difficulties in high-pressure academic settings.

Therefore, this study aims to answer the following questions:

• To what extent are storage, attention, and executive complaints interrelated and how do demographic factors such as field of study (STEM vs non-STEM), gender, and university level influence these relationships?

To answer this research question, the following hypotheses were tested:

Direct Relationships:

• H1: Storage complaints are positively associated with attention complaints.

• H2: Attention complaints are positively associated with executive complaints.

• H3: Executive complaints are positively associated with attention complaints.

Indirect Relationships:

• H4: Storage complaints indirectly affect executive complaints through attention complaints.

• H5: Storage complaints indirectly affect executive complaints through attention complaints.

Group Differences

• H6: There are statistically significant differences in the levels of Storage, Attention, and Executive complaints based on students’ field of study (STEM vs non-STEM), gender, and university level.

The conceptual framework illustrates the hypothesized relationships among the dimensions of working memory, storage, attention, and executive control and their impact on cognitive complaints. It also pinpoints the moderating effects of demographic factors, including gender, field of study, and academic year, on these relationships (Figure 1).

Figure 1

Conceptual model of working memory dimensions and cognitive complaints.

3.1 Research Design

This study employed a quantitative correlational design to examine the relationships between the three dimensions of working memory, storage, attention, and executive control, and their influence on cognitive complaints among Saudi university students. SEM was utilized as the primary analytical method because of its capacity to investigate complex relationships between variables, including direct and indirect effects. The research also explored the impact of demographic factors, such as field of study, gender, and university level, on these relationships. Data were collected using a validated questionnaire, and the study adhered to a cross-sectional approach to provide a precise and focused analysis of the proposed model in the Saudi educational context.

3.2 Participants

This study included 255 Saudi university students selected through a stratified random sampling approach to achieve a well-rounded and inclusive representation across various demographic and academic factors. Stratification criteria included age, gender, field of study, and year of study, with participants randomly selected within each stratum to reflect the broader student population accurately. The majority of participants were aged 21–23 years (56.9%), followed by 18–20 years (24.3%), 27 years and above (11.8%), and 24–26 years (7.1%). Gender distribution was predominantly female, comprising 85.1% of the sample, with males representing 14.9%. This higher proportion of female participants aligns with the enrollment patterns in Saudi higher education, where women often constitute the majority of university students. Regarding the field of study, 61.6% of the participants were from non-STEM disciplines, while 38.4% were enrolled in STEM fields. The year of study distribution revealed that most participants (60.8%) were in their 4th year, followed by smaller proportions in the 1st year (13.7%), 3rd year (12.9%), and 2nd year (12.5%). Although the stratified random sampling ensured representation across groups, the predominance of female and non-STEM students introduces limitations for generalizability. This imbalance reflects structural characteristics of Saudi higher education, where female enrollment rates are typically higher and non-STEM programs attract larger student numbers. Given Saudi Arabia’s gender-segregated system and academic tracking, these demographic realities must be acknowledged as influencing the scope of inference. While this study provides valuable insights, future research should pursue more balanced samples to strengthen external validity and account for gender and field-based variations in working memory outcomes (Table 1).

3.3 Instrument

The WMQ developed by Vallat-Azouvi et al. (2012) served as the primary instrument for this study, providing a comprehensive assessment of everyday difficulties associated with working memory across three domains: Storage, Attention, and Executive Functioning. The WMQ is divided into two main sections (Appendix 1).

The first section, Part A, collects demographic information, including age, gender, academic year, and field of study (STEM vs Non-STEM), thus providing a detailed contextual profile of the participants. The second section, Part B, comprised 30 self-report items evenly distributed across the three working memory domains. Each item is rated on a six-point Likert scale ranging from Not at All to Extremely, with an additional option for Not Relevant, enabling participants to report the extent to which they experienced challenges in daily tasks related to working memory.

• Storage Dimension (10 items): This domain examines the capacity to temporarily store and retrieve information, with items addressing tasks such as remembering sequences of numbers or understanding text.

• Attention Dimension (10 items): This domain evaluates the ability to maintain focus and manage distractions, including difficulties with multitasking or concentrating in noisy environments.

• Executive Function Dimension (10 items): This domain assesses higher-order cognitive skills such as planning, organizing activities, and adapting strategies to changing situations.

The WMQ has been thoroughly validated in previous studies (Aksoy et al., 2022; Guariglia et al., 2020; Vallat-Azouvi et al., 2012), demonstrating robust psychometric properties, including high internal consistency and construct validity. Recognized as a reliable instrument for assessing cognitive challenges in clinical and non-clinical contexts, the WMQ underwent minimal adaptation in this study to align with Saudi university students’ academic environments, ensuring cultural and contextual relevance while preserving its structural integrity. This approach is consistent with recommendations from recent cross-cultural research (Ikram, Saleem, & Mehmood, 2025b), which stresses the importance of adapting and validating instruments to reflect local educational contexts while maintaining theoretical consistency. This choice was further supported by evidence of the WMQ’s successful use across diverse populations, which confirms its flexibility for cross-cultural applications. In the Saudi context, we used forward–back translation and expert review to ensure conceptual equivalence and cultural appropriateness. Only minor modifications (e.g., wording examples to reflect academic practices standard in Saudi higher education) were needed, demonstrating that the original constructs were applicable without substantive changes. The subsequent validation through exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) in this study provides empirical confirmation that the WMQ is both reliable and appropriate for the Saudi student population.

3.4 Data Collection and Ethical Considerations

Data were collected through a Google Forms survey during the first semester of the 2024–2025 academic year. The participants were informed of the research objectives and provided informed consent before participation. Ethical approval was obtained from the institutional review board, which ensured compliance with ethical standards. Participation was voluntary, and individuals were assured of their right to withdraw at any stage. All responses were anonymized, and data were stored securely for research purposes.

3.5 Data Analysis

Data analysis was conducted in a structured, multistage process to validate the proposed model and test the research hypotheses. Initially, an EFA was performed to uncover the underlying factor structure of the constructs. The KMO measure and Bartlett’s Test of Sphericity confirmed the suitability of the data for factor analysis, and items with low factor loadings or significant cross-loadings were excluded. Subsequently, CFA was employed to validate the measurement model, with fit indices, such as CFI, TLI, and RMSEA, confirming the robustness of the model. Descriptive statistics were then computed to provide a comprehensive overview of cognitive complaints across groups. Moreover, to examine group differences, a one-way Multivariate Analysis of Variance (MANOVA) was conducted to assess the influence of the field of study, gender, and year of study on cognitive complaints. Finally, through path analysis, SEM was used to analyze the direct and indirect relationships among the constructs. The analyses were conducted using SPSS for EFA, descriptive statistics, and MANOVA, while AMOS was used for CFA and SEM.

4.1 EFA

The results shown in Table 2 confirm the suitability of the EFA dataset. The KMO measure of sampling adequacy was 0.903, which exceeded the recommended threshold of 0.50. This indicates that the dataset is highly appropriate for factor extraction, owing to the strong correlations among the variables. Moreover, Bartlett’s Test of Sphericity yielded a statistically significant result (χ ² = 3770.316, df = 435, p < 0.001), verifying that the correlation matrix is not an identity matrix and justifying the application of EFA. Further evidence of suitability can be observed in the correlation matrix (Appendix 2), where coefficients greater than 0.3 support the appropriateness of the data (Field, 2017). Together, these results validate the structure of the dataset and its capacity to reveal meaningful factor dimensions, thereby providing a robust foundation for subsequent analyses.

Principal component analysis (PCA) was employed to examine the underlying structure of the dataset and determine the optimal number of factors to be retained. The initial analysis identified four components with eigenvalues greater than 1, as presented in Table 3. Collectively, these components accounted for 67.155% of the total variance, with the first contributing 38.706%, followed by 16.227, 8.290, and 3.931%. This result aligns with the Kaiser criterion, which suggests retaining factors with eigenvalues exceeding 1. A scree plot (Figure 2) was analyzed to refine the factor structure further, revealing an inflexion point after the third component and suggesting a three-factor solution. To validate this finding, a PA was conducted in line with best practices (Patil, Singh, Mishra, & Donavan, 2007). The results indicate that only the first three components in the PCA had eigenvalues exceeding their corresponding simulated values in the PA. Consequently, a three-factor solution was deemed most appropriate for this dataset. Integrating the Kaiser criterion, scree plot, and PA, this multi-method approach ensures a robust determination of the factor structure. These findings provide a solid foundation for subsequent validation through CFA.

Figure 2

Scree plot.

The scree plot in Figure 2 shows a clear inflexion point after the third component, indicating a three-factor solution. The steep decline in eigenvalues for the first three components, followed by a plateau, suggests that these three components account for the most meaningful variance in the data, whereas the subsequent components contribute minimally. This supports the decision to retain these three factors for further analysis.

Table 4 compares eigenvalues obtained from PCA and criterion values generated through PA. The findings confirmed that the first three components, with eigenvalues of 11.612, 4.868, and 2.487, exceeded the corresponding PA thresholds of 1.704, 1.600, and 1.523, respectively. These results support the retention of three factors, as they account for meaningful variance in the data. At the same time, components with eigenvalues below the PA criterion were excluded, indicating minimal contribution beyond random noise. Specific criteria were applied during the item selection to refine the factor structure further and ensure interpretability. Items with factor loadings below 0.5 were removed to maintain the strength of the retained components. Moreover, items exhibiting cross-loadings, defined as loadings on multiple factors with a difference of less than 0.2, were excluded.

Table 5 illustrates the rotated component matrix, confirming a well-defined, three-factor structure aligned with the theoretical dimensions of Storage, Attention, and Executive complaints. The majority of the items exhibited strong factor loadings above the acceptable threshold of 0.50, demonstrating their relevance to their respective constructs. However, item 2 from the storage complaints construct was removed due to cross-loading (0.537 on Component 1 and 0.558 on Component 3), while item 7 (storage complaints) with a loading of 0.426 and item 9 (executive complaints) with a loading of 0.381 were excluded for falling below the threshold. Following these adjustments, 27 items were retained: 8 for storage complaints, 10 for attention complaints, and 9 for executive complaints. These results reinforce the robustness and validity of the identified factor structures.

4.2 CFA

The CFA path diagram illustrates a robust three-factor structure encompassing the constructs of Storage, Attention, and Executive complaints along with their respective observed variables (Figure 3). The factor loadings for all items exceeded the recommended threshold of 0.50, demonstrating strong and consistent relationships between the latent constructs and their indicators. Specifically, Storage items exhibited loadings ranging from 0.745 to 0.882, attention items from 0.617 to 0.833, and executive items from 0.801 to 0.891, underscoring the reliability of these measures. The inter-factor correlations are moderate and positive, with values of 0.32 between Storage and Attention, 0.35 between Storage and Executive, and 0.37 between Attention and Executive. These correlations confirm the theoretical interdependence of the constructs, while maintaining conceptual distinctiveness. The model fit indices further validated the measurement model, with X ² = 513.324, df = 321, X ²/df = 1.59, CFI = 0.922, TLI = 0.953, and RMSEA = 0.033, all indicating an excellent fit to the data. The model also effectively accounts for measurement errors, ensuring that the unexplained variance for each observed variable is appropriately captured. These results confirm the validity and reliability of the hypothesized three-factor structure, providing strong evidence of the scale’s construct validity and robustness.

Figure 3

CFA model.

Table 6 confirms the constructs’ validity, with AVE values above 0.50 (Storage: 0.726, Attention: 0.603, Executive: 0.625) and CR values exceeding 0.70 (Storage: 0.852, Attention: 0.848, Executive: 0.871), ensuring reliability. Discriminant validity is evident as the square roots of AVE (diagonal values) surpass inter-construct correlations (e.g., Storage: 1.00 > 0.74, 0.64), affirming the distinctiveness and robustness of the constructs (Cheung, Cooper-Thomas, Lau, & Wang, 2024).

4.3 Hyphothesis Testing

The structural model was evaluated to test the hypothesized relationships between Storage, Attention, and Executive complaints. Each hypothesis was assessed using standardized path estimates and significance levels. The findings strongly support all the proposed hypotheses, articulating the presence of both direct and indirect effects within the model. Figure 4 presents the structural pathways, factor loadings, and interrelations among constructs.

Figure 4

Path analysis.

The results of the path analysis provided strong empirical support for all hypothesized relationships, as outlined below.

• H1: Storage complaints are positively associated with attention complaints, with a standardized path coefficient of 0.67.

  Verdict: Supported.

• H2: Attention complaints are positively associated with executive complaints, with a standardized path coefficient of 0.45.

  Verdict: Supported.

• H3: Storage complaints indirectly influence executive complaints through attention complaints. The indirect effect was significant, with a combined path coefficient of 0.30.

  Verdict: Supported.

• H4: Storage complaints have the strongest direct effect on executive complaints compared with attention complaints. The direct effect of Storage on Executive complaints is 0.54, which is higher than the direct effect of Attention on Executive complaints at 0.45, confirming that storage has a stronger influence.

  Verdict: Supported.

• H5: The relationship between storage and executive complaints is hierarchically mediated by attention complaints. Mediation analysis showed a significant indirect effect of 0.30, reinforcing hierarchical mediation.

Verdict: Supported.

To examine the differences in cognitive complaints based on field of study, gender, and year of study, a series of one-way multivariate analyses of variance (MANOVAs) were conducted. The results are summarized as follows:

H6a: Field of Study (STEM vs Non-STEM)

A one-way MANOVA revealed significant differences in cognitive complaints between STEM and non-STEM participants (Wilks’ Lambda = 0.925, F(3, 251) = 6.79, p < 0.01, partial eta squared = 0.075; Table 7). Post-hoc analyses (Table 8) showed that all three complaints – Storage (F(1, 253) = 10.86, p < 0.01, partial eta squared = 0.041), attention (F(1, 253) = 13.63, p < 0.01, partial eta squared = 0.051), and executive (F(1, 253) = 20.13, p < 0.01, partial eta squared = 0.074) – differed significantly by field of study. Non-STEM participants reported higher complaints across all dimensions, with mean scores for storage (M = 2.99), attention (M = 3.36), and executive function (M = 3.32) compared to STEM participants (M = 2.70, 2.96, 2.82, respectively; see Table 9).

H6b: Gender

A one-way MANOVA assessing the effect of gender on cognitive complaints found no significant differences in the combined complaints (Wilks’ lambda = 0.971, p > 0.05, partial eta squared = 0.029; see Table 10). However, the individual analyses (Table 11) revealed a significant difference in executive complaints (F(1, 253) = 5.14, p = 0.024, partial eta squared = 0.020). Female participants reported more executive complaints (M = 3.43) than males (M = 3.08; see Table 12). No significant differences were observed in storage or complaints of attention.

H6c: Year of Study

A one-way MANOVA was conducted to assess the impact of the year of study on cognitive complaints. The results showed no significant differences in the combined complaints (Wilks’ Lambda = 0.965, p > 0.05, partial eta squared = 0.012; Table 13) or for any of the individual dimensions (Table 14). Descriptive statistics (Table 15) indicate relatively consistent mean scores across all years for storage (M = 2.80–2.98), attention (M = 3.12–3.38), and executive complaints (M = 3.09–3.32).

The field of study emerged as the sole factor demonstrating significant differences across all dimensions of cognitive complaints, with non-STEM participants consistently reporting higher levels than their STEM counterparts did. Gender differences were observed exclusively in executive complaints, with females reporting higher scores than males. Conversely, the year of study had no significant impact on cognitive complaints, reflecting consistent patterns across academic levels. These findings demonstrate the potential influence of contextual factors, such as the field of study and gender, on cognitive complaints, suggesting that the study year plays a minimal role.

5.1 Relationships Between Working Memory Dimensions

This study identified significant relationships among the dimensions of working memory, confirming the proposed hypotheses. Storage complaints were positively associated with attention complaints (H1), suggesting that difficulties in retaining information intensify the cognitive demands on the attentional processes. Attention complaints, in turn, were positively associated with executive complaints (H2), demonstrating the reliance on higher-order cognitive functions such as planning and decision-making on sustained attention. Moreover, storage complaints were directly linked to executive complaints (H3), revealing that challenges in memory retention can affect executive functioning without necessarily involving attentional mechanisms.

The relationship between storage and attention complaints reflects the compensatory effort required when the storage processes are strained. Difficulties in retaining sequences of information, comprehending text, or recalling recently acquired knowledge often necessitate additional cognitive resources to maintain a focus on tasks. Shen et al. (2021) demonstrated that limitations in storage capacity heighten reliance on attentional systems to coordinate stored and incoming information. Within academic settings, such constraints may manifest as reduced concentration during lectures or difficulties in managing tasks that demand simultaneous processing, thereby intensifying cognitive complaints.

The link between attention and executive complaints illustrates the critical role of attention in enabling complex cognitive tasks. As a foundational process, attention supports decision-making, planning, and adaptive problem-solving. Any disruption caused by cognitive fatigue or external distractions can hinder the execution of these tasks. Vandierendonck (2020) outlined the dependency of executive functions on effective attentional allocation, while Draheim et al. (2021) illustrated that maintaining focus is crucial for managing competing demands and ensuring cognitive efficiency. For instance, students with diminished attentional control may experience difficulties organizing activities or adapting strategies during problem-solving, leading to increased executive complaints.

The direct relationship between storage and executive complaints reinforces how memory deficits can impair executive functioning independent of attention. This is evident in tasks requiring sequential reasoning or goal-directed actions, where failures in memory retention disrupt the task execution. Toba et al. (2020) discuss how such deficits limit the availability of critical information needed for decision-making and planning. For example, a student who is unable to recall key lecture content may encounter difficulties organizing study materials, ultimately affecting overall academic performance. This finding reflects the cascading effects of storage deficits on executive processes, even when attentional capacity remains intact.

These results align with theoretical frameworks, such as Baddeley’s multicomponent model, which details the interconnectedness of storage, attention, and executive functions in supporting cognitive performance. Moreover, as explored by Barrouillet et al. (2024a), cognitive load theories propose that deficits in one working memory dimension can extend to others and exacerbate cognitive complaints. These interrelations also carry educational implications. For Saudi university students, the cumulative effect of storage, attention, and executive complaints may hinder not only day-to-day classroom performance but also long-term skill acquisition. For instance, tasks requiring sustained reading or integrating lecture material with independent study can overwhelm students with weaker storage functions, resulting in broader executive inefficiencies. Addressing such challenges through strategies that reduce memory load, such as scaffolded instruction, the use of external memory aids, or structured note-taking, may help mitigate the cascading impact of deficits across working memory dimensions. This sets the stage for examining how attention mediates the relationship between storage and executive functions, as discussed in the next section.

5.2 Mediated Relationships Among Working Memory Dimensions

The findings revealed that storage complaints indirectly influence executive complaints through attention complaints (H4), demonstrating the mediating role of attention in this relationship. Challenges in memory retention, such as recalling sequences or integrating new information, place additional demands on the attentional systems, leading to increased cognitive strain. This intensified burden on attention disrupts executive processes, including planning, decision-making, and adaptive problem-solving. These cascading effects illustrate how impairments in one dimension of working memory can extend to others, exacerbating the overall cognitive challenges. This mediation aligns with research by Hu and Hu (2023), who discussed the compensatory demands on attention caused by limitations in working memory storage. Their findings indicate that sustaining task performance under such conditions often demands increased attentional effort, which in turn contributes to greater cognitive complaints. Similarly, Rustichini et al. (2023) examined the interaction between attention and executive functions, showing that impairments in attentional resources can compromise decision-making by limiting the cognitive capacity required for goal-directed activities.

The sequential mediation described in H5, where storage complaints affect executive complaints through attention complaints, provides further insights. Although storage deficits directly impair executive function, attention mechanisms predominantly mediate their impact. For instance, difficulties retaining essential information often necessitate sustained focus, diminishing the efficiency of executive processes that rely on stable attentional resources. Zhong et al. (2024) supported these findings by showing that attentional allocation acts as a pivotal intermediary, reducing the adverse effects of storage limitations on broader cognitive functions.

Li et al. (2023) further validate this pathway, demonstrating the central role of controlled attention in mitigating the downstream impact of storage deficits on complex cognitive tasks, such as mathematical problem-solving. Their research shows that attention disruptions limit compound storage and intensify their adverse effects on executive functioning, particularly in cognitively demanding scenarios. The mediated pathways identified in H4 and H5 reveal the interconnectedness of the working memory dimensions. Attention has emerged as a crucial link between storage and executive complaints, reinforcing the integrated nature of working memory. This perspective advances the understanding of how deficits in one dimension can indirectly, yet significantly, affect others, offering a foundation for targeted strategies to address working memory-related cognitive challenges. From the perspective of Cognitive Load Theory (Barrouillet, Portrat, & Camos, 2024b; Sweller, 2010), this mediation suggests that storage limitations heighten intrinsic load, overextending attentional resources and leaving fewer capacities available for executive control. Such strain reflects the trade-off between storage and processing stressed in working memory models. Furthermore, neurological evidence indicates that executive networks continue maturing through early adulthood (Best & Miller, 2010), suggesting that university students may be especially vulnerable to attentional bottlenecks caused by storage deficits. Recognizing these mechanisms is essential for instructional design, such as chunking material or offering external supports to alleviate attentional burden.

5.3 Group Differences in Cognitive Complaints

The findings affirmed H6, revealing notable differences in cognitive complaints across the demographic groups. Differences emerged based on field of study (H6a) and gender (H6b), whereas no significant variations were observed across academic years (H6c).

Regarding H6a, non-STEM students exhibited significantly higher storage levels, attention, and executive complaints than their STEM counterparts. This outcome reflects the domain-specific cognitive demands of each academic field. STEM disciplines often involve structured tasks, such as mathematical problem-solving and technical analyses, which enhance working memory performance through repetitive cognitive engagement (Ji & Guo, 2023). Conversely, non-STEM fields place greater emphasis on verbal reasoning and creative problem-solving, which may not provide the same reinforcement for working memory capacity. Supporting this, Vergauwe, Von Bastian, Kostova, and Morey (2021) identified how task-specific demands shape the balance between storage and processing capacities, particularly in disciplines with less-structured cognitive frameworks. In the Saudi higher education context, this disciplinary gap may also reflect systemic differences in curriculum design, where STEM programs often embed incremental cognitive challenges. In contrast, non-STEM programs may not consistently scaffold memory and attention skills in the same way.

For H6b, gender differences were evident, with female participants reporting higher executive complaints. This observation aligns with research by Al-Baghdadi and Ashmawy (2021) and Alfonso and Lonigan (2021), who identified heightened executive complaints among female students, potentially linked to greater emotional regulation demands and multitasking pressures. In societies with gender-segregated education, such as Saudi Arabia, cultural expectations around performance and responsibility may place additional cognitive load on female students, thereby amplifying these complaints. As Amzil (2022) suggests, societal norms can compound stress levels and shape gendered differences in cognitive functioning. Moreover, Draheim et al. (2021) demonstrated that variations in executive functioning can often be attributed to stress-related factors, which disproportionately affect women in academic settings that require extensive planning and adaptability.

Regarding H6c, the absence of significant differences across academic years implies that working memory complaints remained consistent throughout university education. This finding contrasts with Nelson et al. (2022), who reported developmental improvements in executive function over time. One explanation may be the relatively uniform curricular structure in Saudi universities, where progression across years does not always translate into greater cognitive complexity or differentiated workload. Alternatively, external factors such as high academic expectations, exam-driven assessment, and limited curricular flexibility may overshadow year-specific influences, masking developmental improvements typically observed in other contexts (Hamza & Helal, 2021).