Mental Effort and Load–Translingual Scale (MEL-TS)

Abstract

The Mental Effort and Load–Translingual Scale (MEL-TS) (Endres et al., 2025) is an advanced psychometric instrument developed to assess learning-related cognitive demand by disentangling two theoretically intertwined yet distinct dimensions of cognitive expenditure: mental load (the task-centered cognitive demands imposed on the cognitive architecture) and mental effort (the actual cognitive capacity voluntarily allocated by the learner to satisfy those demands). Built to overcome decades of measurement conflation in educational psychology, instructional design, and human factors research, the scale was created using a translingual methodology involving qualitative interviews with bilingual individuals and a rigorous translation and back-translation protocol across English, German, Chinese (Mandarin), Dutch, Spanish, and French. The instrument exists as an 8-item long version (composed of two 4-item subscales) and a streamlined 6-item short version (composed of two 3-item subscales), each assessed on a 9-point Likert scale ranging from 1 (strongly disagree) to 9 (strongly agree). Psychometric evaluation across adult cohorts demonstrated strong internal consistency, with Cronbach’s alpha values ranging between .81 and .88 across experimental conditions and language implementations. Criterion-related validity was confirmed using a repeated-measures, within-subjects paradigm comprising eight experimental tasks systematically manipulating task complexity, extraneous cognitive interference, and learner motivation. Furthermore, the MEL-TS demonstrated robust convergent and incremental validity, exhibiting medium-to-large associations with classical single-item metrics and multi-dimensional inventories while capturing unique variance in self-regulated learning and task performance. This article offers an extensive psychometric profile of the MEL-TS, exploring its theoretical underpinnings, dimensional structure, translingual validation procedures, administrative protocols, and future applications across global learning environments.

Keywords

Mental Effort and Load–Translingual Scale, MEL-TS, Cognitive Load Theory, mental load, mental effort, self-regulated learning, translingual psychometrics, cognitive demand, working memory capacity, instructional design, cross-cultural adaptation, educational measurement

Authors

The Mental Effort and Load–Translingual Scale (MEL-TS) was designed and validated by an international consortium of educational psychologists and psychometric researchers:

• Tino Endres (Corresponding Author) — Department of Psychology, University of Freiburg, Freiburg, Germany. ORCID: 0000-0003-3334-4064. Email: [email protected]
• Lisa Bender — Department of Psychology, University of Freiburg, Freiburg, Germany. ORCID: 0000-0002-8972-6558.
• Stoo Sepp — School of Education, University of New England, Armidale, Australia. ORCID: 0000-0002-0509-4603.
• Shirong Zhang — Department of Psychology, Education, and Child Studies, Erasmus University Rotterdam, Rotterdam, Netherlands. ORCID: 0000-0002-6975-5386.
• Louise David — School of Health Professions Education (SHE), Department of Educational Development and Research, Maastricht University, Maastricht, Netherlands. ORCID: 0000-0003-1973-4568.
• Melanie Trypke — Institute of Educational Science, University of Osnabrück, Osnabrück, Germany. ORCID: 0000-0002-6970-5741.
• Dwayne Lieck — Department of Psychology, University of Freiburg, Freiburg, Germany. ORCID: 0000-0001-9814-3488.
• Juliette C. Désiron — Department of Psychology, University of Freiburg, Freiburg, Germany. ORCID: 0000-0002-3074-9018.
• Johanna Bohm — Department of Psychology, University of Freiburg, Freiburg, Germany. ORCID: 0009-0000-1026-4809.
• Sophia Weissgerber — Department of Psychology, University of Kassel, Kassel, Germany. ORCID: 0000-0002-6441-358X.
• Juan Cristobal Castro-Alonso — School of Education, University of Birmingham, Birmingham, United Kingdom. ORCID: 0000-0003-3186-3717.
• Fred Paas — Department of Psychology, Education, and Child Studies, Erasmus University Rotterdam, Rotterdam, Netherlands, and School of Education, University of New South Wales, Sydney, Australia. ORCID: 0000-0002-1647-5305.

Purpose

The primary purpose of the Mental Effort and Load–Translingual Scale (MEL-TS) is to establish a rigorous, psychometrically sound, and translingually equivalent measurement infrastructure for educational psychology, cognitive science, and human-computer interaction. For over three decades, researchers investigating human cognitive architecture within the paradigm of Cognitive Load Theory (CLT) have relied on subjective self-report instruments. However, early research often employed single-item metrics—such as the widely used 9-point scale pioneered by Paas (1992)—which prompted learners to rate either their perceived “mental effort” or perceived “difficulty.” Over time, researchers frequently used the terms “mental load,” “mental effort,” and “task difficulty” interchangeably, introducing conceptual ambiguity into empirical studies.

This terminological conflation presents a critical challenge within modern educational frameworks, particularly in the study of self-regulated learning (SRL). In self-regulated contexts, learners must continuously monitor internal cognitive states and make strategic metacognitive decisions regarding resource allocation. A learning task may impose heavy cognitive requirements (high mental load) due to complex concepts or poor instructional design. Yet, a learner may choose either to expend maximal cognitive resources to conquer the material (high mental effort) or disengage entirely due to frustration, fatigue, or low motivation (low mental effort). When an assessment tool fails to separate these two dimensions, researchers cannot determine whether an elevated score reflects an overwhelmingly complex task, a motivated learner working at full capacity, or an interaction between the two.

The MEL-TS was engineered to resolve these diagnostic ambiguities by providing separate, internally coherent subscales for mental load and mental effort. Its purpose spans several key domains:

• Disentangling Demand from Investment: Providing empirical separation between the objective/subjective constraints of the task environment (mental load) and the energetic engagement deliberately initiated by the learner (mental effort).
• Translingual Equivalence: Overcoming cross-linguistic and cultural discrepancies in cognitive load assessment. Many non-English languages do not have distinct vocabulary that neatly maps onto the English distinction between “effort” and “load.” The MEL-TS was systematically formulated through qualitative interviews and back-translation procedures across multiple global languages, establishing semantic equivalence.
• Evaluating Digital and Automated Learning Systems: Assessing how learners interact with complex multimedia platforms, intelligent tutoring systems, and adaptive learning algorithms, where real-time monitoring of cognitive expenditure is critical for tailoring instructional scaffolding.
• Advanced Research in Self-Regulated Learning: Empowering researchers to track metacognitive monitoring and control processes across diverse experimental tasks without burdening participants with long, intrusive questionnaires.

Psychological Construct

The MEL-TS evaluates learning-related cognitive demand as a multi-dimensional construct comprising two correlated, yet distinct, latent factors: Mental Load and Mental Effort. Although both factors reside within the broader domain of human channel capacity and energetic expenditure, they reflect different psychological mechanisms and behavioral realities.

1. Mental Load (Task-Centered Demand)

Mental load reflects the aspect of cognitive demand that originates primarily from the instructional task, the material’s structural architecture, and the environmental constraints acting upon the learner. Grounded in the classical formulations of Paas and van Merriënboer (1994), mental load is defined as the burden that a task places on a learner’s working memory system. This factor encompasses:

• Element Interactivity and Intrinsic Complexity: The number of cognitive schema elements that must be processed concurrently in working memory to comprehend the instructional material.
• Extraneous Cognitive Interference: Demands introduced by poor instructional design, split-attention effects, redundant multimedia presentations, or environmental distractions that consume working memory capacity without directly assisting schema acquisition.
• Subjective Demand Perception: The participant’s cognitive evaluation of how challenging, demanding, or mentally taxing the activity was, irrespective of whether they elected to meet that demand.

In the MEL-TS, mental load items measure this dimension from the perspective of the task’s inherent strain. Example items include: “Working on this activity was mentally demanding” and reverse-coded statements such as “Working on this activity was effortless.” High scores on the mental load subscale denote that the user perceived the activity as requiring substantial information processing capacity, placing heavy demands on limited cognitive resources.

2. Mental Effort (Person-Centered Energetic Investment)

In contrast to mental load, mental effort represents the actual cognitive capacity that the learner chooses to allocate toward meeting task demands. Drawing upon Kahneman’s (1973) capacity model of attention and motivation-driven cognitive allocation frameworks, mental effort is an internal, volitional, and behavioral phenomenon. It reflects:

• Resource Allocation: The actual proportion of finite central executive capacity directed toward problem solving, schema construction, and metacognitive monitoring.
• Motivational Volition: The learner’s determination and willingness to invest energy in the presence of cognitive challenge, directly linked to self-efficacy, achievement goals, and perceived task value.
• Active Cognitive Engagement: The conscious, effortful exertion of intellectual energy (e.g., “working hard,” “putting heart and soul into the activity”) as opposed to superficial skimming, mindless guessing, or avoidance behaviors.

The MEL-TS captures mental effort through items focusing directly on personal investment. Example statements include: “I invested effort while working on this activity” and “I worked hard on this activity.” A high mental effort score indicates that the individual actively directed their cognitive faculties toward the task, regardless of whether the task was inherently easy or exceptionally demanding.

The Dual-Construct Dynamic

The distinction between mental load and mental effort becomes clear when examining divergent learning scenarios. For example, during a poorly designed or overly complex task, a disengaged or frustrated student may report very high mental load (the task was perceived as overwhelming) but near-zero mental effort (having given up, allocating no real capacity to solving it). Conversely, an intrinsically motivated student tackling an accessible, moderately challenging puzzle may report moderate mental load but exceptionally high mental effort, driven by deep curiosity and focused attention. By capturing both vectors simultaneously, the MEL-TS enables researchers to plot learners across a four-quadrant matrix of cognitive demand, offering deeper diagnostic insights than single-item composite measures.

Theoretical Framework

The theoretical framework of the MEL-TS integrates principles from Cognitive Load Theory (CLT), Kahneman’s Capacity Model of Attention, and contemporary models of Self-Regulated Learning (SRL).

Cognitive Load Theory and Cognitive Architecture

Cognitive Load Theory, originated by John Sweller (1988) and advanced by Fred Paas, Jeroen van Merriënboer, and colleagues, rests on a foundational model of human cognitive architecture comprising an essentially unlimited long-term memory and a strictly capacity-limited working memory. Working memory can process only a small number of novel information elements simultaneously (typically 3 to 5 elements) for a short duration. Consequently, instructional designs must manage the demands imposed on working memory to facilitate the construction and automation of cognitive schemas in long-term memory.

Early CLT formulations differentiated between three types of cognitive load: intrinsic load (inherent complexity arising from element interactivity), extraneous load (unnecessary processing induced by ineffective instruction), and germane load (processing devoted to schema acquisition and automation). However, modern revisions of CLT (e.g., Sweller et al., 2011) re-conceptualized germane load not as an independent source of cognitive load, but rather as the mental effort learners invest to manage intrinsic demands. Under this revised framework, an explicit psychometric distinction between external demand (load) and internal resource allocation (effort) became essential. The MEL-TS operationalizes this updated model by treating mental load as the combined burden of intrinsic and extraneous elements, while mental effort reflects the active processing resources deployed by the learner.

Kahneman’s Energetic Model of Attention

The theoretical separation between load and effort is deeply rooted in Daniel Kahneman’s (1973) capacity model of attention. Kahneman posited that human cognitive capacity is variable, determined by arousal, task demands, and deliberate allocation policies. While task demands establish the theoretical ceiling or floor of required energy (mental load), the individual’s evaluation of demand, combined with motivation and intentionality, dictates the actual amount of capacity supplied (mental effort). The MEL-TS directly reflects this allocation framework: load measures the perceived demands of the task, whereas effort measures the resources allocated through deliberate intent.

Self-Regulated Learning (SRL) Frameworks

Within models of self-regulated learning (e.g., Winne & Hadwin, 1998; Zimmerman, 2000), learners progress through cyclical phases of planning, cognitive execution, and metacognitive reflection. During task performance, learners monitor both subjective strain and task difficulty to determine whether to persist, adopt new strategies, or abandon the activity. When instructional tools fail to distinguish between load and effort, they obscure these self-regulatory mechanisms. The MEL-TS provides a clear measurement framework for evaluating how metacognitive judgements of load inform the voluntary upregulation or downregulation of effort.

Validity

The validation program for the MEL-TS (Endres et al., 2025) involved extensive psychometric testing across multiple experimental conditions, establishing its construct, criterion-related, convergent, and incremental validity.

Content and Translingual Validity

To ensure high content validity, the authors avoided the common pitfall of simply translating English idioms into other languages. Instead, the team conducted qualitative, semi-structured interviews with bilingual and multilingual individuals to uncover how speakers of different languages conceptualize words such as “strain,” “effort,” “load,” “difficulty,” and “working hard.” Items were generated from both classic cognitive load literature and novel candidate pools, followed by a rigorous forward- and back-translation process involving native-speaking psychometricians across English, German, Mandarin Chinese, Dutch, Spanish, and French. This careful process ensured that semantic and operational meanings remained equivalent across linguistic contexts.

Criterion-Related Validity

The criterion-related validity of the MEL-TS was examined using an intensive, within-subjects experimental design featuring eight distinct task conditions. The experimental manipulations systematically altered three primary independent variables:

1. Task Complexity (Intrinsic Load): Manipulating the element interactivity of reasoning and problem-solving tasks.
2. Extraneous Load Manipulations: Introducing visual split-attention and cognitive interference during instructional presentations.
3. Motivational Cues: Implementing goal-framing instructions and incentives designed to encourage cognitive investment.

Repeated-measures analyses confirmed that the scale was sensitive to these experimental manipulations. The Mental Load subscale responded strongly to variations in task complexity and extraneous interference, showing significant increases in subjective load scores under high-complexity and split-attention conditions. Conversely, the Mental Effort subscale was sensitive to motivational framing: when learners were presented with high-complexity tasks paired with incentives for deeper engagement, effort scores rose significantly, even when extraneous load was held constant. Under conditions designed to induce ego-depletion or disengagement, effort scores declined while load scores remained elevated, confirming that the two subscales capture distinct behavioral phenomena.

Convergent and Incremental Validity

Convergent validity was evaluated by correlating the MEL-TS subscales with established cognitive load inventories, including the traditional Paas (1992) 1-item cognitive load scale, the NASA-Task Load Index (NASA-TLX), and the multi-component inventory developed by Kolfschoten and colleagues. The MEL-TS subscales demonstrated medium-to-large positive correlations with these benchmark instruments, confirming that the tool measures the broader construct of cognitive demand.

Crucially, hierarchical multiple regression analyses confirmed the scale’s incremental validity. When predicting post-task knowledge acquisition and diagnostic problem-solving performance, the dual-subscale structure of the MEL-TS accounted for statistically significant unique variance beyond that explained by legacy single-item measures. Specifically, entering mental load and mental effort as independent predictors uncovered an interaction effect: high mental effort predicted superior learning gains primarily when mental load was kept within manageable levels, a finding that single-item measures could not capture.

Reliability

The MEL-TS has demonstrated high internal consistency and measurement precision across multiple experimental trials, linguistic samples, and task environments. Psychometric assessments of both the full 8-item inventory and the shortened 6-item version show robust reliability.

Internal Consistency

In the primary validation study conducted by Endres et al. (2025), Cronbach’s alpha ($lpha$) coefficients were computed across all experimental tasks and linguistic adaptations. The internal consistency metrics for the overall sample and individual subscales showed high reliability:

• Total Scale Internal Consistency: Cronbach’s $lpha$ for the overall scale ranged from .81 to .88 across experimental conditions, reflecting high measurement reliability across diverse tasks.
• Mental Load Subscale: The mental load subscale demonstrated strong internal consistency, with $lpha$ values routinely exceeding .82 in both the 4-item long version and the 3-item short version. Corrected item-total correlations for the load items consistently exceeded .55, confirming that the items coherently reflect task-centered cognitive strain.
• Mental Effort Subscale: The mental effort subscale similarly yielded $lpha$ coefficients between .80 and .87 across test settings. The subscale exhibited clean internal cohesion, with items measuring deliberate energetic engagement loading strongly onto the effort factor.

Split-Half and Temporal Stability

Split-half reliability analyses applying the Spearman-Brown correction produced coefficients exceeding .83 across testing blocks. Because cognitive load is a transient, state-based psychological phenomenon that fluctuates based on immediate task demands, traditional test-retest reliability across long time intervals is neither theoretically expected nor methodologically appropriate. However, within repeated-measures blocks evaluating parallel task variants of identical complexity, the scale demonstrated stable within-subject consistency, showing minimal measurement error across repeated measurement occasions.

Factor Analysis

Structural evaluations of the MEL-TS were conducted to examine its latent dimensionality and verify the theoretical independence of the Mental Load and Mental Effort constructs across multiple languages.

Confirmatory Factor Analysis (CFA)

Confirmatory Factor Analysis was performed on the validation datasets to compare two competing models:

1. Unidimensional Model: A single-factor model positing that all items load onto a unified “general cognitive demand” factor.
2. Two-Factor Correlated Model: An oblique two-factor model positing separate latent factors for Mental Load (items 1–4) and Mental Effort (items 5–8), allowing the two dimensions to correlate freely.

The statistical findings demonstrated that the unidimensional model yielded poor fit indices (e.g., Comparative Fit Index $[\text{CFI}] < .85$, Root Mean Square Error of Approximation $[\text{RMSEA}] > .10$), failing to adequately account for the data. In contrast, the two-factor model demonstrated good fit to the empirical data across multiple evaluation criteria:

• Comparative Fit Index (CFI): Values consistently exceeded .95, indicating excellent model fit.
• Tucker-Lewis Index (TLI): Coefficients remained above .94 across long and short versions.
• Root Mean Square Error of Approximation (RMSEA): Values ranged between .042 and .058 (with 90% confidence intervals staying below the conventional .08 threshold), indicating low residual error.
• Standardized Root Mean Square Residual (SRMR): Observed values remained below .05.

Factor Loadings and Latent Correlation

Standardized factor loadings across all items in both the 8-item and 6-item versions were robust, with primary loadings ranging from .62 to .86 on their designated latent factors. Loadings on non-target factors remained low, confirming simple structure.

The latent correlation between the Mental Load factor and the Mental Effort factor was positive and moderate ($r \approx .35\text{ to }.52$), consistent with theoretical expectations: while task difficulty (load) and learner engagement (effort) often co-occur, they are sufficiently distinct to reject a single-construct interpretation.

Measurement Invariance

To verify translingual equivalence across linguistic cohorts (e.g., English, German, and Mandarin Chinese speakers), multigroup confirmatory factor analyses (MGCFA) were conducted. The scale demonstrated configural invariance (identical factor structure across groups), metric invariance (equivalent factor loadings across languages, $\Delta\text{CFI} < .01$), and scalar invariance (comparable item intercepts, allowing for meaningful cross-cultural comparisons of latent means).

Instrument / Measurement Tool

The MEL-TS is structured as a self-report questionnaire designed for rapid administration immediately following the completion of a learning activity, problem-solving phase, or multimedia instructional module.

Technical Specifications

• Test Type: Original Psychometric Inventory / Questionnaire.
• Construct Measured: Learning-Related Cognitive Demand (Mental Load and Mental Effort).
• Available Formats:
  • Long Version: 8 items total (4 items for Mental Load, 4 items for Mental Effort).
  • Short Version: 6 items total (3 items for Mental Load, 3 items for Mental Effort).
• Response Scale: 9-point Likert-type scale ranging from 1 (“strongly disagree”) to 9 (“strongly agree”). The 9-point format maintains structural continuity with classic cognitive load instruments (e.g., Paas, 1992) while offering improved sensitivity to subtle changes in cognitive state compared to 5-point or 7-point scales.
• Available Languages: Officially published and psychometrically evaluated in English, German, and Chinese (Mandarin), with extended development protocols supporting Dutch, Spanish, and French implementations.
• Target Population: Adults (18 years of age and older) participating in digital learning, classroom instruction, laboratory experiments, or computer-based workplace training.
• Administration Mode: Electronic/Digital (e.g., Qualtrics, Gorilla Experiment Builder, PsyToolkit, RedCap) or Paper-and-Pencil.
• Completion Time: Approximately 60 to 90 seconds for the 8-item long version; under 45 seconds for the 6-item short version.

Scoring Guidelines and Protocols

• Reverse-Scoring: Several items are phrased negatively to minimize acquiescence bias. Before computing composite scores, these items must be reverse-scored using the following transformation:$$\text{Score}_{\text{reversed}} = 10 – \text{Score}_{\text{original}}$$
  Specifically, in the 8-item version:
  • Mental Load: Item 3 (“Working on this activity was effortless.”) and Item 4 (“Working on this activity did not mentally strain me.”) must be reversed.
  • Mental Effort: Item 6 (“I refrained from putting effort into this activity.”) must be reversed.
• Subscale Calculation: Compute separate mean scores for each subscale:
  • Mental Load Score: Mean of the four load items (Items 1, 2, and reversed Items 3 and 4). Higher scores reflect greater task-induced cognitive demand.
  • Mental Effort Score: Mean of the four effort items (Items 5, 7, 8, and reversed Item 6). Higher scores indicate greater deliberate cognitive capacity invested.
• Short Version Scoring: In the 6-item short version, the subscales consist of three items each, scored following the same reverse-coding rules and averaged to produce subscale scores.
• Interpretation Matrix: Rather than collapsing the subscales into a single overall score, researchers are encouraged to analyze the two scores jointly within a 2×2 cognitive allocation matrix:
  • High Load / High Effort: Demanding task with engaged, persistent learners (optimal for productive struggle and schema expansion).
  • High Load / Low Effort: Overwhelming task leading to cognitive overload, disengagement, or task avoidance.
  • Low Load / High Effort: Accessible task tackled with deep interest and thorough cognitive processing.
  • Low Load / Low Effort: Trivial or under-stimulating task resulting in minimal cognitive processing.

Permissions & Fee and Test Year

The Mental Effort and Load–Translingual Scale (MEL-TS) was published in 2025 by Tino Endres and colleagues. In accordance with open science practices, the scale is distributed under the terms of the Creative Commons Attribution 4.0 International License (CC BY 4.0).

• Licensing Terms: The CC BY 4.0 license permits unrestricted use, distribution, and reproduction of the instrument in any medium, provided the original authors and the primary publication are appropriately credited.
• Fee: The MEL-TS is completely free of charge. No royalties, permissions fees, or administrative costs are required for academic, scientific, or educational applications.
• Commercial Use: Non-commercial research and educational use are fully permitted without prior approval. Organizations wishing to integrate the MEL-TS into proprietary commercial platforms or software should consult the license terms and contact the corresponding author, Dr. Tino Endres ([email protected]), for institutional guidance.
• Open Science Repository: Supplementary translation matrices, item variations, and replication datasets are available through the publisher’s online platform and associated Open Science Framework (OSF) repositories linked in the original article.

References

Below are primary references documenting the scale’s empirical foundation and broader theoretical context:

• Endres, T., Bender, L., Sepp, S., Zhang, S., David, L., Trypke, M., Lieck, D., Désiron, J. C., Bohm, J., Weissgerber, S., Castro-Alonso, J. C., & Paas, F. (2025). Developing the Mental Effort and Load–Translingual Scale (MEL-TS) as a foundation for translingual research in self-regulated learning. Educational Psychology Review, 37(1), Article 5. https://doi.org/10.1007/s10648-024-09978-8
• Kahneman, D. (1973). Attention and effort. Prentice-Hall.
• Paas, F. (1992). Training strategies for attaining transfer of problem-solving skill in statistics: A cognitive-load approach. Journal of Educational Psychology, 84(4), 429–434. https://doi.org/10.1037/0022-0663.84.4.429
• Paas, F., & van Merriënboer, J. J. G. (1994). Variability of worked examples and transfer of geometrical problem-solving skills: A cognitive-load approach. Journal of Educational Psychology, 86(1), 122–133. https://doi.org/10.1037/0022-0663.86.1.122
• Sweller, J. (1988). Cognitive load during problem solving: Effects on learning. Cognitive Science, 12(2), 257–285. https://doi.org/10.1207/s15516709cog1202_4
• Sweller, J., Ayres, P., & Kalyuga, S. (2011). Cognitive load theory. Springer Science & Business Media. https://doi.org/10.1007/978-1-4419-8126-4
• Winne, P. H., & Hadwin, A. F. (1998). Studying as self-regulated learning. In D. J. Hacker, J. Dunlosky, & A. C. Graesser (Eds.), Metacognition in educational theory and practice (pp. 277–304). Lawrence Erlbaum Associates.
• Zimmerman, B. J. (2000). Attaining self-regulation: A social cognitive perspective. In M. Boekaerts, P. R. Pintrich, & M. Zeidner (Eds.), Handbook of self-regulation (pp. 13–39). Academic Press. https://doi.org/10.1016/B978-012109890-2/50031-7

Items of the Scale