Drivers of Mobile Learning App Usage–Measurement Model Inventory

1. Abstract

The Drivers of Mobile Learning App Usage—Measurement Model Inventory (DMLU-MM), developed by Yeh, Wang, Wang, and Liao (2023), represents a psychometric instrument engineered to investigate the structural, dispositional, and motivational determinants governing mobile learning (m-learning) application acceptance. Designed to address the rapid evolution of handheld educational technology and digital pedagogy, the instrument integrates three primary theoretical domains: personality psychology, individual technology readiness, and contemporary motivational paradigms. Specifically, the measurement model synthesizes operationalizations derived from John, Donahue, and Kentle’s (1991) Big Five Inventory (BFI; assessing Extraversion, Conscientiousness, Openness to Experience, Agreeableness, and Neuroticism, rephrasing reverse-scored indicators positively), Carducci’s (2009) abbreviated forced-choice internal-external Locus of Control scale adapted from Rotter (1966), Lin, Chou, Wang, and Liang’s (2016) multi-dimensional Mobile Learning Readiness construct (encompassing m-learning self-efficacy, optimism, and self-directed learning), Davis, Bagozzi, and Warshaw’s (1992) conceptualization of intrinsic and extrinsic motivation (operationalized through perceived enjoyment and perceived usefulness), and Wang, Wu, and Wang’s (2009) behavioral intention metrics.

Comprising 48 initial candidate items across ten focal constructs, the instrument employs a dual-format scaling strategy: 38 items utilize a standardized 7-point Likert scale, while the 10-item locus of control dimension employs a binary forced-choice response format yielding an index from 0 to 10. Psychometric validation conducted with adult respondents across young adulthood, thirties, and middle-age cohorts demonstrated high internal consistency and construct validity. Cronbach’s alpha coefficients across reflective multi-item constructs ranged from 0.87 to 0.95, and composite reliability (CR) values ranged from 0.90 to 0.97. Confirmatory structural and measurement model evaluations established robust convergent validity, with all retained indicators yielding factor loadings exceeding 0.70 and average variance extracted (AVE) values spanning 0.59 to 0.91. Discriminant validity was empirically corroborated through both cross-loading inspections and the Fornell-Larcker criterion. The DMLU-MM provides educational researchers, organizational development specialists, and instructional software engineers with an empirically rigorous diagnostic tool to evaluate how learner dispositions and perceived utility drive engagement in ubiquitous digital learning ecosystems.

2. Keywords

Drivers of Mobile Learning App Usage, Big Five Personality Model, M-Learning App Acceptance, Mobile Learning Readiness, Intrinsic Motivation, Extrinsic Motivation, Locus of Control, Technology Acceptance Model, Self-Efficacy, Behavioral Intention, Educational Psychometrics, Digital Learning Environments

3. Authors

The Drivers of Mobile Learning App Usage—Measurement Model was developed and validated by an interdisciplinary team of researchers specializing in international business, information systems, and instructional technology management:

• Ching-Hsuan Yeh — Department of International Business, Feng Chia University, Taichung, Taiwan. ORCID: 0000-0003-1308-3602.
• Yi-Shun Wang (Corresponding Author) — Department of Information Management, National Changhua University of Education, Changhua, Taiwan. Email: [email protected].
• Yu-Min Wang — Department of Information Management, National Chi Nan University, Nantou, Taiwan. ORCID: 0000-0002-2096-6801.
• Ting-Jun Liao — Department of Information Management, National Chung Hsing University, Taichung, Taiwan.

4. Purpose

The primary purpose of the Drivers of Mobile Learning App Usage—Measurement Model Inventory (DMLU-MM) is to provide an empirical instrument capable of measuring the complex web of individual differences, psychological readiness, and cognitive-motivational appraisals that govern how adult learners adopt, utilize, and sustain engagement with mobile educational software. As smartphone and tablet technologies have proliferated globally, educational institutions and corporate enterprises have increasingly transitioned instructional content onto mobile platforms. However, mere access to ubiquitous mobile technology does not guarantee meaningful usage or cognitive adoption. Existing assessment frameworks historically overemphasized macro-level technological affordances—such as system speed, interface aesthetics, or platform architecture—while often neglecting how stable personality traits, self-regulatory learning agency, and psychological attribution structures interface with functional utility and hedonic enjoyment.

From a theoretical perspective, the DMLU-MM addresses the critical need for an integrative measurement framework that bridges differential psychology and information systems acceptance models. While traditional models like the Technology Acceptance Model (TAM) or the Unified Theory of Acceptance and Use of Technology (UTAUT) primarily emphasize instrumental beliefs such as perceived usefulness and perceived ease of use, they often treat individual user characteristics as unmodeled exogenous error or superficial demographic covariates. The DMLU-MM explicitly conceptualizes personality traits (the Big Five dimensions and Locus of Control) and Mobile Learning Readiness as foundational, distal antecedents that actively shape proximal motivational beliefs (intrinsic and extrinsic motivation), which subsequently drive behavioral intention to use mobile learning applications.

In applied academic, corporate, and clinical training contexts, the DMLU-MM fulfills multiple diagnostic and prescriptive functions:

• Educational Diagnostics and Learner Profiling: Academic institutions can utilize the inventory to evaluate incoming student populations, identifying cohorts exhibiting low mobile learning self-efficacy, high neuroticism, or pronounced external locus of control who may require targeted onboarding, scaffolding, or instructional support before enrolling in app-mediated courses.
• Human Resource and Professional Development: Corporate talent management departments can administer the scale to assess employee readiness for self-directed mobile micro-learning programs, enabling tailored learning interventions that align with employee motivational profiles and dispositional traits.
• User Experience (UX) and Instructional Software Design: Software architects and instructional designers can deploy the measurement model during beta-testing phases to evaluate whether gamified features, social networking elements, or modular content architectures differentially stimulate intrinsic enjoyment versus extrinsic productivity across varying user personality segments.
• Psychometric and Longitudinal Research: Academic researchers can utilize the validated factor structure to conduct structural equation modeling (SEM), latent growth curve analysis, or cross-cultural invariance investigations into technology-mediated lifelong learning.

5. Psychological Construct

The DMLU-MM is a multidimensional measurement inventory that operationalizes ten focal psychological constructs. These constructs are structured into three distinct conceptual tiers: dispositional personality traits, technology-specific readiness, and proximal cognitive-motivational perceptions.

5.1 Big Five Personality Traits

Grounded in the lexical and structural traditions of personality psychology, the Big Five dimensions capture enduring patterns of cognition, affect, and behavior that shape how individuals approach novel technological environments:

• Extraversion: Characterizes an individual’s sociability, assertiveness, energy levels, and tendency to seek positive emotional stimulation. In mobile learning contexts, highly extraverted individuals often gravitate toward collaborative, socially interactive, and community-driven mobile learning applications, actively engaging in forums, peer reviews, and interactive synchronous features.
• Conscientiousness: Reflects self-discipline, organization, goal-directed behavior, diligence, and persistence. Highly conscientious learners utilize mobile learning apps systematically, relying on calendar integrations, task checklists, and self-monitoring tools to structure independent study habits.
• Openness to Experience: Encompasses intellectual curiosity, aesthetic sensitivity, cognitive flexibility, and a willingness to explore novel ideas and methodologies. Within m-learning, high openness manifests as an enthusiasm for innovative digital interfaces, exploratory problem-solving, and novel learning technologies.
• Agreeableness: Denotes prosocial orientation, trust, empathy, cooperativeness, and modesty. Agreeable users thrive in mobile learning environments that encourage reciprocal peer assistance, constructive community discussions, and collaborative knowledge construction.
• Neuroticism: Captures emotional instability, vulnerability to stress, anxiety, and a tendency to experience negative affect. High neuroticism in mobile learning contexts is frequently associated with technology-related anxiety, apprehension regarding software failure or errors, and cognitive overload during complex, unstructured tasks.

5.2 Locus of Control

Rooted in social learning theory, Locus of Control represents a generalized expectancy regarding the degree to which life outcomes are governed by internal personal agency versus external situational forces:

• Internal Locus of Control: The conviction that educational achievements, skill mastery, and academic performance stem directly from personal effort, discipline, and strategic behavior. Internal learners view mobile learning apps as instrumental tools that amplify their personal agency.
• External Locus of Control: The belief that life events, academic success, and failures are primarily determined by external forces such as luck, fate, chance, task difficulty, or systemic authority. Individuals with an external locus of control may exhibit lower proactive engagement with self-directed mobile software, viewing technological outcomes as outside their personal control.

5.3 Mobile Learning Readiness

Mobile Learning Readiness is a domain-specific, tripartite psychological construct assessing an individual’s perceived capability and psychological preparedness to succeed in mobile-mediated educational spaces:

• M-Learning Self-Efficacy: An individual’s judgment of their technical and cognitive capability to operate mobile hardware, navigate instructional applications, troubleshoot minor usability challenges, and successfully execute learning tasks via mobile devices.
• Optimism: A positive, forward-looking psychological disposition characterized by the belief that mobile learning systems enhance educational efficiency, provide temporal and spatial flexibility, and enrich the overall learning experience.
• Self-Directed Learning: The proactive capacity to set autonomous educational goals, plan learning schedules, exercise cognitive self-regulation, and maintain self-initiated momentum without continuous external supervision.

5.4 Motivation and Acceptance Constructs

The proximal cognitive appraisal constructs capture the dynamic motivational drivers and immediate behavioral commitments of the learner:

• Intrinsic Motivation (Perceived Enjoyment): The extent to which utilizing a mobile learning app is anticipated to be inherently interesting, pleasant, engaging, and enjoyable, independent of external instrumental performance outcomes.
• Extrinsic Motivation (Perceived Usefulness): The subjective assessment that interacting with the mobile learning app will directly elevate learning productivity, enhance academic performance, improve task efficiency, and facilitate the attainment of tangible academic or professional objectives.
• Intention to Use M-Learning Apps: The explicit behavioral commitment, conscious planning, and subjective probability that an individual will adopt and continuously utilize mobile learning applications for educational purposes in the near future.

6. Theoretical Framework

The architecture of the DMLU-MM is anchored in the theoretical integration of trait theory, social cognitive theory, and self-determination theory, unified under the conceptual umbrella of advanced information systems acceptance modeling.

Historically, educational technology adoption was predominantly analyzed through the lens of Fred Davis’s (1989) Technology Acceptance Model (TAM). Derived from the Theory of Reasoned Action (Fishbein & Ajzen, 1975), TAM posited that actual system usage is governed by behavioral intention, which is directly determined by two cognitive beliefs: Perceived Usefulness (PU) and Perceived Ease of Use (PEOU). In 1992, Davis, Bagozzi, and Warshaw expanded this paradigm by incorporating motivational psychology, establishing that user acceptance is jointly driven by extrinsic motivation (PU) and intrinsic motivation (conceptualized as Perceived Enjoyment). While TAM and its subsequent extensions provided robust explanatory power regarding technological characteristics, they maintained a mechanistic view of user cognition, treating the learner largely as a uniform, rational decision-maker.

To overcome this limitation, the DMLU-MM embeds TAM’s motivational core within Julian Rotter’s (1954, 1966) Social Learning Theory and Albert Bandura’s (1986, 1997) Social Cognitive Theory. Bandura’s principle of triadic reciprocal determinism posits that human functioning is an ongoing interaction between cognitive and biological personal factors, environmental events, and behavioral actions. Within this paradigm, personal agency is operationalized through self-efficacy—the conviction in one’s capacity to execute designated performances. In mobile learning environments, where physical classrooms and immediate instructor oversight are absent, self-efficacy and self-directed learning readiness become decisive regulators of persistence. Concurrently, Rotter’s locus of control construct establishes whether a learner attributes academic reinforcement to internal competence or external luck, fundamentally altering the perceived utility and personal relevance of self-directed mobile software.

Furthermore, the integration of John, Donahue, and Kentle’s (1991) Big Five taxonomy acknowledges that technological appraisals do not occur in a psychological vacuum. Personality traits function as enduring cognitive and affective filters through which technological environments are perceived. For instance, learners high in neuroticism interpret interface complexities as threatening stressors, suppressing perceived enjoyment and amplifying perceived difficulty, whereas individuals high in openness to experience perceive novel features as stimulating, catalyzing intrinsic motivation. By synthesizing Costa and McCrae’s (1992) structural personality insights with Deci and Ryan’s (1985, 2000) Self-Determination Theory, the DMLU-MM constructs a comprehensive nomological network wherein distal personality traits and technological readiness dictate proximal intrinsic and extrinsic motivational states, which in turn crystallize into concrete behavioral adoption intentions.

7. Validity

The psychometric validation of the DMLU-MM was conducted through rigorous structural equation modeling and measurement model evaluations, utilizing data collected from a diverse adult general population cohort spanning young adulthood (18–29 years), thirties (30–39 years), and middle age (40–64 years).

7.1 Convergent Validity

Convergent validity evaluates the extent to which indicator items of a specific construct share a high proportion of common variance. In the validation study by Yeh et al. (2023), convergent validity was evaluated via indicator factor loadings (outer loadings) and Average Variance Extracted (AVE) values in accordance with guidelines established by Hair, Hult, Ringle, and Sarstedt (2017):

• Indicator Reliability: All retained reflective indicators exhibited standardized outer factor loadings exceeding the recommended 0.70 threshold. During the initial measurement model evaluation, items that failed to meet this threshold were systematically pruned. Specifically, 1 item from Extraversion (EXT_6), 1 item from Conscientiousness (CON_3), 4 items from Openness to Experience (OTE_3, OTE_6, OTE_7, OTE_9, OTE_10), 2 items from Agreeableness (AGR_1, AGR_5), 3 items from Neuroticism (NEU_3, NEU_4, NEU_7), and 1 item from Mobile Learning Readiness (MLR_6) were eliminated. All remaining items demonstrated statistically significant, high indicator loadings.
• Average Variance Extracted (AVE): The AVE reflects the grand mean value of the squared loadings of the indicators associated with the construct. The AVE values for all multi-item constructs surpassed the established benchmark of 0.50, demonstrating that the latent constructs explained substantially more than half of their indicators’ variance. Specific empirical AVE values include: Extraversion (0.73), Conscientiousness (0.59), Openness to Experience (0.61), Agreeableness (0.59), Neuroticism (0.67), Mobile Learning Readiness (0.71), Intrinsic Motivation (0.83), Extrinsic Motivation (0.87), and Intention to Use M-Learning Apps (0.91).

7.2 Discriminant Validity

Discriminant validity assesses whether constructs within the measurement model are empirically unique and capture phenomena not represented by other constructs:

• Cross-Loading Inspection: Analysis of indicator cross-loadings demonstrated that each indicator loaded at its highest magnitude on its assigned latent construct, with cross-loadings on all non-target latent constructs falling substantially lower, confirming indicator specificity.
• Fornell-Larcker Criterion: In strict alignment with the Fornell-Larcker (1981) testing methodology, the square root of each construct’s AVE was compared against its bivariate correlation coefficients with every other construct in the measurement model. For all constructs, the square root of the AVE was demonstrably higher than the highest construct correlation, verifying that each construct shared more variance with its own indicators than with any other latent variable in the model.

7.3 Construct and Predictive Validity

Construct validity is further substantiated by the theoretical alignment of structural path relationships. The empirical findings demonstrated that intrinsic motivation (perceived enjoyment) and extrinsic motivation (perceived usefulness) positively predicted behavioral intention to use mobile learning applications, while mobile learning readiness and personality dimensions functioned as significant antecedents of these motivational appraisals, demonstrating strong predictive validity.

8. Reliability

The reliability of the DMLU-MM was evaluated using dual metrics of internal consistency: Cronbach’s alpha (α) and composite reliability (CR). While Cronbach’s alpha assumes tau-equivalence (equal factor loadings across all items), composite reliability accounts for differential indicator outer loadings, offering a more robust estimation of scale precision within structural equation modeling environments.

Empirical analyses confirmed high internal consistency reliability across all nine reflective multi-item constructs, comfortably exceeding the standard academic thresholds of 0.70 for exploratory research and 0.80 for established measurement scales:

• Extraversion: Cronbach’s α = 0.94; Composite Reliability (CR) = 0.95. Demonstrates exceptional internal consistency among items capturing talkative, communicative, and lively behaviors.
• Conscientiousness: Cronbach’s α = 0.90; Composite Reliability (CR) = 0.92. Reflects high stability in capturing diligence, organization, and task perseverance.
• Openness to Experience: Cronbach’s α = 0.87; Composite Reliability (CR) = 0.90. Confirms consistent assessment of intellectual curiosity, imagination, and conceptual reflection.
• Agreeableness: Cronbach’s α = 0.89; Composite Reliability (CR) = 0.91. Indicates reliable measurement of prosocial warmth, kindness, and cooperative orientation.
• Neuroticism: Cronbach’s α = 0.89; Composite Reliability (CR) = 0.91. Demonstrates robust measurement of emotional reactivity, nervousness, and depressive affect.
• Mobile Learning Readiness: Cronbach’s α = 0.90; Composite Reliability (CR) = 0.93. Demonstrates high structural reliability across technical self-efficacy, learning freedom, and academic initiative.
• Intrinsic Motivation: Cronbach’s α = 0.89; Composite Reliability (CR) = 0.93. Reflects tight indicator coherence measuring fun, pleasantness, and enjoyment.
• Extrinsic Motivation: Cronbach’s α = 0.95; Composite Reliability (CR) = 0.96. Indicates near-perfect internal reliability in measuring performance gains, productivity, and perceived usefulness.
• Intention to Use M-Learning Apps: Cronbach’s α = 0.95; Composite Reliability (CR) = 0.97. Establishes exceptional consistency across future behavioral planning and intention indicators.

For the Locus of Control dimension, which uses a 10-item forced-choice pairing adapted from Carducci (2009) and Rotter (1966), scoring follows an additive index (0–10) representing external orientation rather than a continuous reflective latent variable; consequently, traditional internal consistency metrics were not computed for this index in the structural model, consistent with methodological conventions for single-index or formative-type structural representations (Hair et al., 2017).

9. Factor Analysis

The structural and dimensional integrity of the DMLU-MM was rigorously examined via confirmatory factor analysis (CFA) within a reflective structural equation modeling framework. All ten constructs were specified as reflective latent entities, where items serve as manifestations of their respective underlying latent variables.

The initial measurement model incorporated 48 total items. To ensure indicator reliability and dimensional purity, factor loadings (outer loadings) were systematically scrutinized against the standard 0.70 benchmark. During the preliminary confirmatory factor iteration, a subset of items exhibited outer loadings below 0.70, suggesting excessive indicator measurement error or cross-dimensional contamination. To optimize model fit and construct validity, the following 12 items were formally pruned from the measurement model:

• Extraversion: EXT_6 (“Has an assertive personality”) was removed.
• Conscientiousness: CON_3 (“Is a reliable worker”) was removed.
• Openness to Experience: OTE_3 (“Is ingenious, a deep thinker”), OTE_6 (“Values artistic, aesthetic experiences”), OTE_7 (“Prefers work that is flexible”), OTE_9 (“Has many artistic interests”), and OTE_10 (“Is sophisticated in art, music, or literature”) were eliminated.
• Agreeableness: AGR_1 (“Tends to find merit with others”) and AGR_5 (“Is generally trusting”) were removed.
• Neuroticism: NEU_3 (“Can be tense”), NEU_4 (“Worries a lot”), and NEU_7 (“Fails to be calm in tense situations”) were eliminated.
• Mobile Learning Readiness: MLR_6 (“I manage time well”) was removed.

Following item elimination, the subsequent confirmatory factor analysis revealed that all remaining indicators exhibited factor loadings well above the 0.70 threshold, with outer loadings spanning between 0.72 and 0.96. The structural model demonstrated outstanding overall fit, confirming that the ten-dimensional architecture accurately reflects the empirical covariance matrix. The factor analysis confirmed that personality traits, technology-specific readiness, and motivational constructs operate as distinct structural entities, allowing for the precise estimation of direct, indirect, and mediated pathways driving mobile learning app adoption.

10. Instrument / Measurement Tool

The DMLU-MM is structured as a comprehensive, modular questionnaire designed for direct paper-and-pencil or online administration. Below are the operational specifications of the measurement tool:

• Test Type: Multi-construct self-report psychometric inventory / technology acceptance measurement model.
• Target Population: Adult learners, undergraduate and graduate students, corporate trainees, and general mobile technology users aged 18 years and older (spanning young adulthood, thirties, and middle-age cohorts).
• Administration Format: Self-administered questionnaire (digital web survey, mobile learning portal, or paper-and-pencil).
• Total Item Pool: 48 original items (36 items retained across reflective subscales following factor refinement; 10 forced-choice pairs for Locus of Control).
• Response Scale:
  • Likert-type Constructs (Sections 1, 3, 4, 5, 6): Scored using a 7-point Likert scale (typically ranging from 1 = Strongly Disagree to 7 = Strongly Agree).
  • Locus of Control (Section 2): Scored using a forced-choice binary format across 10 paired statements (A vs. B / External vs. Internal).
• Subscales & Structural Composition:
  • Extraversion: 8 original items (7 retained in final CFA).
  • Conscientiousness: 9 original items (8 retained in final CFA).
  • Openness to Experience: 10 original items (5 retained in final CFA).
  • Agreeableness: 9 original items (7 retained in final CFA).
  • Neuroticism: 8 original items (5 retained in final CFA).
  • Locus of Control: 10 forced-choice paired items.
  • Mobile Learning Readiness: 6 original items (5 retained in final CFA).
  • Intrinsic Motivation: 3 items.
  • Extrinsic Motivation: 4 items.
  • Intention to Use M-Learning Apps: 3 items.
• Scoring and Index Calculation:
  • Likert Constructs: Individual subscale scores are computed by calculating the arithmetic mean of the retained items within each dimension. Higher scores denote stronger expression of the measured construct (e.g., higher technological readiness, higher perceived enjoyment, or stronger intention to use apps). All Big Five items originally phrased as negative in John et al. (1991) were reworded positively in this inventory, eliminating the need for reverse-scoring Likert items.
  • Locus of Control Index: Respondents select the statement from each pair that most accurately reflects their belief. External control statements (designated ‘E.’) are assigned a score of 1, whereas internal control statements (designated ‘I.’) are assigned a score of 0. Summing the 10 items yields an overall score ranging from 0 to 10. Higher composite scores reflect greater external locus of control (attributing events to luck, fate, or external factors), whereas lower scores indicate an internal locus of control (attributing outcomes to personal agency and effort).

11. Permissions & Fee and Test Year

• Publication Year: 2023 (published online in Interactive Learning Environments, 2021).
• Commercial Status: Non-commercial research instrument.
• Fee: Free for academic, scientific, and educational research purposes.
• Permissions and Licensing: The measurement model was published under academic copyright in Interactive Learning Environments (Taylor & Francis). Academic researchers wishing to utilize, adapt, or translate the DMLU-MM for non-commercial scholarly investigations may do so with appropriate citation of the original authors (Yeh et al., 2023). For commercial exploitation, corporate licensing, or digital redistribution, permission must be formally requested from Taylor & Francis or the corresponding author (Dr. Yi-Shun Wang, [email protected]).

12. References

• Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Prentice-Hall.
• Bandura, A. (1997). Self-efficacy: The exercise of control. W. H. Freeman.
• Carducci, B. J. (2009). The psychology of personality: Viewpoints, research, and applications (2nd ed.). Wiley-Blackwell.
• Costa, P. T., & McCrae, R. R. (1992). Revised NEO Personality Inventory (NEO-PI-R) and NEO Five-Factor Inventory (NEO-FFI) professional manual. Psychological Assessment Resources.
• Davis, F. D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly, 13(3), 319–340. https://doi.org/10.2307/249008
• Davis, F. D., Bagozzi, R. P., & Warshaw, P. R. (1992). Extrinsic and intrinsic motivation to use computers in the workplace. Journal of Applied Social Psychology, 22(14), 1111–1132. https://doi.org/10.1111/j.1559-1816.1992.tb00945.x
• Deci, E. L., & Ryan, R. M. (1985). Intrinsic motivation and self-determination in human behavior. Plenum Press. https://doi.org/10.1007/978-1-4899-2271-7
• Deci, E. L., & Ryan, R. M. (2000). The “what” and “why” of goal pursuits: Human needs and the self-determination of behavior. Psychological Inquiry, 11(4), 227–268. https://doi.org/10.1207/S15327965PLI1104_01
• Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention and behavior: An introduction to theory and research. Addison-Wesley.
• Fornell, C., & Larcker, D. F. (1981). Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research, 18(1), 39–50. https://doi.org/10.1177/002224378101800104
• Hair, J. F., Hult, G. T. M., Ringle, C. M., & Sarstedt, M. (2017). A primer on partial least squares structural equation modeling (PLS-SEM) (2nd ed.). SAGE Publications.
• John, O. P., Donahue, E. M., & Kentle, R. L. (1991). The Big Five Inventory—Versions 4a and 54. Institute of Personality and Social Research, University of California, Berkeley.
• Lin, H.-C., Chou, C.-Y., Wang, Y.-M., & Liang, J.-C. (2016). Reading a computer-simulated physics experiment: The role of reading engagement and mobile learning readiness. Interactive Learning Environments, 24(5), 999–1018. https://doi.org/10.1080/10494820.2014.961486
• Rotter, J. B. (1954). Social learning and clinical psychology. Prentice-Hall. https://doi.org/10.1037/11382-000
• Rotter, J. B. (1966). Generalized expectancies for internal versus external control of reinforcement. Psychological Monographs: General and Applied, 80(1), 1–28. https://doi.org/10.1037/h0092976
• Wang, Y.-S., Wu, M.-C., & Wang, H.-Y. (2009). Investigating the determinants and age and gender differences in the acceptance of mobile learning. British Journal of Educational Technology, 40(1), 92–118. https://doi.org/10.1111/j.1467-8535.2007.00809.x
• Yeh, C.-H., Wang, Y.-S., Wang, Y.-M., & Liao, T.-J. (2023). Drivers of mobile learning app usage: An integrated perspective of personality, readiness, and motivation. Interactive Learning Environments, 31(6), 3577–3594. https://doi.org/10.1080/10494820.2021.1937658

13. Items of the Scale