Table of Contents
Abstract
The ChatGPT Usage Scale is a newly developed psychometric instrument designed to quantify how Postgraduate Students interact with Artificial Intelligence, specifically ChatGPT, in their academic endeavors. As generative AI rapidly permeates Higher Education, educators and administrators face a pressing need to understand its impact on learning and Academic Integrity. This tool provides a structured, empirically validated method to measure these interactions, moving beyond anecdotal observations to rigorous quantitative assessment.
The development of this scale addresses a critical gap in the current Educational Psychology and technology literature. While previous frameworks like the Technology Acceptance Model have explored general AI adoption, this specific instrument drills down into the nuanced ways graduate Students utilize ChatGPT for complex academic tasks. By capturing multidimensional usage patterns—from drafting manuscripts to brainstorming and establishing trust in AI outputs—the scale offers researchers a granular view of modern academic workflows.
Ultimately, the significance of this instrument lies in its potential to inform evidence-based educational policies and curriculum design. By accurately measuring how Students rely on AI, institutions can better navigate the ethical and pedagogical challenges of the digital age. The scale serves as a foundational tool for future empirical studies investigating the intersection of Artificial Intelligence, cognitive load, and academic performance.
📊 Psychometric Scorecard
15
Multidimensional
0.85
0.85
0.917
📍 Egypt
Authors
Purpose
The primary objective of the ChatGPT Usage Scale is to provide a standardized, psychometrically sound measure of how advanced Students integrate AI conversational agents into their scholarly routines. Prior to its development, the field lacked a validated tool specifically tailored to the postgraduate population, a group whose academic demands—such as extensive literature reviews, complex data synthesis, and rigorous academic writing—differ significantly from those of undergraduates. Researchers and educational policymakers needed a reliable way to assess not just whether Students use AI, but exactly how they use it and to what extent they depend on its outputs.
For researchers and clinicians working in Educational Psychology, this scale is invaluable. It allows for the systematic investigation of the cognitive and behavioral shifts occurring as a result of AI integration in academia. By quantifying specific usage behaviors, scholars can explore correlations between AI reliance and variables such as academic anxiety, self-efficacy, and critical thinking skills, ultimately guiding the development of targeted interventions and ethical guidelines for AI use in Higher Education.
Construct
The psychological construct measured by this scale encompasses the behavioral and cognitive dimensions of AI utilization in an academic context. Grounded in theoretical frameworks such as the Technology Acceptance Model (TAM) and Cognitive Load Theory (CLT), the construct conceptualizes ChatGPT usage not merely as a binary adoption metric, but as a multifaceted interaction. It captures the intrinsic and extrinsic motivations driving Students to offload cognitive tasks to an AI, reflecting a complex interplay between perceived utility, ease of use, and academic demands.
This overarching construct is operationalized through three distinct but interrelated sub-domains. The first captures the use of AI as a direct aid in the mechanics of academic writing, such as drafting and paraphrasing. The second broadens the scope to general academic support, encompassing tasks like brainstorming and organizing study materials. The final dimension delves into the psychological aspects of reliance and trust, evaluating the degree to which Students depend on the AI's accuracy and view it as a reliable academic partner. Together, these dimensions form a comprehensive higher-order construct of AI academic engagement.
Validity
The validation process for this instrument was rigorous, establishing strong evidence for its structural and convergent validity. The researchers demonstrated convergent validity by calculating the Average Variance Extracted (AVE), which yielded a value of 0.664. This comfortably exceeds the standard psychometric threshold of 0.50, indicating that the latent constructs explain a substantial portion of the variance in their respective indicators. Such robust convergent validity assures researchers that the items within each subscale are genuinely measuring the same underlying concept.
Additionally, the structural validity of the scale was confirmed through advanced modeling techniques. The items demonstrated strong, statistically significant standardized factor loadings ranging from 0.434 to 0.728, ensuring that each question meaningfully contributes to its designated factor. The presence of a strong second-order overarching factor further validates the theoretical assumption that the three distinct dimensions collectively represent a unified construct of overall ChatGPT usage.
Reliability
The instrument demonstrates excellent internal consistency, making it a highly dependable tool for researchers. Psychometric evaluation revealed a Cronbach’s alpha of 0.848 for the overall scale, which is well above the widely accepted benchmark of 0.70 for research instruments. This high alpha value indicates that the 15 items are highly cohesive and reliably measure the same underlying construct without excessive redundancy.
To further solidify the reliability evidence, the developers also calculated McDonald’s omega, which yielded a nearly identical value of 0.849, alongside a composite reliability score of 0.855. The use of McDonald's omega is particularly noteworthy as it does not rely on the strict tau-equivalence assumptions required by Cronbach's alpha, providing a more robust estimate of reliability. These metrics collectively confirm that the scale will produce stable and consistent results across different samples of Postgraduate Students.
Factor Analysis
The structural integrity of the scale was established through a two-step factor analytic approach, beginning with an Exploratory Factor Analysis (EFA) using principal component analysis with Varimax rotation. This initial exploration successfully reduced the original 39-item pool down to a lean 15-item structure, revealing three distinct factors that together account for approximately 49% of the total variance. This data-driven reduction ensures that only the most psychometrically sound items were retained, minimizing participant fatigue while maximizing measurement precision.
Following the EFA, a Confirmatory Factor Analysis (CFA) was executed to test the fit of this three-factor model. The results demonstrated an acceptable fit to the data, with a Comparative Fit Index (CFI) of 0.917 and a Tucker-Lewis Index (TLI) of 0.900, both meeting the conventional threshold of 0.90. Furthermore, the Root Mean Square Error of Approximation (RMSEA) was an excellent 0.060, indicating a strong model fit. These indices collectively confirm that the theoretical three-dimensional structure accurately reflects the observed data.
Subscales
| Subscale | Items | Description |
|---|---|---|
| Academic Writing Aid | 34, 38, 13, 5, 18, 23 | Measures the extent to which Students use ChatGPT for direct writing assistance, including paraphrasing, generating ideas, drafting, and developing counterarguments. |
| Academic Task Support | 8, 30, 14, 3, 10 | Assesses the use of ChatGPT for broader academic activities such as overcoming writer's block, organizing thoughts, finding information, and creating study materials. |
| Reliance and Trust | 12, 2, 16, 22 | Evaluates Students' psychological dependence on the AI, their trust in its accuracy, and their use of the tool for feedback and brainstorming. |
Instrument
| Test Type | Self-report questionnaire |
| Format | 15 items, 5-point Likert scale (1 = strongly disagree to 5 = strongly agree) |
| Scoring | Items are summed or averaged to create subscale and total scores. |
| Language | English |
| Population | College students, Students |
| Age Group | 21-48 years |
| Administration | Online survey |
Scoring & Interpretation Guidelines
| Scoring Instructions | All items are rated on a 5-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). |
ChatGPT Usage Scale Items
Items are currently not available
The individual items of this scale are not publicly available. Researchers interested in using this instrument should contact the original authors directly to request the scale materials.
Sample
The validation sample consisted of 443 Postgraduate Students (194 males, 249 females) recruited from two Egyptian universities (Kafr el-Sheikh University and Al-Azhar University). Participants had an average age of 27.4 years (SD = 4.8) and included Students enrolled in postgraduate diploma (64.6%), master's (35.0%), and doctoral (23.0%) programs.
Cite This Paper
Mohamed Nemt-Allah, Waleed Khalifa, Mahmoud Badawy, Yasser Elbably, Ashraf Ibrahim (2024). ChatGPT Usage Scale. BMC psychology. https://doi.org/10.1186/s40359-024-01983-4
References
26 references
- Zawacki-Richter O, Marín V, Bond M, Gouverneur F. Systematic review of research on Artificial Intelligence applications in Higher Education–where are the educators? Int J Educ Technol High Educ. 2019;16(1):1–27. 🔗 https://doi.org/10.1186/s41239-019-0171-0
- Floridi L, Chiriatti M. GPT-3: its nature, scope, limits, and consequences. Minds Mach. 2020;30:681–94. 🔗 https://doi.org/10.1007/s11023-020-09548-1
- Aydin Ö, Karaarslan E. Is ChatGPT leading generative Ai? What is beyond expectations? Acad Platf J Eng Smart Syst. 2023;11(3):118–34. 🔗 https://doi.org/10.21541/apjess.1293702
- Hartley K, Hayak M, Ko U. Artificial Intelligence supporting Independent Student Learning: an evaluative case study of ChatGPT and Learning to Code. Educ Sci. 2024;14(2):120. 🔗 https://doi.org/10.3390/educsci14020120
- Elbably Y, Nemt-allah M. Grand challenges for ChatGPT usage in education: psychological theories, perspectives and opportunities. Psychol Res Educ Soc Sci. 2024;5(2):31–6.
- Bin-Nashwan SA, Sadallah M, Bouteraa M. Use of ChatGPT in academia: Academic Integrity hangs in the balance. Technol Soc. 2023;75:102370. 🔗 https://doi.org/10.1016/j.techsoc.2023.102370
- İpek Z, Gözüm A, Papadakis S, Kallogiannakis M. Educational Applications of the ChatGPT AI system: a systematic Review Research. Educ Process Int J. 2023;12(3):26–55. 🔗 https://doi.org/10.22521/edupij.2023.123.2
- Henderson M, Finger G, Selwyn N. What’s used and what’s useful? Exploring digital technology use (s) among taught Postgraduate Students. Act Learn High Educ. 2016;17(3):235–47. 🔗 https://doi.org/10.1177/1469787416654798
- Sain ZH, Hebebci MT. ChatGPT and beyond: The rise of AI assistants and chatbots in Higher Education. In: Curle SM, Hebebci MT, editors. Proceedings of International Conference on Academic Studies in Technology and Education 2023. ARSTE Organization; 2023. pp. 1–12.
- Schön EM, Neumann M, Hofmann-Stölting C, Baeza-Yates R, Rauschenberger M. How are AI assistants changing Higher Education? Front Comput Sci. 2023;5:1208550. 🔗 https://doi.org/10.3389/fcomp.2023.1208550
- Wang T, Díaz DV, Brown C, Chen Y. Exploring the Role of AI Assistants in Computer Science Education: Methods, Implications, and Instructor Perspectives. In: 2023 IEEE Symposium on Visual Languages and Human-Centric Computing (VL/HCC). IEEE; 2023. pp. 92–102. 🔗 https://doi.org/10.1109/VL-HCC57772.2023.00018
- Davis F. Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Q. 1989;13(3):319–40. 🔗 https://doi.org/10.2307/249008
- Venkatesh V, Davis F. A theoretical extension of the Technology Acceptance model: four longitudinal field studies. Manag Sci. 2000;46(2):186–204. 🔗 https://doi.org/10.1287/mnsc.46.2.186.11926
- Deci E, Ryan R. The general causality orientations scale: self-determination in personality. J Res Pers. 1985;19(2):109–34. 🔗 https://doi.org/10.1016/0092-6566(85)90023-6
- Ng J, Ntoumanis N, Thøgersen-Ntoumani C, Deci E, Ryan R, Duda J, et al. Self-determination theory applied to health contexts: a meta-analysis. Perspect Psychol Sci. 2012;7(4):325–40. 🔗 https://doi.org/10.1177/1745691612447309
- Sweller J. Cognitive load during problem solving: effects on learning. Cogn Sci. 1988;12(2):257–85. 🔗 https://doi.org/10.1207/s15516709cog1202_4
- Chen O, Kalyuga S, Sweller J. The worked example effect, the generation effect, and element interactivity. J Educ Psychol. 2015;107(3):689–704. 🔗 https://doi.org/10.1037/edu0000018
- Sallam M, Salim N, Barakat M, Al-Mahzoum K, Ala’a B, Malaeb D, et al. Assessing health Students’ attitudes and usage of ChatGPT in Jordan: validation study. JMIR Med Educ. 2023;9(1):e48254. 🔗 https://doi.org/10.2196/48254
- Abdaljaleel M, Barakat M, Alsanafi M, Salim N, Abazid H, Malaeb D et al. Factors influencing attitudes of university Students towards ChatGPT and its usage: a multi-national study validating the TAME-ChatGPT survey instrument. Preprints 2023:2023090541. 🔗 https://doi.org/10.20944/preprints202309.1541.v1
- Worthington RL, Whittaker TA. Scale Development Research: a content analysis and recommendations for best practices. Couns Psychol. 2006;34(6):806–38. 🔗 https://doi.org/10.1177/0011000006288127
- Hooper D, Coughlan J, Mullen M. Structural equation modelling: guidelines for determining Model Fit. Electron J Bus Res Methods. 2008;6(1):53–60.
- Hu L, Bentler PM. Cutoff criteria for fit indexes in covariance structure analysis: conventional criteria versus new alternatives. Struct Equ Model Multidiscip J. 1999;6(1):1–55. 🔗 https://doi.org/10.1080/10705519909540118
- Nunnally J, Bernstein I. Psychometric theory. 3rd ed. McGraw-Hill; 1994.
- Hair JF, Black WC, Babin BJ, Anderson RE. Multivariate data analysis. 7th ed. Pearson Education Limited; 2014.
- Fornell C, Larcker DF. Evaluating Structural equation models with unobservable variables and measurement error. J Mark Res. 1981;18(1):39–50. 🔗 https://doi.org/10.1177/002224378101800104
- Huallpa J. Exploring the ethical considerations of using Chat GPT in university education. Period Eng Nat Sci. 2023;11(4):105–15.
Cite this article
Mohammed looti (2026). ChatGPT Usage Scale. PSYCHOLOGICAL SCALES. Retrieved from https://scales.arabpsychology.com/s/chatgpt-usage-scale/
Mohammed looti. "ChatGPT Usage Scale." PSYCHOLOGICAL SCALES, 14 Aug. 2026, https://scales.arabpsychology.com/s/chatgpt-usage-scale/.
Mohammed looti. "ChatGPT Usage Scale." PSYCHOLOGICAL SCALES, 2026. https://scales.arabpsychology.com/s/chatgpt-usage-scale/.
Mohammed looti (2026) 'ChatGPT Usage Scale', PSYCHOLOGICAL SCALES. Available at: https://scales.arabpsychology.com/s/chatgpt-usage-scale/.
[1] Mohammed looti, "ChatGPT Usage Scale," PSYCHOLOGICAL SCALES, vol. X, no. Y, ص Z-Z, August, 2026.
Mohammed looti. ChatGPT Usage Scale. PSYCHOLOGICAL SCALES. 2026;vol(issue):pages.