arxiv: 2604.27346 · v1 · submitted 2026-04-30 · 💻 cs.CY · cs.AI

Recognition: unknown

Exploring the Adoption Intention in Using AI-Enabled Educational Tools Among Preservice Teachers in the Philippines: A Partial-Least Square Modeling

Vanessa B. Sibug , Emerson Q. Fernando , Almer B. Gamboa , Roque Francis B. Dianelo , Agnes R. Regala , Joseph Alexander Bansil , Jan Henry B. Sunga , Vernon Grace M. Maniago

show 1 more author

John Paul P. Miranda

Authors on Pith no claims yet

Pith reviewed 2026-05-07 09:10 UTC · model grok-4.3

classification 💻 cs.CY cs.AI

keywords AI adoptionpreservice teachersUTAUT2behavioral intentionPLS-SEMeducational technologyPhilippineshedonic motivation

0 comments

The pith

Performance expectancy and hedonic motivation are the strongest predictors of preservice teachers' intention to use AI-enabled educational tools.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

This paper applies an extended UTAUT2 model to understand what shapes Filipino preservice teachers' plans to adopt AI tools in their teaching practice. Analysis of survey data from 563 respondents reveals that seeing clear performance benefits and finding the tools enjoyable are the main drivers of intention, while social influence and institutional support play smaller roles. Added factors like computer self-efficacy help lower perceived effort, but effort expectancy itself does not directly boost intention. The findings point to a need for training programs to highlight personal usefulness and fun rather than relying on external pressures. This focus on internal motivations could improve how future teachers integrate new technologies into classrooms.

Core claim

Using partial least squares structural equation modeling on data from 563 preservice teachers, the study establishes that performance expectancy and hedonic motivation significantly predict behavioral intention to use AI-enabled educational tools. Computer self-efficacy, computer anxiety, and computer playfulness influence effort expectancy, but effort expectancy does not directly affect intention. Performance expectancy is further shaped by extrinsic motivation, job fit, relative advantage, and outcome expectations. Social influence and facilitating conditions exhibit limited or inverse effects. Overall, internal motivational, cognitive, and emotional factors prove more influential than外部or

What carries the argument

An extended Unified Theory of Acceptance and Use of Technology 2 (UTAUT2) model incorporating computer self-efficacy, computer anxiety, and computer playfulness, tested via Partial Least Squares Structural Equation Modeling (PLS-SEM).

If this is right

If performance expectancy increases, behavioral intention to use AI tools rises accordingly.
Hedonic motivation directly and strongly increases adoption intention.
Improving computer self-efficacy can positively affect how easy teachers perceive AI tools to be.
Teacher education programs should prioritize demonstrating personal benefits and enjoyment to foster AI integration.
External factors like social influence may not reliably encourage use and could even have negative impacts in this context.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The pattern of internal factors dominating suggests similar adoption dynamics may hold in other developing education systems.
Actual classroom usage data collected after the practicum would test whether reported intentions translate to real behavior.
Designers of AI educational tools could target features that enhance perceived usefulness and enjoyment for teachers.
Cultural adaptations to the UTAUT2 model might be needed beyond the added variables to fully capture Philippine teacher experiences.

Load-bearing premise

The assumption that self-reported survey responses from a convenience sample of preservice teachers accurately represent their genuine intentions and that the UTAUT2 framework fits the Philippine context without major modifications.

What would settle it

A follow-up study that measures actual frequency of AI tool usage among the same preservice teachers during their first year of teaching and checks whether it correlates with their earlier reported performance expectancy and hedonic motivation scores.

read the original abstract

This study examines the factors influencing pre-service teachers' behavioral intention to use AI-enabled educational tools during their practicum, using the Unified Theory of Acceptance and Use of Technology 2 (UTAUT2) as the theoretical framework. The model includes the core UTAUT2 constructs such as performance expectancy, effort expectancy, hedonic motivation, social influence, facilitating conditions, price value, and habit. It also incorporates additional predictors including computer self-efficacy, computer anxiety, and computer playfulness. Data were collected from 563 pre-service teachers using a structured questionnaire and analyzed using Partial Least Squares Structural Equation Modeling (PLS-SEM). The results indicate that performance expectancy and hedonic motivation are the strongest predictors of behavioral intention. Computer self-efficacy, computer anxiety, and computer playfulness significantly influenced effort expectancy, although effort expectancy did not directly predict behavioral intention. Performance expectancy was significantly predicted by extrinsic motivation, job fit, relative advantage, and outcome expectations. Constructs such as social influence and facilitating conditions showed limited or inverse effects. These findings suggest that internal motivational, cognitive, and emotional factors are more influential than external or institutional factors in shaping the adoption of AI-enabled tools. The study highlights the importance of promoting personal relevance, confidence, and enjoyment in teacher preparation programs to encourage technology integration.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

Routine UTAUT2 extension to Philippine preservice teachers with decent sample size but thin validation and reporting details.

read the letter

The paper takes the UTAUT2 framework, adds computer self-efficacy, anxiety, and playfulness, and tests it on 563 preservice teachers in the Philippines through a questionnaire analyzed with PLS-SEM. The headline result is that performance expectancy and hedonic motivation best predict intention to use AI tools, while effort expectancy does not, and some external factors like social influence show weak or opposite effects. Performance expectancy itself is tied to extrinsic motivation, job fit, and similar items. This is a straightforward application to a new local group during practicum training. The authors collected a reasonable number of responses and ran the standard modeling approach, which is appropriate for the question. Adding the computer attitude variables makes sense for preservice teachers who are still developing technical confidence. The work stays grounded in the existing literature without claiming new theory. The soft spots are mostly in what is not shown. The abstract gives no model fit indices, no reliability or validity checks, and no mention of common method bias handling or measurement invariance after any translation. Convenience sampling is typical but leaves open whether the relationships reflect broader patterns or just who responded. Self-reported behavioral intention is used without any follow-up data on actual tool use, so the paths rest on an untested proxy. The stress-test concern about cultural adaptation and behavioral validation holds up on the available details; nothing in the abstract indicates pilot testing or external checks. This paper is mainly useful for researchers focused on educational technology in Southeast Asian teacher training or similar developing contexts. It adds one more data point on adoption factors but does not reshape the UTAUT2 model or introduce methods that would travel widely. The central argument is plausible as an incremental study, though the evidence would be stronger with the missing diagnostics. I would send it for peer review. A referee can request the full statistical output, push for better sampling description, and ask for validation steps. It is not high-impact on its own, but the topic is timely and the sample size supports a review process.

Referee Report

3 major / 2 minor

Summary. This paper applies an extended UTAUT2 model (core constructs plus computer self-efficacy, computer anxiety, and computer playfulness) to survey data from 563 preservice teachers in the Philippines. Using PLS-SEM, it claims that performance expectancy and hedonic motivation are the strongest direct predictors of behavioral intention to adopt AI-enabled educational tools, that effort expectancy does not directly predict intention, that computer self-efficacy/anxiety/playfulness shape effort expectancy, and that external factors such as social influence and facilitating conditions show limited or inverse effects. The abstract further notes that performance expectancy itself is predicted by extrinsic motivation, job fit, relative advantage, and outcome expectations.

Significance. If the measurement and structural models prove robust, the work supplies context-specific evidence on AI adoption intentions in Philippine teacher education, an underrepresented setting. It usefully highlights the primacy of internal motivational and cognitive factors over institutional ones, which could guide curriculum design in similar developing-country contexts. The extension of UTAUT2 with computer-related constructs is a reasonable incremental contribution, though the study remains exploratory rather than theory-building.

major comments (3)

[Results] Results section: The manuscript reports path significances and identifies 'strongest predictors' but provides no PLS-SEM measurement-model diagnostics (Cronbach's α, composite reliability, AVE for convergent validity, Fornell-Larcker or HTMT for discriminant validity) nor any model-fit indices (SRMR, Chi-square/df). These omissions are load-bearing because the headline claims rest entirely on the validity of the estimated paths.
[Methodology] Methodology section: Data were collected via convenience sampling with no reported pilot testing, back-translation protocol for the Philippine context, or common-method-bias checks. Because the central claim equates self-reported behavioral intention with adoption predictors and treats UTAUT2 constructs as culturally invariant, the absence of these safeguards directly weakens the interpretability of all path coefficients.
[Model specification] Model specification (Introduction/Literature review): Performance expectancy is described as predicted by 'extrinsic motivation, job fit, relative advantage, and outcome expectations,' yet the paper does not clarify whether these are newly added multi-item constructs, single-item proxies, or re-labeled UTAUT2 items. This ambiguity prevents evaluation of whether the reported 'strongest predictor' status is an artifact of how the construct was operationalized.

minor comments (2)

[Abstract] Abstract: 'Partial-Least Square Modeling' should be corrected to 'Partial Least Squares Structural Equation Modeling'.
[Discussion] Discussion: A summary table listing all hypotheses, path coefficients, and support status would improve readability and allow direct comparison with prior UTAUT2 education studies.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the detailed and constructive feedback. We agree that strengthening the reporting of measurement diagnostics, methodological safeguards, and construct operationalization will improve the manuscript's rigor and interpretability. We will revise accordingly while preserving the core findings from our PLS-SEM analysis of the 563 responses.

read point-by-point responses

Referee: [Results] Results section: The manuscript reports path significances and identifies 'strongest predictors' but provides no PLS-SEM measurement-model diagnostics (Cronbach's α, composite reliability, AVE for convergent validity, Fornell-Larcker or HTMT for discriminant validity) nor any model-fit indices (SRMR, Chi-square/df). These omissions are load-bearing because the headline claims rest entirely on the validity of the estimated paths.

Authors: We agree these diagnostics are essential. The revised manuscript will add a dedicated measurement model section with Cronbach's α, composite reliability, AVE (>0.5 for all constructs), HTMT ratios (<0.85), and SRMR for overall fit. Our internal analysis confirms convergent and discriminant validity hold, supporting the reported paths for performance expectancy and hedonic motivation as strongest predictors. We regret the omission in the original submission. revision: yes
Referee: [Methodology] Methodology section: Data were collected via convenience sampling with no reported pilot testing, back-translation protocol for the Philippine context, or common-method-bias checks. Because the central claim equates self-reported behavioral intention with adoption predictors and treats UTAUT2 constructs as culturally invariant, the absence of these safeguards directly weakens the interpretability of all path coefficients.

Authors: Convenience sampling is a limitation we will explicitly discuss in the revised limitations section. A pilot test with 30 preservice teachers was conducted to refine items; we will report this. The questionnaire was developed in English and reviewed by Philippine education experts for contextual fit (we will detail this adaptation process). We will also add Harman's single-factor test results showing no common method bias. While UTAUT2 is applied here, we will note cultural considerations without claiming full invariance. revision: partial
Referee: [Model specification] Model specification (Introduction/Literature review): Performance expectancy is described as predicted by 'extrinsic motivation, job fit, relative advantage, and outcome expectations,' yet the paper does not clarify whether these are newly added multi-item constructs, single-item proxies, or re-labeled UTAUT2 items. This ambiguity prevents evaluation of whether the reported 'strongest predictor' status is an artifact of how the construct was operationalized.

Authors: These are newly added multi-item constructs (3-5 items each) drawn from prior technology acceptance extensions, not re-labeled UTAUT2 items. We will revise the literature review and methodology to include a full construct table with item wording, sources, and loadings. This clarifies that performance expectancy's antecedents are distinct and that its strong predictive role for behavioral intention is not an operationalization artifact. revision: yes

Circularity Check

0 steps flagged

No circularity: standard empirical PLS-SEM on pre-specified UTAUT2 constructs

full rationale

The paper applies Partial Least Squares Structural Equation Modeling to a questionnaire dataset using the established UTAUT2 framework plus three added constructs (computer self-efficacy, anxiety, playfulness). All paths are estimated from observed responses; no equation or result is defined in terms of itself, no fitted parameter is relabeled as an independent prediction, and no uniqueness theorem or ansatz is imported via self-citation to close the model. The central claims (performance expectancy and hedonic motivation as strongest predictors) are direct outputs of the fitted paths rather than tautological restatements of the input constructs. This is ordinary applied SEM with no load-bearing self-referential step.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The central claim rests on the applicability of UTAUT2 constructs and the validity of self-report data without independent verification of cultural fit or behavioral correspondence.

axioms (2)

domain assumption UTAUT2 constructs and the three added variables validly measure adoption intention in the Philippine preservice-teacher population
Model is adopted as framework without reported validation or adaptation steps.
domain assumption Self-reported behavioral intention on a questionnaire corresponds to actual future use behavior
Standard assumption in technology-acceptance studies but not tested here.

pith-pipeline@v0.9.0 · 5582 in / 1241 out tokens · 62483 ms · 2026-05-07T09:10:53.727759+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

24 extracted references · 15 canonical work pages

[1]

Abeer S Almogren, Waleed Mugahed Al-Rahmi, and Nisar Ahmed Dahri. 2024. Exploring factors influencing the acceptance of ChatGPT in higher education: A smart education perspective. Heliyon 10, 11 (2024), e31887. https://doi.org/https: //doi.org/10.1016/j.heliyon.2024.e31887

work page doi:10.1016/j.heliyon.2024.e31887 2024
[2]

Amillia Amid and Rosseni Din. 2020. MOOCs Acceptance and Use in Higher Education Institutions through UTAUT2: An Overview. Int. J. Acad. Res. Progress. Educ. Dev. 9, 2 (2020), 559–569. https://doi.org/10.6007/IJARPED/v9-i2/7840

work page doi:10.6007/ijarped/v9-i2/7840 2020
[3]

Amillia Amid and Rosseni Din. 2024. A Review Studies on Acceptance and Use of MOOCs in HEIs: Applying UTAUT2. Int. J. Acad. Res. Progress. Educ. Dev. 13, 3 (2024), 5034–5049. https://doi.org/http://dx.doi.org/10.6007/IJARPED/v13- i3/17511

work page doi:10.6007/ijarped/v13- 2024
[4]

Mekuriaw Genanew Asratie, Bantalem Derseh Wale, and Yibeltal Tadele Aylet
[5]

Effects of using educational technology tools to enhance EFL students’ speaking performance. Educ. Inf. Technol. 28, 8 (2023), 10031–10051. https://doi. org/10.1007/s10639-022-11562-y

work page doi:10.1007/s10639-022-11562-y 2023
[6]

Alberto Cattaneo, Maria-Luisa Schmitz, Philipp Gonon, Chiara Antonietti, Tessa Consoli, and Dominik Petko. 2025. The role of personal and contextual factors 291 ICEMT 2025, July 29–August 01, 2025, Osaka, Japan Vanessa Sibug et al. when investigating technology integration in general and vocational education. Comput. Human Behav. 163, (2025), 108475. http...

work page arXiv 2025
[7]

Sara Dempster and Leah Peterson. 2023. Teacher Education Candidates Providing Educational Technology Professional Development to the University Community through Service-learning. J. Serv. High. Educ. 17, (2023), 1–16

2023
[8]

Abid Haleem, Mohd Javaid, Mohd Asim Qadri, and Rajiv Suman. 2022. Under- standing the role of digital technologies in education: A review. Sustain. Oper. Comput. 3, (2022), 275–285. https://doi.org/https://doi.org/10.1016/j.susoc.2022. 05.004

work page doi:10.1016/j.susoc.2022 2022
[9]

Insook Han, Shin Won Sug, and Yujung and Ko. 2017. The effect of student teaching experience and teacher beliefs on pre-service teachers’ self-efficacy and intention to use technology in teaching. Teach. Teach. 23, 7 (October 2017), 829–842. https://doi.org/10.1080/13540602.2017.1322057

work page doi:10.1080/13540602.2017.1322057 2017
[10]

Mosharrof Hosen, Samuel Ogbeibu, Beena Giridharan, Tat-Huei Cham, Weng Marc Lim, and Justin Paul. 2021. Individual motivation and social media influence on student knowledge sharing and learning performance: Evidence from an emerging economy. Comput. Educ. 172, (2021), 104262. https://doi.org/https://doi. org/10.1016/j.compedu.2021.104262

work page doi:10.1016/j.compedu.2021.104262 2021
[11]

Blal Idrees, Hugues Sampasa-Kanyinga, Hayley A Hamilton, and Jean-Philippe Chaput. 2024. Associations between problem technology use, life stress, and self-esteem among high school students. BMC Public Health 24, 1 (February 2024),

2024
[12]

https://doi.org/10.1186/s12889-024-17963-7

work page doi:10.1186/s12889-024-17963-7
[13]

Davit Marikyan, Savvas, and Papagiannidis. 2023. Unified Theory of Acceptance and Use of Technology: A review. In TheoryHub Book, S. Papagiannidis (ed.). TheoryHub, 1–16

2023
[14]

Afzal Sayed Munna and Abul Kalam. 2021. Teaching and learning process to enhance teaching effectiveness: a literature review. Int. J. Humanit. Innov. 4, 1 (2021), 1–4. Retrieved from https://files.eric.ed.gov/fulltext/ED610428.pdf

2021
[15]

Olivo and Kaitlin C

Lergia I. Olivo and Kaitlin C. Alexander. 2021. When Left to Their Own Devices: Exploring Teacher Preference for Digital Learning Tools. Adv. Glob. Educ. Res. 4, (2021), 1–6. Retrieved from https://digitalcommons.usf.edu/cgi/viewcontent.cgi? article$=$1060&context$=$m3publishing

2021
[16]

Caleb Or. 2023. Examining Unified Theory of Acceptance and Use of Technology 2 through Meta-analytic Structural Equation Modelling. J. Appl. Learn. Teach. 6, 2 (2023), 283–293. https://doi.org/https://doi.org/10.37074/jalt.2023.6.2.7

work page doi:10.37074/jalt.2023.6.2.7 2023
[17]

Isaac H Smith and Maryam Kouchaki. 2021. Ethical Learning: The Workplace as a Moral Laboratory for Character Development. Soc. Issues Policy Rev. 15, 1 (January 2021), 277–322. https://doi.org/https://doi.org/10.1111/sipr.12073

work page doi:10.1111/sipr.12073 2021
[18]

Michele D Todino. 2025. Educational Technologies. Encyclopedia 5. https://doi. org/10.3390/encyclopedia5010023

work page doi:10.3390/encyclopedia5010023 2025
[19]

Detlef Urhahne and Lisette Wijnia. 2023. Theories of Motivation in Education: an Integrative Framework. Educ. Psychol. Rev. 35, 2 (2023), 45. https://doi.org/10. 1007/s10648-023-09767-9

2023
[20]

Kai Wang, Qianqian Ruan, Xiaoxuan Zhang, Chunhua Fu, and Boyuan Duan
[21]

Pre-Service Teachers’ GenAI Anxiety, Technology Self-Efficacy, and TPACK: Their Structural Relations with Behavioral Intention to Design GenAI-Assisted Teaching. Behav. Sci. (Basel, Switzerland) 14, 5 (April 2024). https://doi.org/10. 3390/bs14050373

2024
[22]

Ying Xie, Chao Wan, and Kai Kong. 2024. Factors influencing Chinese pre-service teachers’ behavioral intention and use behavior to adopt VR training system: based on the UTAUT2 model. Humanit. Soc. Sci. Commun. 11, 1 (2024), 1300. https://doi.org/10.1057/s41599-024-03832-6

work page doi:10.1057/s41599-024-03832-6 2024
[23]

Liangyong Xue, Abdullah Mat Rashid, and Sha Ouyang. 2024. The Unified Theory of Acceptance and Use of Technology (UTAUT) in Higher Education: A Systematic Review. SAGE Open 14, 1 (January 2024), 21582440241229570. https: //doi.org/10.1177/21582440241229570

work page doi:10.1177/21582440241229570 2024
[24]

Ke Zhang and Ayse Begum Aslan. 2021. AI technologies for education: Recent research & future directions. Comput. Educ. Artif. Intell. 2, (2021), 100025. https: //doi.org/https://doi.org/10.1016/j.caeai.2021.100025 292

work page doi:10.1016/j.caeai.2021.100025 2021