Exploring the factors influencing college students’ learning satisfaction in generative AI-supported MOOCs learning environment: a learning experience framework perspective

Probably not a substantial claim: may not pass S2 (checkable). May combine two claims.

Learning satisfaction serves as a subjective indicator for assessing the learning process and outcomes and more directly reflects learners' experiences and perceptions within GenAI-supported learning environments.

“learning satisfaction serves as a subjective indicator for assessing the learning process and outcomes. It more directly reflects learners’ experiences and perceptions within GenAI-supported learning environments”

• cites Banihashem et al., 2023 sourcePartly supported

  Section 2.6 (p. 6) measures students’ satisfaction with learning activities through self-reported ratings of interactions, feedback, and assignments, supporting its use as an indicator of their experiences in blended education. Section 2.7 (p. 6) measures perceived learning performance separately. The paper does not establish satisfaction as an indicator of learning outcomes or examine GenAI-supported environments. (LLM, conf=0.88)

  Experimental — accuracy not yet measured
• cites Belkina et al., 2025 sourceNot checked — no full text of the source
• cites Chu, 2025 sourceNot checked — no full text of the source

May combine two claims.

In online learning, GenAI enables output to provide personalized learning resources based on students' different styles of input, thus promoting group and peer-to-peer interaction, learning engagement, and effectiveness.

“In online learning, GenAI enables output to provide personalized learning resources based on students’ different styles of input, thus promoting group and peer-to-peer interaction, learning engagement, and effectiveness (Lu and Ba, 2025).”

• cites Lu and Ba, 2025 sourcePartly supported

  The Discussion reports that groups using higher-quality prompts received more personalised GAI feedback, and that the feedback facilitated social interaction and online group discussion performance. The Abstract reports improved behavioural and social engagement and knowledge-building performance among pre-service teachers. The paper describes personalised feedback, not personalised learning resources tailored to students’ different input styles, and its findings are specific to inquiry-based online discussions rather than online learning generally. (LLM, conf=0.88)

  Experimental — accuracy not yet measured

Classified as probably not a substantial claim. May combine two claims.

Inappropriate assessment, lack of support, and inappropriate teacher-student interactions can lead to lower learner satisfaction, which in turn affects student learning outcomes.

“inappropriate assessment, lack of support, and inappropriate teacher-student interactions can lead to lower learner satisfaction, which in turn affects student learning outcomes (Nguyen, 2022)”

• cites Nguyen, 2022 sourcePartly supported

  The Abstract and Discussion (section 5) link dissatisfaction with inappropriate assessment, lack of support, and insufficient teacher–student interaction. The Results (section 4.2) find a correlation between satisfaction and cognitive measures, but do not establish that lower satisfaction in turn affects learning outcomes. (LLM, conf=0.88)

  Experimental — accuracy not yet measured

Classified as probably not a substantial claim.

Interaction as an external supportive behavior from the teacher or classmates positively affects learning outcomes.

“Interaction as an external supportive behavior from the teacher or classmates positively affects learning outcomes (She et al., 2021).”

• cites She et al., 2021 sourceSupported

Classified as probably not a substantial claim. May combine two claims.

Effective interactive content reduces network loss and promotes high levels of learner engagement in learning activities, and has a positive impact on learning outcomes.

“Effective interactive content, which reduces network loss and promotes high levels of learner engagement in learning activities, has a positive impact on learning outcomes (Kim and Kim, 2021).”

• cites Kim and Kim, 2021 sourcePartly supported

  The Discussion (§4.4; page unavailable) reports that course structure positively affects academic achievement and that student-student interaction positively affects engagement. It does not report that interactive content reduces network loss. It also finds that engagement has no significant effect on academic achievement, so the claimed pathway to learning outcomes is not established. (LLM, conf=0.87)

  Experimental — accuracy not yet measured

May combine two claims.

If teacher support becomes perfunctory or duplicates GenAI feedback, students may perceive no added value and feel annoyed, negatively impacting learning satisfaction.

“However, if such support becomes perfunctory or duplicates GenAI feedback, students may perceive no added value and instead feel annoyed, negatively impacting LS (Yang et al., 2022).”

• cites Yang et al., 2022 sourceCould not verify

  The Discussion and Findings discuss autonomy-supportive teaching and student engagement, but do not address perfunctory support, duplicated GenAI feedback, annoyance, or learning satisfaction. Searches for those terms and for negative effects of teacher support found no passage supporting the attributed claim. (LLM, conf=0.76)

  Experimental — accuracy not yet measured

May combine two claims.

Learners perceive a conflict between teacher guidance and GenAI support, diminishing their expectations for teacher-student interactions, which leads to TSI negatively impacting learning satisfaction.

“Consequently, learners perceive a conflict between teacher guidance and GenAI support, diminishing their expectations for teacher-student interactions. This ultimately leads to TSI negatively impacting learning satisfaction (Yang Y. et al., 2025).”

• cites Yang Y. et al., 2025 sourceCould not verify

  The Introduction discusses controversy over replacing teacher feedback with AI, and Discussion §6.2 recommends teacher-led cooperation with AI as an auxiliary tool. Neither section reports that learners perceive a conflict, expect less teacher–student interaction, or experience lower learning satisfaction as a result. Searches for those terms found no passage addressing the claimed relationship. (LLM, conf=0.76)

  Experimental — accuracy not yet measured

• AssumptionRests on an assumption: Instructor guidance in MOOC settings often exhibits delays and focuses primarily on knowledge transmission.

May combine two claims.

The introduction of GenAI alters the form of traditional teacher-student interaction; some students may prefer to autonomously use GenAI to find answers rather than engage frequently with teachers.

“Furthermore, the introduction of GenAI alters the form of traditional TSI; some students may prefer to autonomously use GenAI to find answers rather than engage frequently with teachers (Lu and Ba, 2025).”

• cites Lu and Ba, 2025 sourcePartly supported

  The paper describes students receiving feedback through a GAI chatbot and discusses ChatGPT as a complement to human feedback (Abstract and Discussion; pages not provided). This supports a change in how students can obtain feedback, but the paper does not report that students prefer using GenAI to find answers instead of interacting frequently with teachers. (LLM, conf=0.76)

  Experimental — accuracy not yet measured

The quote only partly states this claim.

Teacher-student interaction (TSI) has a significant positive impact on learning outcomes (LO).

“TSI has a significant positive impact on LO”

The quote only partly states this claim.

In GenAI-supported MOOC learning, student-student interaction enhanced learning outcomes.

“(3) student–student interaction, although they could enhance learning outcomes, did not have a significant impact on satisfaction.”

May combine two claims.

Among the learning experience dimensions, LO had the lowest mean (4.517), followed by SSI (4.652), TSI (4.759), and LE (4.798).

“The lowest scores in the dimensions of learning experience were for learning outcomes LO (M = 4.517, SD = 0.970), SSI (M = 4.652, SD = 1.152), TSI (M = 4.759, SD = 1.025), and LE (M = 4.798, SD = 1.002).”

Classified as probably not a substantial claim.

The learning experience involves evaluating multiple dimensions including cognition, behavior, and emotion.

“The learning experience, as the source of learners’ subjective perceptions and satisfaction with their participation, involves evaluating multiple dimensions including cognition, behavior, and emotion (Li et al., 2023).”

• cites Li et al., 2023Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing).

GenAI-supported MOOCs learning is a complex learning environment.

“GenAI-supported MOOCs learning is a complex learning environment (Gutiérrez-Páez et al., 2023; Lang et al., 2025).”

• cites Gutiérrez-Páez et al., 2023 sourceNot checked — no full text of the source
• cites Lang et al., 2025 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Performance expectations and effort expectations significantly influence students' willingness to integrate ChatGPT into their learning practices.

“performance expectations and effort expectations significantly influence students’ willingness to integrate ChatGPT into their learning practices (Khlaif et al., 2024).”

• cites Khlaif et al., 2024 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Learners were ambivalent about using GenAI, relying on its efficiency but also experiencing anxiety due to a sense of "cheating".

“learners were ambivalent about using GenAI, relying on its efficiency but also experiencing anxiety due to a sense of “cheating” (Rudolph and Tan, 2023).”

• cites Rudolph and Tan, 2023Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

GenAI-supported MOOCs learning environments have the ability to implement learning analytics, recommender systems, adaptive learning, intelligent assessment, and more, and can provide students with a richer learning experience.

“GenAI-supported MOOCs learning environments have the ability to implement learning analytics, recommender systems, adaptive learning, intelligent assessment, and more, can provide students with a richer learning experience (Anderson et al., 2025).”

• cites Anderson et al., 2025 sourceNot checked — no full text of the source

May combine two claims.

GenAI creates an environment that is conducive to fostering curiosity, helping to create a strong connection between instructor and student, and fostering the learner's desire to learn based on knowledge.

“GenAI creates an environment that is conducive to fostering curiosity, helping to create a strong connection between instructor and student, and fostering the learner’s desire to learn based on knowledge (Evans et al., 2023).”

• cites Evans et al., 2023 sourceNot checked — no full text of the source

Classified as probably not a substantial claim.

Cognition does not occur in isolation but is distributed across interactions among learners, learning environments, and cultures.

“cognition does not occur in isolation. Instead, it is distributed across interactions among learners, learning environments, and cultures”

• cites Bosch et al., 2019 sourceNot checked — no full text of the source

May combine two claims.

When teachers lack experience in using GenAI to support learning or organize collaborative activities, learners feel less engaged and less motivated.

“When teachers lack experience in using GenAI to support learning or organize collaborative activities, learners feel less engaged, less motivated (Kuo et al., 2014).”

• cites Kuo et al., 2014 sourceNot checked — no full text of the source

May combine two claims.

Embedding AI learning assistants in MOOCs platforms can answer students' questions and provide personalized advice in real time.

“embedding AI learning assistants in MOOCs platforms can answer students’ questions and provide personalized advice in real time (Yılmaz and Yılmaz, 2022).”

• cites Yılmaz and Yılmaz, 2022Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

May combine two claims.

GenAI intervention may lead to emotional detachment or excessive expectations.

“indicating that GenAI intervention may lead to emotional detachment or excessive expectations”

Probably not a substantial claim: may not pass S2 (checkable). May combine two claims.

Learning environment is the foundation on which learning takes place, including the technological and resource environments.

“Learning environment is the foundation on which learning takes place, including the technological and resource environments (Mi and Wu, 2017).”

• cites Mi and Wu, 2017 sourceNot checked — no full text of the source

May combine two claims.

Teacher-student interaction is directly related to guided feedback and personalized support for teaching and learning, and is associated with learner satisfaction and learning outcomes.

“It is directly related to guided feedback and personalized support for teaching and learning, and is associated with learner satisfaction, and learning outcomes (Molinillo et al., 2018).”

• cites Molinillo et al., 2018 sourceNot checked — no full text of the source

May combine two claims.

The study collected two independent samples, Sample A and Sample B, to conduct EFA and CFA respectively.

“Therefore, we collected Samples A and B.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

All participants in Sample B completed the GenAI-supported MOOC course and fulfilled related learning tasks.

“All participants in Sample B completed the GenAI-supported MOOC course Modern Educational Technology and fulfilled related learning tasks.”

May combine two claims. The quote only partly states this claim.

The failure of peer interaction to significantly improve learning outcomes may stem from multiple factors, including inadequate form and quality of student-student interaction in MOOC environments and GenAI partially replacing student-student interaction.

“However, in the context of GenAI-supported MOOCs, the failure of peer interaction to significantly improve LO may stem from multiple factors.”

Probably not a substantial claim: may not pass S3 (load-bearing).

The sampling considered gender ratios to prevent gender-based sample sizes from affecting the generalizability of results.

“Additionally, sampling considers gender ratios to prevent gender-based sample sizes from affecting the generalizability of results.”

May combine two claims.

The generalizability of the study's findings is limited because the sample primarily consists of Chinese university students, whose learning habits and knowledge foundations may differ from students in other countries and regions.

“the generalizability of the findings is limited. The sample primarily consists of Chinese university students, whose learning habits and knowledge foundations may differ significantly from those of students in other countries and regions.”

• AssumptionRests on an assumption: Chinese university students' learning habits and knowledge foundations may differ significantly from those of students in other countries and regions.

May combine two claims.

GenAI has partially replaced certain aspects of student-student interaction; learners may prefer interacting with GenAI over peers, diminishing student-student interaction's impact and making learning outcomes more dependent on learner-GenAI interactions.

“Second, GenAI has partially replaced certain aspects of SSI. When encountering difficulties, learners may prefer interacting with GenAI over engaging with peers. This scenario diminishes the impact of SSI, making LO more dependent on learner-GenAI interactions rather than peer collaboration.”

May combine two claims.

The adapted questionnaire has high reliability and validity and can be used for subsequent model testing.

“In conclusion, the questionnaire has high reliability and validity and can be used for subsequent model testing.”

May combine two claims.

The introduction of GenAI technology did not negatively impact learner satisfaction; rather, it enhanced student satisfaction through timely Q&A support and personalized guidance.

“This indicates that the introduction of GenAI technology did not negatively impact learner satisfaction; rather, it further enhanced student satisfaction through timely Q&A support and personalized guidance.”

May combine two claims.

CFA showed significant loadings for all variables (p < 0.001, α > 0.7); all variables satisfied AVE > 0.5 and CR > 0.7 except LE, which had AVE = 0.472 and was acceptable under the relaxed criteria (0.4 < AVE < 0.5 and α > 0.6).

“The internal consistency of the questionnaire was checked and the results showed significant CFA loadings for all variables (p < 0.001, α > 0.7). AVE and CR values were used to discriminate the validity of the dataset, and all variables satisfied AVE > 0.5 and CR > 0.7 except LE. however, when 0.4 < AVE < 0.5 and α > 0.6, the results were acceptable as shown in Table 5.”

Classified as probably not a substantial claim. May combine two claims.

Students' perceptions of learning satisfaction, learning environment, teacher-student interaction, and peer-student interaction may change over time.

“students’ perceptions of learning satisfaction, learning environment, teacher-student interaction, and peer-student interaction may change over time (Sidik and Syafar, 2020).”

• cites Sidik and Syafar, 2020 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S4 (based).

Sample A from Shaanxi Normal University can represent the typical characteristics of Chinese college students to a certain extent.

“its student population can represent the typical characteristics of Chinese college students to a certain extent”

May combine two claims.

The learning experience section had a KMO of 0.875 (p < 0.001) and the satisfaction section had KMO = 0.774 and alpha = 0.786, indicating acceptable questionnaire reliability.

“The reliability of the questionnaire was checked and the results showed that the KMO value for the learning experience section was 0.875 > 0.8, p < 0.001, and for the four topics of satisfaction the KMO = 0.774, alpha = 0.786, and the results indicated that the questionnaire reliability was acceptable.”

May combine two claims.

Improving the quality of the learning environment, teacher-student interaction, and student-student interaction leads to better learning outcomes and subsequently better learning satisfaction.

“This suggests that once the quality of the learning environment improves, teacher-student interaction, and student–student interaction improves, students obtain better learning outcomes and subsequently have better learning satisfaction.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Learning activities with higher interactivity will increase motivation, improve learning outcomes and satisfaction compared to less interactive learning environments.

“Learning activities with higher interactivity will increase motivation, improve learning outcomes and satisfaction compared to less interactive learning environments (Liaw and Huang, 2013).”

• cites Liaw and Huang, 2013 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

When conducting self-directed learning, learning peers' interactions mainly take place through topic discussions and dialogs, and effective interactions can improve learning outcomes and develop higher-order thinking skills.

“When conducting self-directed learning, learning peers’ interactions mainly take place through topic discussions and dialogs, and effective interactions can improve learning outcomes and develop higher-order thinking skills (Lee and Choi, 2017).”

• cites Lee and Choi, 2017 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

There are differences in the needs of students in different majors for GenAI, with art students preferring GenAI to stimulate creativity and provide a variety of outputs, while technology students are more concerned with accuracy and efficiency.

“There are differences in the needs of students in different majors for GenAI, with art students preferring GenAI to stimulate creativity and provide a variety of outputs, while technology students are more concerned with accuracy and efficiency (Kim et al., 2025b).”

• cites Kim et al., 2025bNot matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

May combine two claims.

In GenAI-supported MOOC learning environments, when learners enhance their learning outcomes through GenAI tools, their satisfaction with course arrangements and tools increases; if learning outcomes are unsatisfactory, their satisfaction decreases.

“In GenAI-supported MOOC learning environments, when learners enhance their LO through GenAI tools, their satisfaction with course arrangements and the tools used also increases accordingly. Conversely, if LO are unsatisfactory, their satisfaction decreases.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

The sampling ensured that each academic achievement stratum contained at least 10 samples, and the total sample size for each major stratum was no less than 30.

“This ensured that each academic achievement stratum contained at least 10 samples, and the total sample size for each major stratum was no less than 30.”

Probably not a substantial claim: may not pass S3 (load-bearing).

The study employed stratified random sampling to maximize representativeness and randomness.

“To maximize representativeness and randomness, we employed stratified random sampling”

May combine two claims.

MOOC learners are numerous and possess diverse educational backgrounds; interactions are constrained by professional backgrounds and foundational knowledge, making student-student interaction superficial when knowledge disparities are excessive.

“MOOC learners are numerous and possess diverse educational backgrounds. Interactions among learners are often constrained by their professional backgrounds and foundational knowledge. When learners at different levels exhibit excessive knowledge disparities, interactions may become inefficient, rendering SSI largely superficial (Johnson, 1981).”

• cites Johnson, 1981 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

All participants used GenAI-supported learning to varying degrees in their daily studies.

“All of them used GenAI-supported learning to varying degrees in their daily studies”

Classified as probably not a substantial claim. May combine two claims.

Learning can be significantly enhanced once supportive teacher interactions are increased and a more dynamic and responsive learning environment is created.

“Learning can be significantly enhanced once supportive teacher interactions are increased and a more dynamic and responsive learning environment is created (Pahi et al., 2024).”

• cites Pahi et al., 2024 sourceNot checked — no full text of the source

May combine two claims.

All factor loadings in the EFA were greater than 0.50, ranging from 0.637 to 0.847, indicating acceptable measurement of the constructs.

“Further EFA analysis using the rotated component matrix showed that all factor loadings of the study variables were greater than the recommended threshold of 0.50, with factor loadings ranging from 0.637 to 0.847, indicating that the items used to measure the various constructs were acceptable (Table 4).”