Palestinian Students Embrace AI in Education but Face a Knowledge-Practice Gap

University students in Palestine know a great deal about artificial intelligence, hold strongly positive views of its role in learning, and report using it regularly in their studies, according to a new cross-sectional study published in Discover Education. Yet the research also reveals a striking contradiction at the heart of this enthusiasm: the students’ theoretical knowledge and favorable attitudes have not fully translated into deep, structured, or institutionally supported practice, leaving a visible gap between what students understand about AI and what they are actually able to do with it in their academic work.

The study, conducted by Firas Asmar of the Department of Education at Al-Ummah University College in Jerusalem, surveyed 236 students drawn from a college population of 550, a sample size calculated with a 95 percent confidence level and a 5 percent margin of error. Participants were recruited through an online questionnaire distributed via Facebook and WhatsApp between March 25 and May 25, 2025, and represented five academic departments: Education, Hebrew Language, Business Administration, Engineering and Technology, and Graphic Design. Students ranged from first-year to fourth-year levels and came from cities, villages, and refugee camps, giving the survey a breadth of demographic coverage unusual for a single-institution study in the region.

Methodologically, the research followed a descriptive, correlational, quantitative, non-experimental, cross-sectional design. The instrument was a structured questionnaire organized into four sections: sociodemographic information, followed by three ten-item domains measuring knowledge, practice, and attitudes toward AI in education, each scored on a five-point Likert scale from strongly disagree to strongly agree. Content validity was established by a panel of three expert judges, including two professors specializing in information technology and artificial intelligence and a specialist in statistics and research methodology. Each item was rated for relevance on a four-point scale, and items achieving an Item-level Content Validity Index of at least 0.78 were retained, yielding an average Scale-level CVI of 0.94, a figure the author characterizes as indicating excellent content validity. A pilot study with 36 students, who were excluded from the final sample, produced a Cronbach’s alpha of 0.862, confirming good internal consistency.

The statistical analysis, performed in SPSS version 26, was rigorous for a survey of this scale. Kolmogorov-Smirnov tests with Lilliefors correction and visual inspection of Q-Q plots confirmed that knowledge, attitude, and practice scores did not significantly deviate from normality, justifying the use of parametric tests. Independent samples t-tests and one-way ANOVA compared mean scores across demographic groups, while Pearson correlation coefficients examined the relationships among the three core variables. Domain means were transformed to a 0-10 scale using the formula Transformed score = [(Mean − 1) / 4] × 10, with a cutoff of 6.00 used to categorize scores descriptively as good or appropriate.

The headline results are encouraging on the surface. Transformed mean scores reached 7.20 for knowledge, 7.48 for practice, and 7.76 for attitude, all comfortably above the 6.00 threshold. Most participants agreed or strongly agreed with statements about understanding AI concepts and expert systems, and most reported that AI facilitates access to information, improves communication, supports cognitive and research skills, and helps with big data analysis. Students recognized AI’s potential to enrich learning experiences and foster more effective, interactive, and personalized educational environments, findings the author links to the widespread accessibility of tools such as ChatGPT across academic disciplines.

Beneath the averages, however, the demographic breakdown tells a more nuanced story. Male students scored significantly higher than female students on knowledge (mean 7.58 versus 7.12, p = 0.028), though the author cautions that the small male subsample of 36 students limits interpretation, and no significant gender differences emerged for attitude or practice. Academic level was a significant factor across all three domains: first- and second-year students outperformed third- and fourth-year students in knowledge (p < 0.001), attitude (p = 0.002), and practice (p = 0.044), a pattern suggesting that younger cohorts arrive with greater digital fluency and reliance on technology.

Discipline and place of residence also shaped the results in unexpected ways. Differences across specialties were significant only for practice (p = 0.008), with Engineering and Technology students achieving the highest mean practice score of 7.85 and Graphic Design students the lowest at 6.57. Perhaps most striking, students living in refugee camps reported the highest mean practice score of all residential groups at 7.98, compared with 7.55 for rural students and 7.31 for urban students. The author speculates that this may reflect increased reliance on digital tools when physical educational resources are scarce, though he notes the relationship warrants further investigation.

The correlational analysis provided some of the study’s most theoretically meaningful findings. Knowledge correlated moderately with attitude (r = 0.396, p < 0.001) and with practice (r = 0.344, p < 0.001), while attitude showed the strongest association with practice (r = 0.553, p < 0.001). This pattern aligns closely with the Technology Acceptance Model, which holds that perceived usefulness and attitude strongly predict technology adoption behavior. In practical terms, the results suggest that building positive attitudes, through success experiences, role modeling, and exposure to well-designed AI applications, may be the most powerful lever for increasing actual use, more so than simply transmitting factual knowledge about AI.

The attitude domain itself revealed the barriers students perceive. Participants generally agreed that inadequate infrastructure, insufficient training, poor institutional planning, and limited AI awareness within educational organizations significantly hinder AI integration. The author situates these findings within a broader regional literature: Arab studies have repeatedly identified weak technical infrastructure, shortages of qualified personnel, high application costs, and limited institutional readiness as fundamental obstacles to AI adoption in higher education. In Palestine, these challenges are compounded by political and economic constraints, and the pace of AI development has simply outstripped the capacity of local educational systems to adapt. The author is explicit that the gap between knowledge and practice is not merely a matter of individual motivation but reflects systemic failures, including the absence of hands-on training, lack of institutional support, and deficient infrastructure.

The study concludes with a clear set of recommendations for Palestinian higher education: invest in technological infrastructure, integrate AI-related competencies into curricula, provide specialized and continuous training for students and academic staff, and foster a culture of responsible and ethical AI use. The author acknowledges limitations, including the online administration of the survey, the use of convenience sampling from a single institution, which means the sample should not be considered statistically representative of all Palestinian university students, and the exclusion of many colleges in the West Bank and Gaza Strip due to the current political situation. Even so, as one of the first field analyses of AI knowledge, attitudes, and practices among Palestinian university students, the study fills a significant regional gap and offers a data-driven roadmap for institutions seeking to ensure that students can use AI effectively, ethically, and productively as the technology reshapes global education.

Subject of Research: University students' knowledge, attitudes, and practices regarding artificial intelligence in education in Palestine

Article Title: Assessing university students’ knowledge, attitudes, and practices towards artificial intelligence in education in Palestine

Article References: Asmar, F. (2026). Assessing university students’ knowledge, attitudes, and practices towards artificial intelligence in education in Palestine. Discover Education, 5(1), Article 956. https://doi.org/10.1007/s44217-026-02195-1

Image Credits: AI Generated

DOI: 10.1007/s44217-026-02195-1

Keywords: artificial intelligence, higher education, Palestine, KAP study, technology acceptance, students, ChatGPT, educational technology, cross-sectional study, AI literacy, institutional barriers, Discover Education