Mixed-methods study links social norms, attitudes and learning-oriented smartphone use with university students’ self-reported scientific literacy

A mixed-methods study of 370 university students in China found that subjective norms, attitudes toward smartphone-based science learning, and actual learning-oriented smartphone use were positively associated with self-reported scientific literacy. The cross-sectional findings do not establish that smartphone use causes higher scientific literacy.

The study applied the Theory of Planned Behavior to examine smartphone-related attitudes, subjective norms, perceived behavioral control, usage intention, actual use and scientific literacy. In this context, subjective norms referred to perceived social pressure surrounding smartphone use for science learning, while perceived behavioral control reflected students’ perceived ease or difficulty of learning science with a smartphone.

Researchers first conducted semi-structured interviews with 20 university students. Their accounts consistently represented attitudes, subjective norms, perceived behavioral control, usage intention and actual use, which informed the quantitative phase. Participants described both possible benefits and drawbacks: astronomy applications could broaden access to scientific information, while smartphones could also distract from learning.

The quantitative phase used an online questionnaire distributed through university and class WeChat groups at four universities in Shandong, Jiangxi and Guangxi provinces. Of 406 questionnaires received, 36 were excluded as rushed, careless or identical, leaving 370 valid responses. Participants were recruited through non-probability convenience sampling and voluntary participation, so the sample was not a random representation of Chinese university students. The final group included 101 men and 269 women across four undergraduate years.

The investigators analyzed the survey with partial least squares structural equation modeling. Reliability coefficients for the measured constructs ranged from 0.822 to 0.922, and the model met the reported checks for convergent and discriminant validity. Scientific literacy was assessed with a self-report scale rather than an objective test of scientific knowledge or performance. After measurement-model refinement, nine of the original 38 scientific-literacy indicators were retained.

Subjective norms had the largest direct positive association with self-reported scientific literacy, with a standardized path coefficient of 0.305. Actual use of smartphones for learning scientific knowledge followed at 0.286, while attitudes toward smartphone-based science learning had a coefficient of 0.179; all three associations were statistically significant. Together, the modeled predictors explained 49.1% of the variance in self-reported scientific literacy.

Subjective norms and attitudes also had significant indirect associations with scientific literacy through pathways involving usage intention and actual smartphone use. The estimated indirect coefficient was 0.126 for subjective norms and 0.116 for attitudes, indicating partial mediation in the specified model. These results describe statistical associations rather than causal boosts. Perceived behavioral control was positively associated with usage intention, with a coefficient of 0.134, but its indirect association with scientific literacy was not statistically significant. No direct path from perceived behavioral control to scientific literacy was estimated.

An Importance-Performance Matrix Analysis ranked subjective norms first in predictive importance, followed by attitudes, actual smartphone use, usage intention and perceived behavioral control. Subjective norms and actual use were placed in a quadrant combining relatively high importance with lower performance, leading the authors to identify them as areas for attention. Based on the results, the authors recommend supportive social environments, purposeful smartphone-based learning strategies, richer digital science resources and instructional designs that encourage support for smartphone learning from peers, teachers and parents.

The findings require several cautions. All major constructs were measured by the same self-report questionnaire at one time point, so temporal ordering cannot be established and common-method bias cannot be completely ruled out. Convenience sampling and self-selection limit generalizability, while the small fourth-year subgroup prevented reliable analysis of differences by academic year. The reduced scientific-literacy scale also leaned more heavily toward the nature-of-science dimension than the science, technology and society dimension, making it a narrower representation of the original concept. The authors say broader representative samples, multi-source or temporally separated data, further validation of the scale’s dimensional structure, and longitudinal or experimental designs could address these limitations.

Subject of Research: Associations between smartphone-use factors and self-reported scientific literacy among Chinese university students

Article Title: Impact of smartphone usage on university students' scientific literacy: a mixed-methods approach with importance-performance analysis

Article References: Luan, J., & Xin, S. (2026). Impact of smartphone usage on university students' scientific literacy: a mixed-methods approach with importance-performance analysis. Frontiers in Psychology, 17, Article 1888350. https://doi.org/10.3389/fpsyg.2026.1888350

Image Credits: AI Generated

DOI: 10.3389/fpsyg.2026.1888350

Keywords: scientific literacy, smartphone learning, university students, subjective norms, student attitudes, mixed-methods research, structural equation modeling, cross-sectional study, China, mobile learning, Impact, smartphone