Mobile phone addiction as a mediator in the association between alexithymia and self-directed learning among Chinese nursing students: a structural equation modeling analysis

Abstract

Background

Self-directed learning ability is an essential competency for nursing undergraduates to develop professional skills and adapt to evolving healthcare needs. However, the psychological and behavioral factors affecting self-directed learning remain understudied.

Objective

The aim of this study is to explore the relationship between alexithymia, mobile phone addiction, and self-directed learning among nursing undergraduates.

Methods

This study involved 234 nursing undergraduates enrolled at a medical college located in southern China. The Scale of Competencies of Autonomous Learning of Nursing Undergraduates (SCALNU), the Mobile Phone Addiction Index Scale (MPAI) and the Toronto Alexithymia Scale (TAS-20) were employed to assess self-directed learning ability, alexithymia, and mobile phone addiction respectively. Correlation analyses were employed to examine pairwise relationships among the three variables. Subsequently, multiple linear regression analysis was used to identify significant influence factors of self-directed learning. Then a structural equation modeling approach was utilized to examine the mediating effect of mobile phone addiction in the relationship between alexithymia and self-directed learning.

Results

The mean scores for the SCALNU were 90.19 ± 10.10. Correlation analysis revealed a significant negative correlation between higher SCALNU scores and lower scores on the MPAI (r=-0.294, p < 0.01) and the TAS-20 (r=-0.383, p < 0.01). A multiple linear regression analysis showed that grade, only-child status, experience as class officials in high school, alexithymia, and mobile phone addiction were associated with SDL (all p < 0.01). The structural equation model analysis with a relatively good fit (χ²/df = 2.228, RMSEA = 0.073, CFI = 0.905, GFI = 0.917, and SRMR = 0.066) indicated that alexithymia has both a direct effect on SDL (β=-0.305, p = 0.014) and an indirect effect through mobile phone addiction on SDL (β=-0.120, p = 0.007), with a mediation effect accounting for 28.2% of the total effect.

Conclusion

Mobile phone addiction significantly mediates the relationship between alexithymia and self-directed learning ability among nursing undergraduates, highlighting its potential as a modifiable target for interventions to enhance self-directed learning.

Introduction

Nursing is a hands-on field inherently linked to the safety and well-being of individuals [1]. The rapid iteration of medical knowledge makes traditional educational models inadequate for the dynamic needs of the healthcare industry. It highlights the urgency of self-directed learning (SDL) abilities for keeping pace with the latest medical advancements and addressing academic challenges [2]. SDL is characterized by an individual’s proactive identification of learning needs, setting of learning objectives, selection and implementation of appropriate strategies, and self-assessment of learning outcomes, with minimal or no guidance [3]. Self-directed learners exhibit excellent self-control abilities in both cognitive and behavioral strategies, which are reflected in their active engagement in learning activities and superior academic achievements [4]. Furthermore, these learners demonstrate exceptional competency in resource management and adaptive help-seeking behaviors when confronted with academic challenges [5]. As a cornerstone of lifelong learning, SDL empowers nursing students to independently acquire vital knowledge and skills, cultivate professional self-efficacy, continuously prepare them for professional skills, and construct sustainable development capabilities to address technological transformations in the healthcare industry [5–7]. However, studies in China reveal that nurses generally exhibit only moderate proficiency in SDL [8], with many viewing learning as an onerous task beyond their professional duties [9]. This evidence highlights the crucial need to assess influential factors while placing particular emphasis on cultivating SDL among student nurses.

Alexithymia is often manifested as a multidimensional impairment in recognizing, understanding, and describing emotions [10]. Individuals with alexithymia exhibit significant difficulty distinguishing between physiological symptoms of emotional arousal and the underlying emotions themselves, and they tend to have a reduced capacity for imagination and a predisposition towards externally oriented thinking [11]. A cross-sectional study of 1950 medical students in China found that 15.7% of participants were considered alexithymic [12]. This deficit may produce rigid and inflexibility thinking patterns [11], exacerbate academic burnout [13], compromise learning motivation [14], impede collaborative support-seeking behaviors [15], and simultaneously undermine self-control [16]. All of these collectively constitute critical indicators of impaired SDL, as posited by Zimmerman’s Self-Regulated Learning Theory [17]. Previous studies also showed that students with alexithymia are more likely to exhibit low academic performance [13].

Mobile phone addiction is common in students with alexithymia [18–21]. Mobile phone addiction indicated an excessive reliance on these devices during daily activities, including studying, socializing, and even driving. Individuals with mobile phone addiction demonstrate a persistent inability to control their screen time [22]. Research has indicated that the prevalence of mobile phone addiction among Asian medical students is alarmingly high, reaching up to 41.93% [23]. According to Compensatory Internet Use Theory [24], individuals with alexithymia, who often experience experiential avoidance and difficulty in describing feelings [11, 25], may develop excessive or problematic internet use behaviors when their psychological needs remain unfulfilled in offline life. Mobile phones, which offer social networking, gaming, and entertainment, provide emotional avoidance mechanisms that enable temporary escape from real-life challenges [24, 26].

Mobile phone addiction may be associated with deficits in SDL. Among individuals with online game addiction, learning motivation as the intrinsic driver of SDL [14], has been found to be impaired [27]. This motivational decline may further impair critical SDL components, including knowledge exploration, goal-setting, and learning process monitoring [14]. Moreover, individuals with mobile phone addiction often exhibit immaturity in inhibitory control, self-regulation, attentional control, and time management [28]– [29], which may further erode critical SDL elements such as self-planning and self-monitoring of learning activities. In conclusion, while alexithymia, mobile phone addiction, and SDL exhibit significant intercorrelations, the complex relationship among the three still needs to be rigorously clarified.

This study aimed to: (1) investigate the SDL abilities in nursing undergraduates; (2) examine the association between alexithymia and SDL, and the potential mediating role of mobile phone addiction in this association. By doing so, we hope to highlight the importance of contemporary psychological (alexithymia) and behavioral (mobile phone addiction) challenges for SDL. A conceptual framework is depicted in Fig. 1.

Fig. 1.

The hypothetical mediation model showing alexithymia mediates self-directed learning through mobile phone addiction

Methods

Design and recruitment

This cross-sectional survey was conducted targeting nursing undergraduates at a medical university in Southern China. Due to the COVID-19 pandemic, a web-based approach was employed to minimize in-person contact. The survey was administered online through the Wenjuanxing platform (www.wjx.cn) and disseminated via the popular social media network WeChat. Voluntary participation was encouraged, with each participant accessing the survey through a unique WeChat account and IP address, ensuring that each participant could complete the questionnaire only once. Only fully completed responses were considered for this study. The research protocol was approved by the Institutional Review Board of Shantou University Medical College (SUMC-2020-05), and written informed consent was obtained from all participants before their study involvement. This study complied with the Declaration of Helsinki.

Participants

The survey was conducted over a two-week period from February 27 to March 10, 2020 by using convenience cluster sampling method. Eligibility for participation was limited to medical college students enrolled in the nursing program, proficient in reading and writing Chinese, and willing to participate. Students diagnosed with depression or sleep disorders, or those enrolled in vocational education, were excluded. The criteria for inclusion and exclusion were clearly stated in the invitation link distributed to participants. Participants were informed of their right to withdraw from the study at any time. And their personal information was anonymized and securely stored to ensure privacy. The required sample size was determined using the formula: . This calculation was based on data from a prior cross-sectional study among Chinese medical students, reporting a 15.7% prevalence of alexithymia [12]. The formula used a Z-score of 1.96 for a 95% confidence level and a margin of error (d) of 0.05. Thus, a convenience sample of 245 nursing undergraduates was successfully recruited.

Instruments

A self-administered questionnaire was used to collect detailed socio-demographic data from participants. It included questions on gender, grade, residence, only-child status, annual household income, the kind of college entrance examination and experience as class officials in high school or university. Additionally, the questionnaire also collected participants’ intentions to engage in nursing-related work, as well as whether they chose nursing as a major due to personal interest.

Currently, there is no widely established scale for assessing SDL in nursing undergraduates in academic research. The Scale of Competencies of Autonomous Learning of Nursing Undergraduates (SCALNU), developed by Chinese researchers Lin [30], was widely utilized to assess SDL among nursing undergraduates and achieved satisfactory results in China [5, 9, 31]. This scale is composed of 28 items across three subscales: self-management (10 items), information management (11 items), and learning cooperation (7 items). Participants responded to a 5-point Likert scale, where 1 indicates ‘completely inconsistent’ and 5 indicates ‘completely consistent’, with total scores ranging from 28 to 140. A higher total score indicates a greater level of independent learning ability. The scale showed strong internal consistency, with a Cronbach’s α of 0.82 in this study.

The Mobile Phone Addiction Index (MPAI) was employed to assess the degree of mobile phone addiction among participants, which was developed by Leung [32]. The Chinese version has shown robust reliability and validity among Chinese college students, with a Cronbach’s α of 0.91 [33]. MPAI comprises 17 questions across four subscales: inability to control craving, feeling anxious and lost, withdrawal or escape and productivity loss. Each item is scored on a five-point scale, with total scores ranging from 17 to 85, with higher scores reflecting greater addiction to mobile phones. The scale showed strong internal consistency, with a Cronbach’s α of 0.82 in this study.

The 20-item Toronto Alexithymia Scale (TAS-20) was translated into Chinese by Zhu [34], which is a self-reporting tool commonly used to measure alexithymia. It is designed to measure three subscale of alexithymia: (1) difficulty identifying feelings (7 items), (2) difficulty describing feelings (5 items), and (3) externally-oriented thinking (8 items). The TAS-20 utilizes a five-point rating scale, with total scores ranging from 20 to 100, where scores above 60 suggest the alexithymia and higher scores indicate greater severity. The scale has a Cronbach’s α of 0.87 [12], indicating high internal consistency.

The scale showed acceptable internal consistency, with a Cronbach’s α of 0.77 in this study.

Data analysis methods

Descriptive statistics were calculated for continuous variables (means ± standard deviations) and categorical variables (frequencies and percentages) to summarize demographic characteristics. The between-group comparison of SDL in participants with different demographic factors were examined using independent samples t-tests for binary variables and one-way ANOVA for multi-category variables. Correlation relationships were assessed via Spearman’s correlation analysis. Accordingly, those variables significantly associated with SDL in between-group comparison or correlation analysis were considered as covariables in a multiple linear regression model. The hypothesized mediation model was tested through structural equation modeling (SEM) with same covariables, employing maximum likelihood estimation. Confirmatory factor analysis (CFA) was used to validate the measurement model. We conducted 5000 bootstrap resampling iterations to estimate bias-corrected 95% confidence intervals for mediation effects. Confidence intervals were then tested for significance by examining if they crossed zero. Model fit was evaluated using multiple indices: the chi-square statistic (χ2) divided by degrees of freedom (df) (χ²/df), root mean square error of approximation (RMSEA), comparative fit index (CFI), goodness-of-fit index (GFI), and standardized root mean square residual (SRMR). Following established guidelines [35], acceptable model fit was defined by: χ²/df < 3, RMSEA < 0.08, CFI > 0.90, GFI > 0.90, and SRMR < 0.08. All statistical analyses were performed using SPSS and AMOS plugin (SPSS 26, USA). A p-value less than 0.05 was considered statistically significant.

Results

Characteristics of the participants and univariable analysis of factors associated with SDL

Demographic characteristics of the participants are presented in Table 1. Out of the 245 questionnaires, 11 were eliminated for incomplete responses, resulting in 234 valid questionnaires for analysis. 191 participants were female (81.6%), 124 were freshmen (53%), 74 were from rural areas (31.6%), and 73 were an only-child (31.2%). In the college entrance examination, 217 participants were science majors (92.7%), 160 and 72 were class officials in high school and in university, respectively (68.4% and 30.8%). 125 participants made clear that they would engage in nursing-related work (53.4%), while 94 were not sure (40.2%). 165 participants indicated their choice of the nursing major was influenced by practical considerations or other factors, not personal interest (70.5%). Table 1 also shows results of the univariable analysis of factors associated with SDL. The analysis indicates that non-freshmen, only children, higher annual household income, and those with experience as class officials in high school or in college tend to have higher SDL scores (all p < 0.05).

Table 1.

Univariable analysis of the effects of socio-demographic characteristics on the self-directed learning ability in nursing undergraduates

Variables	n (%) (n = 234)	Mean± standard deviations	t/F value	p-value
Gender			0.814	0.417
Male	43 (18.4)	91.33 ± 10.40
Female	191 (81.6)	89.94 ± 10.05
Grade			-3.073	0.002
Freshman	124 (53.0)	88.31 ± 10.52
Non-freshman	110 (47.0)	92.31 ± 9.21
Residence			0.852	0.395
City	160 (68.4)	90.58 ± 10.70
Rural	74 (31.6)	89.36 ± 10.20
Only-child status			2.879	0.004
Yes	73 (31.2)	92.97 ± 10.7
No	161 (68.8)	88.93 ± 9.90
Annual household income (RMB x 103)			2.741	0.044
< 10	102 (43.5)	88.44 ± 11.21
10–20	90 (38.5)	90.58 ± 9.01
20–30	30 (12.8)	93.50 ± 8.82
> 30	12 (5.2)	93.92 ± 8.54
College entrance examination			0.143	0.887
Humanities	17 (7.3)	90.53 ± 10.41
Science	217 (92.7)	90.17 ± 10.10
Class officials in high school			3.720	0.000
Yes	160 (68.4)	91.82 ± 9.76
No	74 (31.6)	86.68 ± 10.01
Class officials in university			2.628	0.009
Yes	72 (30.8)	92.76 ± 10.35
No	162 (69.2)	89.05 ± 9.81
Intentions to engaged in nursing-related work after graduation			1.203	0.302
Yes	125 (53.4)	91.03 ± 9.523
Not sure	94 (40.2)	88.93 ± 10.364
No	15 (6.4)	89.53 ± 11.698
Choosing the nursing major due to personal interest			1.923	0.056
Yes	69 (29.5)	92.14 ± 10.25
No	165 (70.5)	89.38 ± 9.96

Self-directed learning, mobile phone addiction, and alexithymia in nursing undergraduates

The participants achieved a total SCALNU score of 90.19 ± 10.10. The subscale scores for self-management, information management, and learning cooperation were 32.47 ± 4.07, 34.92 ± 5.03, and 22.80 ± 3.44, respectively.

For mobile phone addiction, the participants’ total MPAI score was 40.12 ± 11.10. The subscale scores for inability to control craving, feeling anxious and lost, withdrawal or escape, and productivity loss were 16.97 ± 4.95, 8.01 ± 3.24, 8.12 ± 2.73, and 7.02 ± 2.69, respectively.

Regarding alexithymia, the total TAS-20 score for participants was 51.32 ± 8.59. The subscale scores for difficulty identifying feelings, difficulty describing feelings, and externally-oriented thinking were 17.26 ± 4.76, 13.03 ± 2.86, and 21.03 ± 3.25, respectively. Notably, 29 participants (11.8%) were identified with alexithymia.

Correlations between self-directed learning, mobile phone addiction and alexithymia

SDL exhibited significant negative correlations with mobile phone addiction (r=-0.294) and alexithymia (r=-0.383, both p < 0.01). Conversely, mobile phone addiction correlated positively with alexithymia (r = 0.339, p < 0.01, Table 2).

Table 2.

Spearman correlation coefficients between self-directed learning, mobile phone addiction and alexithymia

Variable	1	2	3
1 Self-directed learning	1.000
2 Mobile phone addiction	-0.294**	1.000
3 Alexithymia	-0.383**	0.339**	1.000

Note: *p<0.05，**p<0.01

A multiple regression analysis was performed to identify the factors influencing SDL among nursing undergraduates, incorporating seven variables demonstrating significant bivariate correlations with SDL: grade, only-child status, annual household income, experience as class officials in high school, experience as class officials in university, alexithymia, and mobile phone addiction. The final model revealed five statistically significant influence factors (all p < 0.01): grade (β = 0.182), only-child status (β = 0.178), experience as class officials in high school (β = 0.184), alexithymia (β=-0.227), and mobile phone addiction (β=-0.240), as presented in Table 3.

Table 3.

Multiple regression analysis of influence factors on self-directed learning

Independent variables	B	S.E	β	t/F value	p-value
Constant	101.075	4.563		22.152	0.000
Gradea	3.675	1.169	0.182	3.145	0.002
Only-child statusb	3.874	1.244	0.178	3.114	0.002
Experience as class officials in high schoolc	3.998	1.224	0.184	3.266	0.001
Alexithymia	-0.264	0.071	-0.227	-3.731	0.000
Mobile phone addiction	-0.221	0.055	-0.240	-4.042	0.000

Note: F = 13.975, p < 0.01; R² = 0.550, Adjust-R² = 0.280;  ^a:Freshman = 1 and Non-freshman = 2; ^b: Yes = 1 and No = 0; ^c: Yes = 1 and No = 0

B Unstandardized coefficient, SE Standardized error, β Standardized coefficient

Validation of the hypothesized model

To validate the theoretical model (Fig. 1), we conducted a CFA to validate the measurement model. The model fit indices were: χ²/df = 2.228, RMSEA = 0.073, CFI = 0.905, GFI = 0.917, and SRMR = 0.066, all of which indicating a relatively good fit model. Alexithymia positively influenced mobile phone addiction (β = 0.399, SE = 0.122, C.R.=3.68, p < 0.001). Mobile phone addiction and Alexithymia negatively influenced SDL (β=-0.301, SE = 0.213, C.R.=-3.215, p = 0.001; β=-0.305, SE = 0.243, C.R.=-3.219, p = 0.001). Grade, only-child status, and experience as class officials in high school positively predicted SDL (β = 0.229, SE = 0.490, C.R.=3.236, p = 0.000; β = 0.229, SE = 0.513, C.R.=3.323, p < 0.00; β = 0.220, SE = 0.507, C.R.=3.214, p = 0.001). Results of the SEM analysis with standardized coefficients are illustrated in Fig. 2.

Fig. 2.

Structural equation model for alexithymia on self-directed learning, mediated by mobile phone addiction

Additionally, the structural equation model analysis indicated that alexithymia has both a direct effect on SDL (β=-0.305, 95%CI -0.426~-0.138, p = 0.014) and an indirect effect through mobile phone addiction on SDL (β=-0.12, 95%CI -0.202~-0.055, p = 0.007), accounting for a medication effect of 28.2% of the total effect, shown in Table 4.

Table 4.

Indirect effect, direct effect and total effect of alexithymia on SDL through mobile phone addition

Effect Type	Estimate	95% Confidence Interval		p-value	Effect Proportion
		Lower	Upper
Indirect Effect	-0.120	-0.202	-0.055	0.007	28.2%
Direct Effect	-0.305	-0.426	-0.138	0.014	71.8%
Total Effect	-0.425	-0.546	-0.286	0.011	100%

In summary, our findings underscore the pivotal mediating role of mobile phone addiction in the relationship between alexithymia and SDL, offering novel insights into the underlying psychological mechanisms.

Discussion

Our study provides novel empirical evidence on the psychological mechanisms underlying self-directed learning among nursing undergraduates in Southern China. The study reveals that alexithymia exerted both direct and indirect (via mobile phone addiction) negative effects on SDL, with 28.2% mediation effect. These results advance the theoretical framework of SDL, and underscore mobile phone addiction as a modifiable behavioral target for clinical interventions.

This study established a positive correlation between alexithymia and mobile phone addiction, which is consistent with findings from a meta-analysis of mainland Chinese students [20]. Alexithymia can impede the expression of emotions and empathizing with others during face-to-face interactions. According to Compensatory Internet Use Theory [24], individuals with alexithymia are increasingly substitute virtual interaction to escape from the real world and avoid genuine emotional expression. Specifically, mobile devices are perceived as tools for entertainment or escape, enabling excitement and offering temporary relief from feelings of hopelessness and isolation [36]. This creates a vicious cycle in which reduced face-to-face interaction further diminishes essential SDL skills, such as collaborative learning, help-seeking, and focused attention.

The structural equation model revealed that alexithymia exerts both a direct and an indirect effect to SDL. The direct effect demonstrated a significant inverse relationship between alexithymia and SDL. This finding supports the consensus in the literature that alexithymia adversely affects college students’ academic performance [13]. The indirect effect means that mobile phone addiction emerged as a behavioral mediator, exacerbating the impact of alexithymia on SDL, with significant mediation effects accounting for 28.2% of the total effect. According to Zimmerman’s Self-Regulated Learning Theory [17], this phenomenon may interfere with all three phases of self-regulation. (1) During the forethought phase, excessive mobile phone use may hinder the formulation and execution of learning plans, while persistent negative emotions can diminish cognitive abilities and academic motivation, thereby adversely affecting SDL [14, 31, 37]. (2) In the performance phase, alexithymia may both impair peer collaboration abilities [11], and weaken self-control over mobile phone use [20], leading to poor time management and attentional distraction that disrupting deep learning [29]. (3) During the self-reflection phase, alexithymia may compromise metacognitive monitoring, hindering effective learning evaluation and strategy adjustment [21]. In summary, we recommend that these cumulative effects may reduce SDL.

Our findings suggest that reducing alexithymia may enhance SDL by decreasing mobile phone addiction, which underscores mobile phone addiction as a modifiable behavioral target for clinical interventions. Furthermore, adopting positive interventions such as increasing physical activity [38]– [39], strengthening social support [40], enhancing family cohesion [41], and cultivating professional identity [42] can effectively mitigate mobile phone addiction and thereby strengthen SDL outcomes.

Our findings also revealed that non-freshmen level, only-child status, and experience as class officials in high school exhibited stronger SDL abilities among nursing undergraduates. Specifically, non-freshmen exhibited superior SDL, which align with previous findings [43]. This may be due to their extended involvement in PBL and medical curriculum and the gradual improvement of their adaptive learning strategies [44]. Students who are only children showed higher SDL, likely due to focused family educational investments and personalized academic guidance during their early development. Furthermore, students with leadership experience may have better self-awareness, self-planning, self-execution, self-assessment, and self-correction skills [45].

The current study had some limitations. Firstly, the research was conducted during the winter break of college students, coinciding with the early phase of the COVID-19 pandemic. Although there was no outbreak in the participants’ hometowns at that time, the pandemic’s influence on SDL, alexithymia, and mobile phone addiction among college students can be considered negligible. Secondly, given the cross-sectional nature of our study, results should be interpreted with caution. Due to the inability to establish causality in cross-sectional studies, these observed associations should be considered preliminary and necessitate further investigation using longitudinal or experimental approaches.

Conclusions

Mobile phone addiction mediates the relationship between alexithymia and SDL among nursing undergraduates. Interventions at reducing mobile phone addiction among undergraduate students should be implemented to enhance their SDL ability.

Abbreviations

SDL

Self-directed learning

SCALNU

Scale of Competencies of Autonomous Learning of Nursing Undergraduates

MPAI

The Mobile Phone Addiction Index Scale

TAS-20

20-item Toronto Alexithymia Scale

SEM

Structural equation modeling

CFA

Confirmatory factor analysis

n

Number of subjects

χ2

Chi-square statistic

df

Freedom

RMSEA

Root mean square error of approximation

CFI

Comparative fit index

GFI

Goodness-of-fit index

SRMR

Standardized Root Mean Square Residual

PBL

problem-based learning

Authors’ contributions

Author contribution: M.G. wrote the research protocol, supervised the survey, and drafted the original draft. Q.L. and X.C. did the literature review and statistical analyses. M.C., Y.F. and J.W. collected the data, checked the data. Q.W. revised the manuscript. J.S. designed the study and revised the manuscript.

Funding

This study was funded by the 2025 Teaching Research and Reform Project of Shantou University Medical College (Grant No. 25JXGG20) and the 2022 Guangdong Province Undergraduate Higher Education Quality and Teaching Reform Project (Grant No. 589).

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

The research protocol was approved by the Institutional Review Board of Shantou University Medical College (SUMC-2020-05), and written informed consent was obtained from all participants before their study involvement. This study complied with the Declaration of Helsinki.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Clinical trial number

Not applicable.