Cross-lagged analysis of mobile phone addiction and bedtime procrastination: a comparison of gender and perceived stress levels among Chinese college students

Xiujian Lin, Xingshuai Fu, Yuhao Shen, Gaoyang Liu, Ningning Ding, Guohua Zhang, Jun QianFrontiers in Psychology 2025 · DOI

Introduction

Mobile Phone Addiction (MPA)—characterized by emotional dependence and a loss of control—has become a significant public health concern, particularly among young adults. One of its most pervasive consequences is Bedtime Procrastination (BP), the intentional delay of sleep without external cause. While the link between these two behaviors is well-documented, previous research has largely relied on cross-sectional data, leaving the direction of causality unclear. This study, published in Frontiers in Psychology, utilizes a longitudinal approach to investigate the "vicious cycle" between MPA and BP among Chinese college students, exploring how gender and perceived stress levels moderate this relationship.

Methodology

The researchers conducted a two-wave longitudinal study involving 1,423 college students from Wenzhou, China. Data was collected in November 2021 (T1) and May 2022 (T2). To determine temporal precedence, the team employed Cross-Lagged Panel Modeling (CLPM). This statistical technique examines how a variable at T1 predicts a different variable at T2 while controlling for the stability of each variable over time. Participants were assessed using the Mobile Phone Addiction Index (MPAI), the Bedtime Procrastination Scale, and the stress subscale of the DASS-21.

Key Results: The Bidirectional Loop

Analysis of the total sample confirmed a bidirectional reinforcement loop between mobile phone addiction and bedtime procrastination. Both behaviors showed high stability over the six-month period.

Figure 2: Cross-lagged path model between MPA and BP for the total sample.

As shown in Figure 2, MPA at T1 significantly predicted BP at T2 (β = 0.147, p < 0.001), and BP at T1 significantly predicted MPA at T2 (β = 0.102, p < 0.001). This suggests that while phone addiction leads to later nights, the act of procrastinating on sleep also feeds back into increased phone dependency.

The "Stress Switch" and Gender Dynamics

A critical finding of this research is that the direction of influence shifts based on a student's perceived stress level.

The Impact of High Stress

In high-stress environments, the primary driver of the cycle changes. For students reporting high stress, the path from phone addiction to bedtime procrastination became non-significant. Instead, bedtime procrastination itself became the dominant predictor of future phone addiction.

Figure 3: Cross-lagged path model between MPA and BP for the high-stress group.

In the high-stress group (Figure 3), BP at T1 significantly predicted MPA at T2 (β = 0.172, p < 0.001), while the reverse path was not significant (β = 0.035). This suggests that under pressure, students may use their phones as a maladaptive coping mechanism to escape negative emotions during the hours they should be sleeping. Conversely, in low-stress groups, MPA remained the primary driver of BP (β = 0.198).

Gender Differences

Gender also moderated the relationship. For female students, the relationship remained bidirectional. However, for male students, the influence was unidirectional: MPA predicted subsequent BP, but procrastinating on sleep did not significantly increase their phone addiction levels.

Implications and Limitations

These findings support the I-PACE model, suggesting that addictive behaviors are driven by an interaction between personal predispositions (gender) and situational factors (stress). Practically, the study highlights the need for personalized interventions. For high-stress students, interventions should focus on stress management and healthy coping strategies to break the procrastination-to-addiction link.

Limitations: The study's sample was predominantly female and drawn from a single university, which may limit generalizability. Additionally, the reliance on self-reported data suggests that future research should incorporate objective metrics like screen-time logs or sleep trackers to validate these behavioral patterns.

Summary updated

Abstract

MPA and BP mutually influence each other among college students, with this relationship being moderated by stress levels and gender. Personalized intervention measures should be adopted for college students of different genders and stress levels to effectively prevent and alleviate MPA and BP, promoting the healthy growth and comprehensive development of college students.
BibTeX
@inproceedings{lin2025cross,
  author = {Xiujian Lin and Xingshuai Fu and Yuhao Shen and Gaoyang Liu and Ningning Ding and Guohua Zhang and Jun Qian},
  title = {{Cross-lagged analysis of mobile phone addiction and bedtime procrastination: a comparison of gender and perceived stress levels among Chinese college students}},
  booktitle = {Frontiers in Psychology},
  year = {2025},
  doi = {10.3389/fpsyg.2025.1588090},
}
Figure