A Reproducible Survey Workflow for Examining Psychological Resilience, Sense of School Belonging, and Higher-Order Thinking in College Students

Summary

This protocol presents a reproducible survey workflow for data collection, quality screening, and moderated mediation analysis in college students.

Abstract

This method article presents a reproducible cross-sectional self-report survey protocol for examining associations among psychological resilience, sense of school belonging (SSB), higher-order thinking, and internet gaming disorder (IGD) in college students. In this protocol, new employment patterns are treated as a contextual background shaped by generative artificial intelligence, platform-based work, and changing graduate skill expectations rather than as a participant-level exposure variable. The protocol integrates institutional sampling, questionnaire administration, translation documentation, response-quality screening, missing-data handling, scale scoring, common-method-bias screening, variable centering, and conditional process analysis into one standardized workflow. SSB is specified as a statistical mediator, and IGD is specified as a moderator within an association-based moderated mediation framework. Representative outputs include sample-flow records, descriptive statistics, correlation matrices, mediation and moderated mediation estimates, conditional indirect effects, the index of moderated mediation, and interaction plots. Because the representative dataset is cross-sectional and self-reported, the protocol supports transparent estimation of statistical associations rather than causal or longitudinal inference. By standardizing preprocessing decisions and model specifications before analysis, this workflow improves transparency, comparability, and reproducibility in educational and behavioral survey research.

Introduction

New employment patterns shaped by generative artificial intelligence, platform-based work, and cross-domain occupational integration have changed how universities discuss graduate preparedness and student development¹. In this article, new employment patterns are treated as a contextual framing condition rather than a measured participant-level exposure. The protocol does not classify students according to their direct participation in platform labor, gig work, or artificial intelligence-mediated employment. Instead, this context explains why higher-order thinking (HOT), psychological resilience (PR), and school-based support are relevant const....

Protocol

This study was approved by the Ethics Committee of Liuzhou Vocational and Technical College (protocol code: LVTC-2024-02-879; approval date: February 1, 2024). Obtain electronic informed consent before questionnaire access. Remove direct identifiers before data export and analysis. Store raw, cleaned, analysis-ready, scoring, screening, and output files in restricted-access folders.

1. Recruit participants using a prespecified multistage random sampling workflow

1. Build and lock the institutional sampling frame
   1. Define new employment patterns as the contextual background of the study rather than as a pa....

Results

Sample characteristics and preliminary variable patterns
The workflow successfully generated a complete sample-flow record before statistical modeling. As shown in Table 1, a total of 860 students were invited during the prespecified data-collection window, and 800 questionnaires were returned, giving an overall response rate of 93.0%. After response-quality screening, 24 invalid responses were excluded, leaving 776 valid cases in the final analysis-ready dataset. The retained sample.......

Discussion

This method article presents a reproducible survey-to-analysis workflow for examining association-based mediation and moderated mediation patterns among PR, SSB, HOT, and IGD symptoms in college students. The primary contribution of the workflow is methodological rather than causal because the protocol demonstrates how a cross-sectional self-report dataset can be collected, screened, scored, documented, and analyzed through a transparent sequence of prespecified procedures. In this revised protocol, new employment patter.......

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
Institutions; codes I-01 to I-32	Study team	N/A	Locked before random selection and used for institution-level randomization.
Institutional screening log	Study team	N/A	Documented institutional eligibility criteria, source documents, eligibility decisions, and institution codes.
Internet Gaming Disorder Scale	Lemmens, Valkenburg, and Gentile (2015), The Internet Gaming Disorder Scale, Psychological Assessment, 27(2), 567–582. doi:10.1037/pas0000062	N/A	9-item scale aligned with DSM-5 criteria and used as a continuous self-reported IGD symptom measure rather than as a clinical diagnosis.
Mediation model output	Study team	Model 4 output	Output for the mediation model with PR as X, SSB as M, and HOT as Y.
Moderated mediation model output	Study team	Model 59 output	Output for the moderated mediation model with IGD moderating the PR → SSB, PR → HOT, and SSB → HOT paths.
Online questionnaire platform	Wenjuanxing / Questionnaire Star	N/A	Web-based survey platform used for electronic informed consent, online questionnaire administration, forced-response settings, and survey-data export.
Pilot-check record	Study team	N/A	30-student pilot-check record used to assess item clarity, comprehension, completion burden, and platform function. Pilot participants were not included in the final analytic sample.
Psychological Resilience Scale	Hu and Gan revision based on the Connor-Davidson framework	N/A	25-item scale used to measure psychological resilience. Responses were scored on a 5-point Likert scale and computed as mean scores.
Psychological Sense of School Membership Scale	Chinese revised version based on Goodenow’s school-membership framework	N/A	18-item scale used to measure school belonging. Items 3, 6, 9, 12, and 16 were reverse-scored before mean-score computation.
Randomization procedure	Study team	Random seed: 20241007	Used for institution-level and student-level random selection through a seed-based random-number procedure.
Raw survey dataset	Study team	Raw_Survey_NEP_2024_800returned_locked.xlsx	Raw exported dataset containing 800 returned questionnaires before response-quality exclusion.
Response-quality screening log	Study team	N/A	Documented 24 excluded responses, including duplicate responses, eligibility inconsistencies, completion-time exclusions, straight-line responses, and repeated-pattern responses.
Response-quality screening rules	Study team	N/A	Exclusion criteria included duplicate submission, eligibility inconsistency, completion time <360 s, same option selected for ≥85% of Likert items, and repeated alternating patterns across ≥15 consecutive Likert items.
Scale reliability output	Study team	N/A	Output reporting Cronbach’s alpha values for PR, HOT, SSB, and IGD.
Scoring and coding reference file	Study team	Supplementary Table 2	Documents item count, response range, reverse-coded items, score direction, mean-score computation, demographic coding, and reliability values.
Session administration log	Study team	N/A	Recorded administration date, institution code, class context, counselor identifier, invitation count, returned count, session duration, and deviations.
Statistical software	IBM Corp.	IBM SPSS Statistics, version 27.0	Used for descriptive statistics, correlations, reliability analysis, exploratory factor analysis, regression, mediation, and moderated mediation analyses.
Student randomization log	Study team	Random seed: 20241007	Documented student-level stratified random selection within the three selected institutions.
Supplementary Table 1	Study team	Translation and adaptation record	Documents item source, original wording, forward translation, back translation, final Chinese wording, expert review, pilot feedback, and revision rationale.
Supplementary Table 2	Study team	Scoring and coding rules	Documents scoring rules, reverse coding, demographic coding, variable abbreviations, and reliability values.
Translation and adaptation record	Study team	Supplementary Table 1	Documents forward translation, reconciliation, back translation, expert review, pilot checking, and final Chinese wording.
Windows operating system	Microsoft Corp.	Windows 11 Pro, 64-bit	Operating system used for statistical analysis and output generation.