Codes for the article "Youth Screen Media Activity Patterns and Associations with Behavioral Developmental Measures and Resting-state Brain Functional Connectivity".
Author: Kunru Song, PhD Student at Beijing Normal University
First Upload Date: 2022.10
Manuscript Accepted Update: 2023.3.23
Ciation Info Update: 2023.5.24
NDA Study Update: 2023.5.26
This Github Respository contains all anlytic codes for our published article, entitled Youth Screen Media Activity Patterns and Associations with Behavioral Developmental Measures and Resting-state Brain Functional Connectivity Pubmed ID: 36963562; DOI: 10.1016/j.jaac.2023.02.014
According to terms and conditions in ABCD Data Use of Certification (DUC), we cannot make some of the intermeidate data public. If any one want to use the ABCD data or replicate findings in this manuscript, it would be better to apply the ABCD Study data through the NIMH Data Archive (NDA) with the DUC signed by yourself/your institution. In addition, an NDA Study has been created for this pubilshed article, which contains all data used in this study. If you want to reproduce or replicate these results and findings, you can download the data included in this study from NDA Website with querying NDA Study Number: 1595, or NDA Study DOI: 10.15154/1526465.
For more details about the analyses in this manuscript, see supplemental materials with our published article.
This article now has been accepted by Journal of the American Academy of Child and Adolescent Psychiatry. If you would like to use any codes in this repository, please cite
Song K, Zhang JL, Zhou N, et al. Youth Screen Media Activity Patterns and Associations With Behavioral Developmental Measures and Resting-state Brain Functional Connectivity [published online ahead of print, 2023 Mar 20]. J Am Acad Child Adolesc Psychiatry. 2023;S0890-8567(23)00132-6. doi:10.1016/j.jaac.2023.02.014
Preprint Version of our manuscript is available at SSRN: https://ssrn.com/abstract=4141354 or http://dx.doi.org/10.2139/ssrn.4141354
TXT-format analysis log files (auto-generated by MATLAB codes). These log files recorded some useful information to check the correctness of this analysis pipeline.
Analyses results from MATLAB and R. This folder contains all results reported in this manuscript. Some additional results may only be summarised into one sentence in manuscript while full results could be found in this folder. For example, full results from unweighted LME modelling could be found in 'Table_UnweightedLME_behav.xlsx' and 'Table_UnweightedLME_RSFNC.xlsx'. We have done this supplemental analysis according to reviewer's comments while it is too long to be included in this manuscript.
Intermediate data from the raw ABCD V4.0 release tabulated data. These intermediate data were generated by clustering analysis and statistical analyses, most of which have been visualized in this manuscript and supplemental material. The raw data in these figures could be found in the 'Res_3_IntermediateData' folder. Some intermediate data contain the individual-level ABCD V4.0 Tabulated Data, which are not allowed to be shared according to the ABCD DUC. We really regret for such situation while we think these public codes would be sufficient to reproduce findings in this manuscript.
Codes to perform cluster analysis (unsupervised machine learning analysis) in this manuscript K-means clustering and hierarchical clustering were performed by MATLAB (Statistical and Machine Learning Toolbox). Latent profile analysis was performed by R (tidyLPA R package). This folder also included the sensitivity analysis about the reliability and stability of the identified SMA patterns, which includes subsample clustering, reassignment analysis, cluster evaluation index analysis and so on.
Cross-sectional analysis for ABCD baseline wave data, which mainly included the cross-sectional three-level linear-mixed-effect (LME) modeling. The demographic table (table 1) were generated by codes in this folder.
Longitudinal analysis for ABCD baseline, 1-year follow-up, and 2-year follow-up wave. Implemented by three-level conditional growth LME model. For more details about these LME models (cross-sectional and longitudinal), please refer to our manuscript.
Codes generate all figures in this manuscript and supplemental materials.
All subfunctions supporting these analyses. If you want to run any code in above mentioned folders, please add path this folder (subfunctions) to your MATLAB environment path before you run any scripts.
Some MATLAB and R codes require MATLAB toolbox and R packages. If you cannot run some scripts, please take care of this issue. Make sure you have installed all required MATLAB Toolboxes and R packages.