Replication Package for: Prediction of Diet Adherence Among Young Adults: A Cross-Cultural Study in Iraq and Poland

This repository contains the dataset and documentation supporting the manuscript: Prediction of Diet Adherence Among Young Adults: A Cross-Cultural Study in Iraq and Poland Repository Structure & File Directory The compressed archive contains the following core components: Folder: Survey Survey Instrument.pdf Description: The complete questionnaire as presented to participants, showing the exact questions used to gather data on diet adherence, history, barriers, facilitators, and demographics. Codebook.pdf Description: The operational codebook mapping out the exact variables, column names, and numerical schemas used to transform raw responses into the coded dataset (e.g., categorical mappings for demographics, scale definitions for facilitators and barriers). Folder: Data data_README.md Description: It outlines the specific relationship between the raw and filtered datasets. raw_collected_dataset(N=602_all_countries).xlsx Description: The complete, unfiltered raw dataset containing survey responses from participants across all surveyed countries worldwide (N = 602). country_filtered_dataset(N=558_Poland_and_Iraq_only)_coded_data.xlsx Description: The numerically coded dataset restricted exclusively to participants originating from Poland and Iraq (N = 558). This subset forms the analytical basis for the main findings in the accompanying manuscript. Folder: Outputs 03_Data_Cleaning/data_full_processed.xlsx Description: The fully cleaned and processed analytical dataset post-wrangling (N = 289). 03_Data_Cleaning/data_Poland_subset.xlsx Description: The cleaned and processed standalone dataset is restricted strictly to the Poland participant cohort. 03_Data_Cleaning/data_Iraq_subset.xlsx Description: The cleaned and processed standalone dataset is restricted strictly to the Iraq participant cohort. 04_Reliability_Cronbach/cronbach_alpha_summary.xlsx & cronbach_alpha.png Description: Internal consistency and psychometric reliability metrics mapping to the Cronbach values cited in Section 2.4 of the manuscript. 07B_Model_Training_Embedded/ALL_cohorts_model_comparison.xlsx Description: The core model evaluation matrix. This file serves as the mathematical source for Tables 2 and 3, as well as the statistical results reported in the manuscript's Abstract. 07B_Model_Training_Embedded/ALL_cohorts_per_fold_metrics.xlsx & ALL_cohorts_per_fold_features.xlsx Description: Cross-validation granularity records. These files map directly to Supplementary Materials S2.1 and S2.2. 07B_Model_Training_Embedded/model_comparison_embedded_<cohort>.png Description: Group-specific performance visualisations comprising the three bar charts found in Supplementary Materials Figure S2.1. 08_Embedded_LASSO_CV/embedded_lasso_auc_comparison.xlsx & embedded_vs_leaky_auc.png Description: Regularized model diagnostic outputs mapping to Supplementary Materials S2.3, S2.4, and Figure S2.2. OUTPUT/07B_Model_Training_Embedded/training_log.txt OUTPUT/08_Embedded_LASSO_CV/embedded_lasso_cv_log.txt Description: Timestamped execution logs capturing exact execution dates, runtime environments, and per-fold training performance timings to mathematically guarantee execution validity. Usage Notes Users seeking to fully replicate or audit the findings of the paper should inspect the project repository URL provided below.Refer to the Codebook.pdf first to map the numerical codes back to the original questions in the Survey instrument.pdf.