Sınırda Kişilik Bozukluğu Ölçeği (BPDI TR)

## Project Title Psychometric Evaluation of the Borderline Personality Disorder Inventory (BPDI): Structural Validity, Measurement Invariance, and Screening Utility ## Description This repository contains the data, materials, and analytic code associated with the study evaluating the psychometric properties and screening performance of the Turkish version of the Borderline Personality Disorder Inventory (BPDI). The study aimed to examine the structural validity, reliability, measurement invariance, and clinical utility of the BPDI as a brief screening instrument for borderline personality pathology using contemporary psychometric and decision-analytic approaches. ## Repository Contents This repository includes: - De-identified, analysis-ready datasets for: - Exploratory factor analysis (EFA sample) - Confirmatory factor analysis (CFA sample) - Study materials: - Turkish version of the BPDI - Analytic code: - Full R script reproducing all analyses reported in the manuscript - Documentation: - README file describing the analytic workflow and reproducibility steps ## Analytic Transparency All analyses reported in the manuscript can be reproduced using the provided datasets and R script. The code includes procedures for: - Item analysis and descriptive statistics - Exploratory and confirmatory factor analyses - Bifactor modeling and reliability estimation - Convergent, known-groups, and incremental validity - ROC analyses, likelihood ratios, and predictive values - Decision curve analysis (DCA) - Differential item functioning (DIF/DTF) - Item response theory (IRT; 2PL model) ## Data Sharing and De-Identification The shared datasets are de-identified and restricted to variables required to reproduce the reported analyses. All direct identifiers and unnecessary variables have been removed prior to sharing. ## Reproducibility To reproduce the analyses: 1. Download all files in this repository 2. Place them in the same working directory 3. Run the R script: source("analysis_code_BPDI_TR_excel_full.R") Outputs (tables and figures) will be generated in an "outputs" folder. ## Reproducibility Check To ensure full computational transparency, the analytic code includes an automated reproducibility check. This procedure compares key output parameters generated from the data (e.g., AUC values, optimal cut-off scores, sensitivity, specificity, and effect sizes) with the corresponding values reported in the manuscript. The script generates a summary report indicating whether the reproduced results match the reported values within predefined tolerance thresholds. This allows independent researchers to verify the consistency between the shared data, analytic code, and published findings. The reproducibility check output is saved in the "outputs" folder as: - REPRO_CHECK_report.txt - REPRO_CHECK_table.csv These files provide a transparent verification of analytic reproducibility. ## Ethical Considerations Data were collected in accordance with institutional ethical standards. Only de-identified data are shared in this repository to ensure participant confidentiality. ## Preregistration This study was not preregistered. ## Contact For questions regarding the dataset or analyses, please contact: Sedat Batmaz