TR-VisAWI-Pos: Turkish Adaptation and Psychometric Evaluation of the Positive-Item Visual Aesthetics of Websites Inventory

This record contains the anonymous item-level data, documentation, and R analysis scripts supporting the article "TR-VisAWI-Pos: Turkish Adaptation and Psychometric Evaluation of the Positive-Item Visual Aesthetics of Websites Inventory" by Aydın Erden and Tutku Tuncalı Yaman (Marmara University), submitted to the International Journal of Human–Computer Interaction. Study. The positive-item Visual Aesthetics of Websites Inventory (VisAWI-Pos; Perrig et al., 2024) was adapted into Turkish (TR-VisAWI-Pos). A total of 230 undergraduate students at a public university in Istanbul, Türkiye, each browsed one of four publicly available Turkish websites on a mobile phone and then completed the 18 TR-VisAWI-Pos items and 17 items from selected dimensions of the Turkish User Experience Questionnaire (UEQ). Data were collected anonymously online between November 2025 and March 2026. The study was approved by the Marmara University Social Sciences Research Ethics Committee (decision no. 2025-142). Contents. The archive VisAWI_Pos_TR_repository_v1.0.zip contains:- data/raw/VisAWI_Pos_TR_raw_data.xlsx: anonymous item-level responses (N = 230) to the 18 TR-VisAWI-Pos items and 17 UEQ items.- documentation/: a codebook (Codebook_VisAWI_Pos_TR.xlsx), a machine-readable data dictionary (data_dictionary.csv), and the UEQ item-scoring map (UEQ_scoring_map.csv).- R/: nine R scripts implementing the analyses reported in the article. These cover data screening and item diagnostics; confirmatory factor analyses of the correlated four-factor, second-order, single-factor, merged-factor, and diagnostic bifactor models under WLSMV and MLR estimation; a response-pattern sensitivity analysis; reliability (Cronbach's alpha, ordinal alpha, CFA-based omega); convergent and discriminant validity (AVE, Fornell–Larcker, HTMT); and criterion-oriented validity against the UEQ dimensions.- README.md: a description of the files and instructions for running the analysis. It is also uploaded separately for preview. Respondent-level demographic variables are not included; the sample description in the article is reported at the aggregate level. Derived data and result tables are not included, because the scripts generate them. Usage. Extract the archive, set the repository folder as the working directory, and run source("R/run_all.R", encoding = "UTF-8"). The workflow requires R with the packages lavaan, psych, readxl, and openxlsx, and it uses a fixed random seed. It was run with R 4.6.1, lavaan 0.7-2, psych 2.6.5, readxl 1.5.0, and openxlsx 4.2.9. License. Data and documentation are licensed under CC BY 4.0. The R code in R/ is licensed under the MIT License. When using these materials, please cite this record and the accompanying article.