The Gap Between Salience and Meaning: Dataset and Analysis Code for AI vs. Human Visual Attention Study

Description (English) This repository contains the anonymized and derived datasets from the study “The Gap Between Salience and Meaning: Why AI Visuals Attract Attention but Miss Emotional Depth.” Eye-tracking, facial coding, and behavioral data were collected at Selçuk University to examine how participants visually engage with AI-generated and human-created images. All personally identifiable information has been removed and replaced with anonymized IDs. The supplementary materials accompanying this study include the datasets, analysis outputs, and the visual and video stimuli used during the experiment. Detailed descriptions of the included files are provided below: study-eye-tracking_tasar__m-ed0efbe8-0c90-4610-8df2-2605dec430a3-facial-coding.xlsx – Contains summary metrics of participants’ facial expressions, including emotion categories, facial action units (AUs), and timestamps. study-eye-tracking_tasar__m-ed0efbe8-0c90-4610-8df2-2605dec430a3-fixations.xlsx – Provides fixation-level data per participant and stimulus, including fixation count, duration, time to first fixation (TTFF), and coordinates. study-eye-tracking_tasar__m-ed0efbe8-0c90-4610-8df2-2605dec430a3-items-key-pressed-overall_stats.xlsx – Includes behavioral keypress responses and task performance statistics. study-eye-tracking_tasar__m-ed0efbe8-0c90-4610-8df2-2605dec430a3-raw-gazes.xlsx – Contains raw gaze samples with x–y coordinates and timestamps. study-eye-tracking_tasar__m-ed0efbe8-0c90-4610-8df2-2605dec430a3-raw-gazes-denoised-normalized-and-split.xlsx – Provides preprocessed gaze data that have been filtered, normalized, and segmented by stimulus. study-eye-tracking_tasar__m-ed0efbe8-0c90-4610-8df2-2605dec430a3-survey-results.xlsx – Contains post-experiment questionnaire data, including participants’ self-reports and attitudinal measures. study-eye-tracking_tasar__m-ed0efbe8-0c90-4610-8df2-2605dec430a3-aoi-overall-stats.xlsx – Summarizes aggregated area-of-interest (AOI) metrics such as mean fixation duration and dwell time per region. Visual and Video Stimuli – The visual and video materials used during the experiment (e.g., gündüzel.png, project_lg_12cddf2d-7b14-4f26-a301-8524ed820e43.jpg, etc.) are also included as part of the research dataset. Ethics and consent:All participants gave written informed consent. The study was approved by the Scientific Ethics Committee of the Faculty of Communication, Selçuk University. Data access and reuse:Derived and anonymized data are openly available under the Creative Commons Attribution 4.0 International License (CC BY 4.0).Raw gaze coordinates and any potentially identifiable content (e.g., facial videos) are shared in anonymized numeric form; no visual recordings are included. For any extended access, please contact the corresponding author and sign a data use agreement (DUA). Additional Notes (Türkçe kısa özet) Bu veri seti, yapay zekâ ve insan üretimi görsellere yönelik estetik algı ve dikkat süreçlerini inceleyen göz izleme, yüz kodlama ve anket verilerini içermektedir. Veriler anonimleştirilmiş olup, ham video veya yüz görüntüleri paylaşılmamıştır. Çalışma, Selçuk Üniversitesi İletişim Fakültesi Bilimsel Etik Kurulu onayıyla yürütülmüştür.