Dual Screen Viewing and Ambient TV - dataset

This dataset and analysis pipeline accompany the poster "Dual-Screen Viewing and Ambient TV" presented at Mind Art Knowledge (MAK), Brussels, 15–17 June 2026. In an eye-tracking experiment (N = 31; Pupil Labs Invisible, 200 Hz), participants watched three TV clips — Emily in Paris (15:23 min), Reacher (16:06 min), and a Slow TV landscape control (4:57 min) — in a home-like setting with a tablet running TikTok. Total viewing time was approximately 36 minutes. Memory for the narrative was tested after the session with recall questions and a confidence rating. The repository contains: Three R analysis scripts (dwell time, gaze switching, memory × TikTok) Output CSV files with descriptive statistics and model results - Poster figures (PNG) A codebook describing all CSV columns The abstract Key findings: TikTok captured 35–40% of fixation time during TV viewing. Gaze switch frequency habituated over the course of Reacher (β = −0.18, p < .001) but remained flat throughout Emily in Paris (β = +0.03, p = .45). More TikTok predicted worse recall and lower confidence for Reacher (β = −0.84, p = .05; β = −1.63, p < .001) but not for Emily in Paris — consistent with the concept of ambient TV (Chayko 2020). Raw eye-tracking data are not included due to participant privacy. All aggregate statistics needed to reproduce the poster findings are provided in the CSV files. Ethical approval: Royal Holloway University of London Research Ethics Committee Approval number: #753 Date granted: 27 August 2025 Approving institution: Royal Holloway, University of London All participants provided informed consent. Eye-tracking recordings and personal data are held securely at Royal Holloway, University of London, and are not included in this public repository. Only aggregate statistics and anonymised outputs are shared. This work was supported from OP JAC Project “MSCA Fellowships at Palacký University III.” CZ.02.01.01/00/22_010/0008685, run at Palacký University in Olomouc, Czech Republic.