The Positive Advantage of Affective Working Memory Valence Retention and Its Cross-Temporal Stability

The dataset for this study consists of two parts: initial evaluation data for the emotional face library and experimental data from the emotion maintenance task. The initial evaluation data for the emotional face library includes intensity, arousal, and pleasantness ratings for each emotional face image. These metrics are quantified using a 1–9 scale, where 1 represents the lowest level and 9 represents the highest level. These ratings for each image were provided by a group of independent evaluators, with the final average serving as the corresponding metric value for that image. This initial evaluation data provides the foundation for subsequent emotion maintenance tasks, ensuring that the selected images possess clearly quantifiable emotional characteristics.The experimental data for the affective maintenance task recorded participants' judgments regarding which of two emotionally valenced facial images was more intense. The experimental design included sequential and simultaneous presentation conditions to investigate how different presentation methods influence affect maintenance. The data detailed participants' basic information, experimental conditions, image IDs for each pair, judgment outcomes, and reaction times. Furthermore, by comparing the intensity values of each image from the initial assessment data with the participants' judgment outcomes, the accuracy of judgments was calculated. This enabled an evaluation of performance differences under positive and negative maintenance conditions, thereby facilitating an in-depth investigation into the mechanisms of valence maintenance. The integrated application of these datasets aims to provide new empirical support for emotion processing theories and reveal the underlying mechanisms of emotional face images in affective maintenance tasks.