A data-driven investigation into whether beautiful visualizations actually perform better — replicating and extending Cawthon & Vande Moere's seminal study on aesthetic-usability trade-offs in data viz.. It is inspired by the research "The effect of aesthetics on the usability of data visualization"by Nick Cawthon & Andrew Vande Moere.
📌 Project Overview Many people assume that beautiful visualizations are always better — but is that true? This project investigates whether increasing visual aesthetics improves or hinders user performance, patience, and completion rates when interpreting data visualizations. Key questions addressed: Does higher aesthetic quality improve user task performance? Does beauty increase user patience? Does aesthetic design reduce abandonment rates?
📂 Dataset The analysis uses survey-based experimental data originally collected by Cawthon & Vande Moere. The dataset records participant responses across different visualization conditions, measuring: Task accuracy Time on task User satisfaction Willingness to continue
📌 Note: The dataset is used strictly for academic and educational purposes.
🛠️ Methodology
Handling missing values Encoding categorical variables Normalizing performance metrics
Distribution analysis Correlation heatmaps Group comparisons (aesthetic vs. non-aesthetic conditions)
Hypothesis testing (t-tests / ANOVA) Effect-size calculations Regression analysis where applicable
Matplotlib & Seaborn for statistical plots Comparative visualizations to highlight aesthetic vs. usability trade-offs
📊 Key Findings Aesthetic improvements do not always guarantee better usability. In some cases, increased visual complexity reduced task accuracy. User patience and perceived satisfaction were positively correlated with aesthetics, but performance was not. The study supports a nuanced view: aesthetics should enhance, not distract from, data interpretation.
🧰 Libraries Used pandas、numpy、matplotlib、seaborn、scipy
📷 Key Visualizations
Correlation between Aesthetic Score & Task Abandonment Rate
Individual & Group Aesthetic Scores
🙋♀️ Author Sight-link