Academic Statistical Figure Generation

Problem

Statistical figures in papers often get rejected because mean comparisons lack error bars, significance marks omit test methods, dimensions do not match journal requirements, or colors fail after printing. This skill focuses on statistical inference visualization, turning CSV/Excel data into publication-ready PNG/SVG charts with an emphasis on verifiability rather than just visual polish.

How It Works

• It first confirms the data source and statistical question, then selects a chart type: bar plot, box plot, violin plot, scatter plot, line plot, heatmap, forest plot, ROC curve, Bland-Altman plot, or radar chart.
• Its core capability is error bars and significance annotation, supporting SEM, SD, and CI; significance marks should include the statistical test and p value, while non-significant results are reported as they are.
• It can apply journal presets for size, color, and export settings, with colorblind-friendly and grayscale-safe options; when charts are embedded in Word, it follows caption and multi-figure layout rules.

Boundaries

It is not a general data-analysis tool, and it must not be used to fabricate, beautify, or selectively remove data. It does not publish automatically and does not make statistical judgments for the user. Inputs must come from real experiments, collected data, or public datasets; outputs should be labeled AI-generated; verify the data source, with users responsible for data authenticity and conclusions.