Plots & Visualizations

Generating Plots

The Analysis Chat can generate publication-ready plots by extracting data from your studies and running statistical code. Plots appear as interactive image cards in the chat that you can export, copy, or add to your review draft.

Forest Plots

Forest plots visualize the results of a meta-analysis, showing individual study effects and the pooled summary estimate.

Creating a Forest Plot

Ask the AI to generate a forest plot by specifying the outcome and data:

  • 'Generate a forest plot for depression outcomes across all included RCTs'
  • 'Create a forest plot using the BDI-II scores at post-treatment'
  • 'Run a random-effects meta-analysis and show the forest plot for mortality outcomes'
  • 'Create a forest plot with subgroups by study design'

The AI will extract the necessary data (means, SDs, sample sizes, or event counts) from your studies, run the meta-analysis, and generate the forest plot.

Customizing Forest Plots

Refine your forest plot with follow-up messages:

  • 'Use a fixed-effect model instead of random-effects'
  • 'Sort studies by year instead of effect size'
  • 'Add subgroup analysis by risk of bias rating'
  • 'Change the x-axis label to Standardized Mean Difference'
  • 'Exclude [study] and regenerate the plot'
  • 'Show the plot in grayscale for journal submission'

Funnel Plots

Funnel plots help assess publication bias by plotting study effect sizes against a measure of precision.

  • 'Generate a funnel plot for the depression meta-analysis'
  • 'Create a funnel plot with confidence interval contours'
  • 'Show me a funnel plot to check for publication bias'

Other Visualizations

The AI can generate various other plots through code execution:

  • Bar charts comparing study characteristics
  • PRISMA flow diagrams
  • Risk of bias traffic-light plots
  • Outcome comparison charts
  • Any custom visualization you describe

Working with Plot Cards

Viewing

Plots appear as image cards in the chat. Click the expand icon to view in full screen for detailed inspection.

Exporting

Each plot card includes export options:

  • PNG — high-resolution image for manuscripts and presentations
  • Copy to clipboard — paste directly into documents or slides
  • Add to draft — insert the plot into your review draft

Tips for Better Plots

  • Specify the outcome and effect size metric clearly in your request
  • For forest plots, indicate whether you want fixed-effect or random-effects model
  • Ask for specific customizations (labels, colors, dimensions) in follow-up messages
  • Request grayscale versions if your target journal requires it
  • Always verify the data underlying the plot matches your extracted values
  • Use the expand view to check that all text labels are readable at publication size

Plots generated in the Analysis Chat are created through real statistical code execution, not templates. This means the AI can produce any visualization you can describe, including custom analyses beyond standard forest and funnel plots.