Statistical Analyzer

What it does

The Statistical Analyzer skill enables AI agents to perform a variety of statistical analyses on data, including hypothesis testing (t-tests, chi-square tests), regression analysis (linear, polynomial, multiple), ANOVA, and correlation analysis. It provides plain-English interpretations of the results and generates visualizations like regression plots, residual analysis, and box plots. Finally, it can create PDF or HTML reports summarizing these analyses.

When to use it

• Analyzing A/B test results to determine if a change significantly impacted performance (using t-tests).
• Identifying relationships between variables in a dataset (e.g., age vs. income using regression analysis).
• Comparing the means of different groups to see if there's a statistically significant difference (ANOVA).
• Generating reports summarizing statistical findings for stakeholders.
• Exploring data distributions and identifying potential outliers.

Key capabilities

• Hypothesis Testing: t-tests, chi-square tests, proportion tests, one-sample t-test.
• Regression Analysis: Linear, polynomial, and multiple regression.
• ANOVA: One-way and two-way ANOVA with post-hoc tests.
• Correlation Analysis: Pearson and Spearman correlation with significance testing.
• Distribution Analysis: Normality tests and Q-Q plots.
• Plain-English Results: Interpretation of statistical outputs in understandable language.
• Visualizations: Regression plots, residual analysis, box plots.
• Report Generation: PDF/HTML reports including interpretations.

Example prompts

• "Perform a t-test comparing the 'control' and 'experimental' groups on the 'score' column of my data."
• "Run a linear regression with 'age' as the independent variable and 'income' as the dependent variable, and save the regression plot as 'regression.png'."
• "Generate an ANOVA report to compare different categories based on their scores in the dataset."

Tips & gotchas

• The skill requires data in a CSV format that can be loaded using load_data or load_csv.
• Ensure your data is properly formatted and cleaned before analysis for accurate results.
• Familiarity with statistical concepts (e.g., p-values, R-squared) will help you interpret the output effectively.