Brown-Forsythe Test Calculator
Brown-Forsythe Test Calculator is a robust variation of Levene's test that uses the median instead of the mean. The best choice for checking homogeneity of variance when your data is skewed or has heavy tails.
Brown-Forsythe Test (Equality of Variances)
A robust modification of Levene's Test using the Median.
Use this to check for Homogeneity of Variance when data is non-normal or skewed.
Data Input
Total Rows: 0
Analysis Options
This plot shows the absolute spread of each group from its Median. Higher boxes
= More Variance.
Download Options
Brown-Forsythe Test Calculator
by Learnbin Lab.
Accessed: January 15, 2026.
Copy
What is the Brown-Forsythe Test?
The Brown-Forsythe test is a statistical test for the equality of group variances based on performing an ANOVA on the absolute deviations from the group medians. While mathematically similar to Levene's test, the use of the median makes the Brown-Forsythe test much more robust against outliers and skewed distributions.
When to use it?
You should prioritize Brown-Forsythe over the standard Levene's (Mean) test when:
- Skewed Data: Your data is not symmetric (e.g., reaction times, income data).
- Heavy Tails: Your data has extreme outliers.
- Pre-ANOVA Check: You need to verify homoscedasticity before running a robust ANOVA or Kruskal-Wallis test.
How our tool works (Accuracy & Stack)
Our tool runs on Python (Google Cloud Functions) using the scipy.stats.levene function with the center='median' parameter. We have cross-validated the F-statistic and P-values against R to ensure 4-decimal place accuracy.
We provide specialized visualizations to help you understand the spread of your data:
- Absolute Deviations Plot: The most accurate way to visualize what the test is actually measuring (the deviation from the median).
- Violin Plots: To see the shape and tails of the distribution.
- Box Plots: For standard outlier detection.
A comprehensive PDF Report is generated instantly for your documentation.
Input Requirements & Limitations
To ensure browser responsiveness for our interactive charts:
- Row Limit: Supports up to 5,000 rows per group.
- Column Limit: Supports up to 10 groups (columns).
- Total Data Points: Maximum 50,000 data points.
CSV import supported. Data is processed in memory for maximum privacy.
Comparison: Brown-Forsythe vs. Levene's (Mean)
Both tests check for equal variances. However, the original Levene's test uses the arithmetic mean. If your data contains outliers, the mean is pulled toward them, which can artificially inflate the variance calculation. The Brown-Forsythe uses the median, which is resistant to outliers, providing a more accurate assessment of variance for non-normal data.
Disclaimer: A Note on Performance, Fair Use & Accuracy
How Our Tools Work:
Our tools are designed for speed and accuracy. Many run instantly in your browser. For advanced statistical analysis (e.g., ANOVA, PCA), we use a high-performance cloud engine to ensure precision. In rare cases where the cloud API is busy, the tool may switch to a backup mode, which takes a few moments to load but guarantees you get your results.
Fair Use Policy:
These tools are free for educational and research purposes. To ensure availability for everyone, excessive automated requests or scraping are prohibited.
Accuracy Disclaimer:
This tool uses industry-standard, open-source scientific libraries to perform its calculations. While we strive for high accuracy, the results are for educational and informational purposes only. All results should be independently verified by a qualified professional before being used for academic publications, medical decisions, or other critical applications.
Our tools are designed for speed and accuracy. Many run instantly in your browser. For advanced statistical analysis (e.g., ANOVA, PCA), we use a high-performance cloud engine to ensure precision. In rare cases where the cloud API is busy, the tool may switch to a backup mode, which takes a few moments to load but guarantees you get your results.
Fair Use Policy:
These tools are free for educational and research purposes. To ensure availability for everyone, excessive automated requests or scraping are prohibited.
Accuracy Disclaimer:
This tool uses industry-standard, open-source scientific libraries to perform its calculations. While we strive for high accuracy, the results are for educational and informational purposes only. All results should be independently verified by a qualified professional before being used for academic publications, medical decisions, or other critical applications.