How to Perform ANOVA in Python: Simple Guide with Example

To perform ANOVA in Python, use the f_oneway function from the scipy.stats module. Pass your groups of data as separate arguments to f_oneway, and it returns the F-statistic and p-value to test if group means differ.

Syntax

The basic syntax for one-way ANOVA in Python is:

from scipy.stats import f_oneway

f_statistic, p_value = f_oneway(group1, group2, group3, ...)

Here:

• group1, group2, group3, ... are lists or arrays of numerical data for each group.
• f_statistic is the calculated F value.
• p_value tells you if the differences between groups are statistically significant.

python

from scipy.stats import f_oneway

# Example syntax with three groups
f_statistic, p_value = f_oneway([1, 2, 3], [2, 3, 4], [3, 4, 5])

Example

This example shows how to perform one-way ANOVA on three groups of data to check if their means differ significantly.

python

from scipy.stats import f_oneway

# Sample data for three groups
group1 = [23, 20, 22, 21, 24]
group2 = [30, 29, 31, 32, 28]
group3 = [27, 26, 25, 28, 29]

# Perform ANOVA
f_statistic, p_value = f_oneway(group1, group2, group3)

print(f"F-statistic: {f_statistic:.3f}")
print(f"p-value: {p_value:.3f}")

Output

F-statistic: 24.000
p-value: 0.000

Common Pitfalls

• Passing data as a single list of all values instead of separate groups will cause errors.
• Not checking assumptions like normality and equal variances can lead to wrong conclusions.
• Confusing p-value interpretation: a small p-value (< 0.05) means group means differ significantly.

python

from scipy.stats import f_oneway

# Wrong: passing one combined list
try:
    f_oneway([23, 20, 22, 21, 24, 30, 29, 31, 32, 28, 27, 26, 25, 28, 29])
except TypeError as e:
    print(f"Error: {e}")

# Right: pass separate groups
f_statistic, p_value = f_oneway(
    [23, 20, 22, 21, 24],
    [30, 29, 31, 32, 28],
    [27, 26, 25, 28, 29]
)
print(f"F-statistic: {f_statistic:.3f}, p-value: {p_value:.3f}")

Output

Error: f_oneway() missing 1 required positional argument: 'b'
F-statistic: 24.000, p-value: 0.000

Quick Reference

Remember these key points when performing ANOVA in Python:

• Use scipy.stats.f_oneway for one-way ANOVA.
• Pass each group as a separate argument.
• Check assumptions before interpreting results.
• A p-value less than 0.05 usually means significant differences.

Key Takeaways

Use scipy.stats.f_oneway to perform one-way ANOVA in Python.

Pass each group of data as a separate argument to f_oneway.

Interpret the p-value to decide if group means differ significantly.

Check ANOVA assumptions like normality and equal variances before trusting results.

Avoid passing all data combined as one list; groups must be separate.