Null Hypothesis (H₀)

Introduction: Null Hypothesis (H₀)

The null hypothesis (H₀) is the default assumption in hypothesis testing. It states that there is no effect, no difference, or no relationship between variables. Researchers use it as a starting point to test whether observed results are due to chance or reflect a real effect.

Background

The concept of the null hypothesis was formalised in the early 20th century by statisticians such as Ronald Fisher and Jerzy Neyman. It has since become a foundation of modern research, ensuring results are tested against a clear baseline before conclusions are drawn.

Key Elements / Features

• Definition: H₀ claims no meaningful change, difference, or effect exists.
• Role: Provides a benchmark against which the alternative hypothesis (Hₐ) is tested.
• Statistical testing: A p-value is calculated to measure how consistent the data are with H₀.
• Decision rule: If the p-value is below the chosen significance level (often 0.05), H₀ is rejected.

Applications / Examples

• Medicine: H₀ might state that a new treatment is no better than a placebo.
• Business: H₀ could claim that a new marketing strategy has no effect on sales.
• Manufacturing: H₀ may assume that a new machine produces parts with the same quality as the old one.

For example, in a drug trial, H₀ = “There is no difference in recovery rates between the drug and placebo groups.” Rejecting H₀ would suggest the drug has an effect.

Relevance / Impact

The null hypothesis provides scientific rigour by requiring evidence before accepting new claims. It prevents researchers from overstating results and reduces the risk of false positives. H₀ ensures that findings are supported by data, not assumptions.

See also

• Alternative Hypothesis (Hₐ)
• p-Value
• Hypothesis Testing
• Alpha Risk (Type I Error)