Null vs Alternative Hypothesis: Key Differences & Easy Guide
A null hypothesis (H₀) claims “no effect” or “no difference”—it’s the default position researchers try to disprove. An alternative hypothesis (H₁ or Ha) states the effect or difference you suspect exists.
People mix them up because H₀ often sounds negative (“no change”), so they assume H₁ must be the “positive” version of the same sentence. In reality, H₀ and H₁ are mirror opposites, not just tone shifts.
Key Differences
H₀ always contains an equality sign (=, ≤, ≥), while H₁ uses ≠, <, or >. Statistical tests calculate a p-value from sample data; if it’s below a chosen threshold (e.g., 0.05), we reject H₀ and accept H₁.
Which One Should You Choose?
You don’t pick one—you test H₀ to see if evidence supports H₁. Set H₀ as the conservative claim you’re willing to abandon only with strong data. Choose H₁ to reflect the smallest meaningful effect you care about.
Examples and Daily Life
Testing a new energy drink: H₀ = “mean alertness equals placebo,” H₁ = “mean alertness is higher.” After a week of A/B testing, if p < 0.05, you conclude the drink works; otherwise, you stick with H₀.
Can both hypotheses be true?
No. They’re mutually exclusive; evidence can only support one at a time.
What if p is exactly 0.05?
Convention treats 0.05 as the cutoff, but real decisions weigh cost, sample size, and context.
Is a directional H₁ always better?
Only if you truly care about one direction; otherwise, use a two-sided H₁ to avoid bias.