Hypothesis: Null and Alternative

Once you have a clear research question, the next step is to formulate a hypothesis — a specific, testable prediction about what you expect to find. In research, we work with two complementary hypotheses. The null hypothesis (H₀) states that there is no effect, no difference, or no association. For example: 'There is no difference in HbA1c reduction between metformin and placebo.' The alternative hypothesis (H₁ or Hₐ) states that there IS an effect, a difference, or an association: 'Metformin reduces HbA1c more than placebo.' Why do we start by assuming nothing works (the null)? Because in science, the burden of proof lies with the person making the claim. We assume the null hypothesis is true and then use data to see if there is enough evidence to reject it. If the evidence is strong enough (typically a p-value < 0.05), we reject the null and accept the alternative. This approach might seem backwards, but it protects against false claims. It is much harder to prove something is true than to find enough evidence to reject the assumption that it is not true. Hypotheses can be one-sided (directional) or two-sided (non-directional). A one-sided hypothesis predicts the direction of the effect: 'Drug A reduces blood pressure MORE than Drug B.' A two-sided hypothesis predicts a difference without specifying the direction: 'There is a DIFFERENCE in blood pressure between Drug A and Drug B.' In clinical research, two-sided hypotheses are usually preferred because they are more conservative. Not all research is hypothesis-driven. Exploratory studies, qualitative research, and descriptive studies may not have formal hypotheses. But any study that uses inferential statistics (comparing groups, testing associations) needs a hypothesis pair.