Explain in Common Man’s Term #2 — Hypothesis Testing, P-value and T-score

Following content is based on my understanding, it might not reflect a comprehensive understanding of this concept. This will get updated as I get better understanding of these concept.

Photo by Chris Liverani on Unsplash

Hypothesis testing is a key procedure in inferential statistics¹.

Inferential statistics is the method of reaching conclusions that extend beyond the immediate data alone. For example, we use it when we trying to infer from sample data what the population might be².

For each hypothesis testing, following steps are required:

1. Design hypothesis
2. Determine what significance level (α) to consider
3. Obtain sample data
4. Find out P-value of the data
5. Compare between significance value and P-value

FIRST STEP — Design Hypothesis

a null hypothesis (H₀) and alternative hypothesis (H₁) is needed. Usually, alternative hypotheses is the hypothesis we are trying to prove. Null hypothesis will be the opposite of that. For example, if we want to check certain drug has an effect on the patient or not. H₀ will be that the drug has no effect on the patient. H₁ will be that the drug has an effect on the patient.