Review of Hypothesis Testing and P-Values

Flashcards about Comparison of Means, False positives, Two-Sample t-test P values, Type I and II errors, and Power, based on lecture notes.

Type I Error Rate

False Positive Rate

t test formula

𝑡 = 𝑥̅ − 𝜃 / 𝑆𝐸𝑥 or 𝑡 = 𝑥̅1 − 𝑥̅2 / 𝑆𝐸𝑥1𝑥2

Null Hypothesis (Ho) in two sample t-test

The sample mean of x1 is equal to the sample mean of x2

Alternative Hypothesis (HA) in two sample t-test

The sample mean of x1 is not equal to the sample mean of x2

P-value

The probability of getting a test statistic equal to or further away from the null expectation given the null is true.

Use of P-value

Measure the confidence we have in the null hypothesis & Allows us to perform deductive reasoning

Type II error rate

The probability of accepting a false null

Power

The likelihood of not committing type II error

Power in a t test

∝ Δ * √(n) / σ

α

Type I error rate

n

Sample size

σ

Standard error