exam 3 introduction to hypothesis testing

research hypothesis

what the researcher thinks is going to happen

statistical hypothesis

formally describes the null & alternative hypothesis

Null

• describes unchanged state (no difference between treatment means)

• H0

• μ1 = μ2 = μ3 = μ4

alternative 

• describes changed state (difference)

• direction of difference between treatment means is not indicated

• H1

• μ1 ≠ μ2 ≠ μ3 ≠ μ4

Significance level (α)

• an arbitrary cutoff point that is chosen before data is collected

• says how much evidence against the null we insist on

• value can change depending on situation

• most common: ≤ 0.05

p-value

a value calculated from data set that describes the chance of getting the observed results (or one as extreme) if the null hypothesis is true

small p-value

• unlikely to have obtained the observed results if the null were true

• more evidence the data provides against null

• reject null

large p-value

• obtained results if null is true

• fail to reject (no statistical difference)

significance level

• p-value < than significance level, reject null & results are statistically significant

• fail to reject null = not significant