range includes the outliers too

spread is only affected by the multiplication of constant (not the addition or subtraction)

binomial experiment conditions

1. fixed trials

2) constant probability of success should exists

blocking and stratification seems similar

blocking - experiment before random assignment

• block 1 and block 2 - are distinct
• to eliminate the confounding (lurking) variable to reduce error

stratification - sampling

• choose some variables within the group
• to find more representative sampling

completely randomized experiment →

randomized block design → matched pair design (eliminate confounding variable)