Without spread

Think about how models are working

Very valid for count data

We rarely care whether a predictor is normal

With spread

Think about how models are working

Very valid for count data

We rarely care whether a predictor is normal

space-evenly

Think about how models are working

Very valid for count data

We rarely care whether a predictor is normal

5 items - spread

Think about how models are working

Very valid for count data

We rarely care whether a predictor is normal

Distribution matters more than normality

Always visualize your data first

5 items - spread-evenly

Think about how models are working

Very valid for count data

We rarely care whether a predictor is normal

Distribution matters more than normality

Always visualize your data first