How can variable importance be assessed in random forest methods?

Group of answer choices

Average a variable$$s error for each tree, then average it across all trees. Determine the difference between the bagged and out$-$of$-$bag errors. The lower the error, the more important the variable.

Use a particular variable to define every split for all trees in a forest. Then, determine which ones result in the highest levels of variation from the bagged samples. The higher the variation, the more important the variable.

Calculate a variable$$s out$-$of$-$bag error for each tree and average it across all trees where it was not used as a splitting variable. The higher the error, the more important the variable.

Determine a variable$$s bagged and out$-$of$-$bag error for each tree that uses it as a splitting variable. Average the difference between the bagged and out$-$of$-$bag values, then rank all variables in order of importance.