Likert Type Items - Analysis

Dear all,

Aileen Fink said that we can treat likert type item values (which are ordinal) as if they are continuous data when we analyse them.

Could we also apply this principle to descriptive statistics?
For example, a question has [strongly disagree, disagree, neutral, agree, strongly disagree] as options, could we simply convert each option to a number and calculate the mean??

Thank you in advance.