A bit of context. 1). I know all respondents answers to a pre-and-post survey after an hour long training session was conducted. 2). I can also match all respondents pre and post answers. For example, I know respondent one's pre-survey answer to question 1 and post-survey answer to question 1 (both questions are the same). 3). There are 14 questions in the pre-and-post survey that are all the same. The questions are all Likert scale items ranged from 1 to 4. 4). I would prefer to compare the overall results of the survey, but I am willing to go item-by-item. However, I would prefer not to go item-by-item.
What is the "best" way to statistically compare my pre-and-post surveys?