Visual encoding and acoustic encoding are shallower forms of processing than s semantic encoding. We process verbal information best when we encode it semantically, especially if we apply the self-reference effect, making information "relevant to me"
Hope this helped! :)