Do the perceptions of relational mobility work as "second sight" of dynamics in the second-order network zone? A longitudinal analysis