This study investigates how content-based network partitioning and tie definitions affect the structure and interpretation of social networks in MOOC discussion forums. By analyzing 817 threads and 3124 posts from a MOOC on statistics in medicine, the study categorized interactions into content-related and non-content networks using five tie definitions: Direct Reply, Star, Direct Reply+Star, Limited Copresence, and Total Copresence. Findings reveal that content-related and non-content networks exhibit distinct characteristics, supporting the utility of content/non-content distinctions for network analysis. Tie definitions impact network properties, with Total Copresence showing unique features that could indicate inflated social status from "superthread" initiation.