People search for information online to accomplish goals or tasks. Tasks can be as simple finding facts about a well-known person, or as complex as writing a report or planning a trip. In either case, people leave a trace of digital activity - such as search behaviors on commercial search engines - and researchers in interactive information retrieval (IIR) researchers have long aimed to understand how people’s behavior can be used to detect a person’s task. Although task is understood to drive and influence a searcher’s behavior, other factors are known to do this as well. Personal and contextual features influence behavior, and some of these in turn are influenced by a person’s task. The work presented here argues and demonstrates that such factors stand in holistic relationships and need to be modeled accordingly. This work naturally extends existing theory, applying structural equation modeling to discover these complex relationships.