Large-scale multi-touch displays provide highly interactive spaces for small group activities. These devices feature the ability to detect concurrent touch inputs, which enable multiple co-located collaborators to manipulate virtual spaces in myriad ways. This paper explores two types of interaction, simultaneous and sequential, with regard to how people engage in shared virtual space during a collaborative ideation task. Our findings suggest that the two types of interactions present different patterns in both temporal and spatial dimensions. Sequential interaction is the major interaction technique, while the simultaneous interaction is actively used for information exploration and manipulating objects in personal space. Observation of semantic actions suggests that some behaviors are preferably performed in turns, while others are used more in simultaneous manner. The relationship between the two interaction types with regard to different collaboration factors is explored. We share lessons learned from the study and suggest design implications for multi-user touch interfaces.