“…However, the data representation and its content evolved through several intermediate states using representation shifts (table, network, heatmap with dendrogram), representation transformations (circular network, force-directed network, layered network with clusters, layered network with significant clinical variables), and content modifications (raw table values, normalized table values, clusters, significance) driven by feedback from different team members, and leading to emergent representations and insights. Therefore, similar to results reported by other studies on computational boundary objects discussed earlier (Daniel et al, 2017;Taylor, 2007), the boundary object was not a static object, but rather one that evolved having its own emergent states, in addition to having emergent effects on team dynamics. However, in contrast to previous studies where the boundary object was reported as changing mainly in content (e.g., text changes in a blog, graphical additions to a 3D model), here we observe that the boundary object in addition to changing in content, also changed in representation by shifting and transforming.…”