Movement vs Interaction on Other

This visualisation shows the movement trajectory of participants over the course of each task, alongside the positions in which they interacted with objects which did not belong to them. This shows a top-down orthogonal projection of the room, with each facet sized to the 2m by 2m room.

1. Lines represent movement of each participant. This movement is based on the head position of the user

2. Points represent locations where that colour participant interacted with someone else's object. This interaction position is based on the interacting controller's position

The aim of this visualisation was to identify any interesting patterns in where participants moved and interacted with other people's objects, if at all. For the purposes of our analysis, ownership of an object is not transferred upon interaction, meaning that 10 interactions in a row on the same object results in 10 instances being logged.

The grey centre box for groups 6 to 10 is the table. Left figure shows the colours of each movement trajectory, whereas right figure does not in order to show the points.

There are numerous observations of this visualisation:

• Some groups were very territorial, never moving away from their original areas

• Participants never swapped territories, instead moving in and out of each other's territories as seen by the long "tendrils"

• Interactions were generally clustered together, either in what appears to be group areas (such as the table in G6, G9), or in areas between territories (such as G8),

• Interactions were prevalent during tightly-coupled collaboration, as they were generally clustered in group areas (such as the table in G6, G9), in areas between territories (such as G8), or in areas where participants were clearly sharing the same panel for extended periods of time (such as G7 and G9)

• Some interactions (particularly for Part A groups) occurred outside of these clusters, but appear to be the result of short trips into each other's territories for some specific purpose (usually collaboration) rather than a complete takeover of resources

There are some corrections in this visualisation as well, after referencing the recorded videos:

• The cluster of interactions by Red in G10, Task 2 was him continuously moving and readjusting Blue's panel seemingly for fun

• The cluster of interactions in G10, Task 4 was the group playing around at the final moments of presentation

Code

Based on PlayerData.

We create two layers, one for movement based off player position, and another for action positions by determining controller positions in timestamps where said actions were performed. We then plot this using ggplot2 (shown on this page):