Page 170 - Kaleidoscope Academic Conference Proceedings 2021

2021 ITU Kaleidoscope Academic Conference

node to pick. The decision maker then needed to inspect
attributes of the nodes’ neighbors to support their final
choice.

We considered that, in alignment with best practices, the
decision makers’ response times to questions would have to
be recorded. Answers were also recorded as they would be
used as proxies to indicate the level of success in the effort
spent by a decision maker. Lastly, we recorded the reasons
decision makers expressed for their decision. The quality of
answers to HCE was evaluated into 3 levels: low (1),
acceptable (2) and outstanding (3), whereas the quality of
answers to LCE questions was recorded as right (1) or wrong
Figure 4 – Aroaro highlighted node properties (0).

In Figure 4, a node is shown just like a decision maker would
see it in Aroaro’s VR space by means of an HMD. Users
could select the information to be displayed associated with
each node. Similarly, Figure 5 shows how a node is viewed
by a decision maker that uses Gephi. Both figures highlight
the viewed node’s neighbors.

Figure 7 – Two nodes compete for the decision maker’s
attention in Aroaro

For each group, a prior questionnaire and a post-
questionnaire were given to the decision maker. A prior
questionnaire collected typical information about a lab
subject: age, gender, and education degrees, besides
Figure 5 – Node properties in a Gephi network view information more acquainted with the subject’s skills such as
prior VR/AR experience, and vision conditions. A post-
How a decision maker would interact with Aroaro to display questionnaire recorded the participants’ experience on
information about the network nodes is shown in Figure 6, maneuverability of, task effectiveness and satisfaction with
where components of the menu can be seen. Aroaro and Gephi.

5.2 Observations and preliminary results

All participants provided answers for all six questions in
Aroaro’s VR immersive environment and almost all
participants made decisions for all six questions using a 2D
desktop platform in Gephi. Two participants failed to make
a final decision for one HCE question in Gephi; the problem
required participants to consider multiple attributes and
discover hidden relationships. Most participants responded
to easy questions quickly with a high quality of decisions on
both the immersive platform and 2D desktop platform, but
more participants made better decisions or discovered new
insights beyond what they had been trained for, besides
Figure 6 – User controls which node’s information to appealing to common sense, in an immersive environment
display by means of a menu
for more complicated problems. A handful of participants
made decisions that were beyond expectations of quality for
An instance of the type of situations a decision maker faced complex questions on the 2D desktop platform. Some
when answering a HCE question is found in Figure 7. The participants spent a longer time to answer one set-up
figure shows two nodes that centrality analytics found to be
of high importance but which are unable to suggest which

– 108 –