|M.Sc Student||Abdu Hisham|
|Subject||Heuristics for Decision-Making in Virtual Reality, Based on|
Visual and Haptic Cues
|Department||Department of Education in Science and Technology||Supervisor||Professor Miriam Reiner|
|Full Thesis text - in Hebrew|
Recent research on decision-making has proved that fast and frugal heuristics achieve good results. It showed that such simple heuristics can compete and even outperform other decision-making strategies which are based solely on linear or statistical methods. As examples of such simple and frugal heuristics one can name: ‘Take the Best’, ‘the Minimalist’, and other heuristics. These fast and frugal heuristics are considered part of the bounded-rationality branch ; while the statistical and the linear methods for decision-making belong mainly to the unbounded-rationality branch.
In the current research we have examined the "Take the Best" heuristic in a real-time decision-making environment embedded in a virtual-reality apparatus. The main goal of the research was to discover how decision-making processes are guided by sensory cues: visual and haptic cues.
The experimental environment was based on a hapto-visual virtual reality. The experiment has investigated also the influence of an intervening questionnaire on the learning processes, i.e. changes in decision-making and its effect on the process of cue sorting within the environment according to their validities. A group of twenty pupils (N=20) boys and girls with an average age of 10.3 years old, was chosen randomly from a fifth grade, in an elementary school in Haifa. Each member of this group has played an immersive decision-making game in this haptic virtual-reality environment. Following a first round of the game, an intervening questionnaire was presented to a ten-pupil sub-group chosen as the experimental group.
The primary hypothesis was that decision-making mean success rates of the experiment in phase A, statistically equals the mean success rates measured by a recent research studying the fast and frugal heuristics. The secondary hypothesis was that a statistically significant difference should be found in decision making between the mean success rates of the experimental group and that of the control group.
The final results of the research confirm both hypotheses of this research with high level of confidence (95%). This suggests that fast and frugal, rather than linear or statistical decision-making methods are employed. In addition, sensory cues are translated into a rational decision-making process which is derived from the ‘most probable prediction’. The ‘Take the Best’ heuristic is applied to create an order of preferences, which leads to a decision. We found also that this process is learnable. Potential implications of these results are mainly in understanding ‘presence’ and in the design of tele-manipulation environments (tele-surgery) and simulators (surgical simulators).