Fig 1.
The red arrow marks an infected facility (red dot). The player’s facility is enclosed by a triangle. Each round spans 6 decision months, where the player can remain at the current level of biosecurity or invest in increased biosecurity from None to Low, Medium, and High. An indicator for the infection status is presented to the participant.
Fig 2.
For each clustering coefficient, K, the sum of the squared errors are plotted using participant’s Biosecurity Adoption Ratings, .
Fig 3.
Participant biosecurity adoption ratings are clustered using the K-means algorithm with K = 3. The circles represent each player’s two dimensional risk attitude rating, . The diamonds portray the center of each cluster. The x axis scores their decisions using a low infection rate (0.08) while the y axis represents scores from a high infection rate (0.3). Near the origin, (0,0) players adopted very little biosecurity during their game-play. In the upper right corner, players adopted the most biosecurity for both infection rates. Points close to the main diagonal do not modify their behavior in response to the game context, while points off the main diagonal show players who differ their behavior in response to simulated opportunities.
Table 1.
Biosecurity level expected returns.
Expected Returns in experimental dollars for each level of biosecurity adopted. The probability of infection (pi) is estimated using several hundred trials at each specified biosecurity level. Recall, the the probability of infection depends on the infection rate and distance to each infected facility.
Fig 4.
Histograms of the proportions of all decisions to remain with no biosecurity (“None”) or increase from None—to Low—Medium—and finally to High as a function of decision month. Biosecurity can only increase one level per month. Less biosecurity was implemented when the infection rate, pinf, was Low (= 0:08); the number of decisions to invest in biosecurity increased with higher infection rates. The Risk tolerant cluster (left column) implements the least biosecurity, while the Risk Averse cluster (right column) invests the most biosecurity under both infection rates. After attaining a High biosecurity level, no more decisions can be logged for the simulated year, which is why most of the decisions from the risk averse cluster are completed by the third decision month. The Opportunistic cluster (middle column) behave like the risk averse group under high infection rate scenarios, and implement less biosecurity during low infection rates.
Fig 5.
Biosecurity Adoption Ratings from 50 industry professionals who attended the 2018 World Pork Expo were compared to 50 participants recruited online from Amazon Mechanical Turk. Each participant’s rating (Ri) was calculated from a single infection rate (0.15) across 192 decision months for a total of 19,200 choices.