Simulated dataset based on the potato_visual data. Based on the rankings in potato_visual, all n-choose-2 = 190 pairs of items were sampled from each assessor. With probability .9, the pairwise preference was in agreement with potato_visual, and with probability .1, they were in disagreement. Hence, the data generating mechanism was a Bernoulli error model (Crispino et al. 2019) with \(\theta=0.1\).
See also
Other datasets:
beach_preferences,
cluster_data,
potato_true_ranking,
potato_visual,
potato_weighing,
sushi_rankings