Results

Bayesian Independent Samples T-Test

Data from Tuk et al. (2011). The authors investigated how increased urination frequency facilitates impulse control in other domains, unrelated to bladder control.

Bayesian Independent Samples T-Test
  BF₁₀ error %
ll_sum 1.766 0.010

The Bayes factor indicates that the data are 1.766 times more likely under the alternative model than onder the null model, which is very weak evidence in favor of HA. Based on the Bayesian results, we cannot conclude that there is a difference in number of rewards chosen between the two urgency groups. The robustness check below indicates that there are no conventional prior width values that would have lead to a different conclusion ("absence of evidence") so this result can be considered robust. The sequential analysis plot shows how the Bayes factor developed, as the data were collected - if only we had stopped after 90 observations, when the Bayes factor briefly dipped above 3...

Inferential Plots

ll_sum

Prior and Posterior
Bayes Factor Robustness Check
Sequential Analysis

Descriptives Plots

ll_sum