HomeOverviewDownload PalamedesSupporting documentationFrequently asked questionsSubmit comments and questionsSubmit a bug reportNews and updates (last update: January 5, 2021)Why "Palamedes"?About usOptions besides PalamedesDemos figure galleryModel Comparison ExamplesConfused about Weibull?Failed Fits?Hierarchical Bayesian PF Fitting

PAL_PFLR_CustomDefine_Demo: Palamedes allows you to custom-define models by reparametrizing parameters to your own liking. The example in the demo deals with a (hypothetical) situation in which an observer is trained on two tasks ('task A' and 'task B'). There are 8 sessions of 500 trials each under each task. A three-parameter learning curve is fitted to the eight sessions under each task (i.e., instead of estimating eight independent thresholds, the three parameters of the learning curve are estimated). It is assumed that slopes of the eight PFs are identical. The model comparison tests whether the four parameters in each task differ between task A and task B. Models do not differ significantly (i.e., no evidence to support that parameters differ between tasks, i.e., the more parsimonious should be preferred). The goodness-of-fit of this model is tested. The distribution of Deviance values may be shifted systematically from the theoretical chi-square distribution of appropriate df. This is a result of data having been collected using an adaptive method.