Posts tagged ‘delta method’

July 30, 2015

Easy fit of psychometric functions

Using psych.function you can fit the data of the single participant (via call of standard glm function), estimate JND and PSE via delta method and plot the model predictions. The names of the Arguments are consistent with standard glm and lines function preceded by “ps.”; example ps.formula, pd.data, ps.col, etc. For the moment only works for univariable psychometric functions (lazy Italians…). See MERdemo.R and the on line tutorial for details. Write me for feedback/suggestions! ūüôā

Advertisements
April 21, 2014

compatibility with lme4.1 (a first step)

Dear all,

I am working on the new version of the package to make it compatible with lme4.1. You can download  it here or in the Blogrol.

The main idea in the new version of the package is to create a function (I called it xplode.mer) extracting all the values from the fitted model, that are necessary for the MERpsychophysics package. This way, if the internal structure of the lme4 will change, I have to modify the xplode.mer function only.

The function based on the delta method is ready. About the bootstrap-based estimate of the PSE: for the moment I wrote a very simple function based on bootMer{lme4.1}. I called it pseMer(). I am still planning to  work on my own boot function (based on my previous function MERboot) Рstay tuned!

Along with the new function you will find a demo in R to use the xplode.mer function. I will add the user manual and the other functions ASAP. As usual, write me for info/suggestions/bug report. For the moment, I will keep in the blogroll also the old version of the code (compatible with lme4.0)

Alessandro