Can I believe this paper? Josep Marco-Pallarés and Juan Lupiáñez were kind enough to invite me to deliver a course on publication and reporting biases at Universitat de Barcelona and Universidad de Granada. I explained how to use p-curve, funnel plots and the test of excess significance to assess the reliability of published research. If you missed it, you can download the slides here. They include some background information and commented R scripts to conduct the analyses on your own. Read at your own risk!

A simple algorithm for the offline recalibration of eye-tracking data. During 2014, I worked with Chris Street, Tom Beesley and David Shanks in a series of eye-tracking experiments that yielded more calibration errors than expected. To correct these errors, we developed a simple algorithm to recalibrate the data. The complete description of the algorithm was published in this paper. If you want to use the algorithm to clean your own data, you can download our MATLAB implementation from this link.

Confidence intervals for Cohen’s d. I recently had to learn how to compute confidence intervals for effect sizes. Maarten De Schryver drew my attention to this nice explanation by David Howell. Based on his procedure, I wrote two MATLAB scripts that compute 95% confidence intervals for Cohen’s d from t-values and sample sizes in within- and between-subjects designs. The scripts also compute the variance of Cohen’s d for meta-analyses. If you find any error in the scripts, your feedback is more than welcome.

Using MATLAB in experimental psychology. In April 2014 I gave a 4-hour course at the University of Málaga focused on how to use MATLAB in experimental psychology. The module covered the basics of the MATLAB/Octave programming language (vectors, matrices, scripts, and functions) and provided a brief introduction to some toolboxes particularly relevant for psychological research (Cogent and Psychtoolbox). You can download the materials of the course (slides and some example scripts) from this link (in Spanish only, sorry).