Despite many cultural, methodological and technical improvements, one of the major obstacle to results reproducibility remains the pervasive low statistical power. In response to this problem, a lot of attention has recently been drawn to sequential analyses. This type of procedure has been shown to be more efficient (to require less observations and therefore less resources) than classical fixed-N procedures. However, these procedures are submitted to both intrapersonal and interpersonal biases during data collection and data analysis. In this tutorial, we explain how automation can be used to prevent these biases. We show how to synchronise open and free experiment software programs with the Open Science Framework and how to automate sequential data analyses in R. This tutorial is intended to researchers with beginner experience with R but no previous experience with sequential analyses is required.
Link to the Shiny application: https://barelysignificant.shinyapps.io/blind_sequential_analyses/