Posted by Joachim Goedhart on October 8th, 2018
Calculation and reporting of p-values is common in scientific publications and presentations (Cristea and Ioannidis, 2018). Usually, the p-value is calculated to decide whether two conditions, e.g. control and treatment, are different. Although a p-value can flag differences, it cannot quantify the difference itself (footnote 1). Therefore, p-values fail to answer a very relevant question:[…]
Posted by Joachim Goedhart on April 18th, 2018
After leaving the bar, what are we to do? I propose to move on to a better bar (and I hope that you will not be disappointed when you find out that I’m actually referring to an interval). In a previous blog I advocated the transparent presentation and reporting of data in graphs. It was[…]
Posted by Joachim Goedhart on March 24th, 2017
Introduction Graphs (or charts or plots) are often used for the display and summary of data. They are essential tools for the communication of results in presentations or manuscripts. One particular type of graph, the bar graph, is often used to quantitatively compare (multiple) conditions. The earliest known example of a bar graph, dates from[…]