Category: Statistics

Does calling a study “under powered” help or hinder criticism?
A common criticism of research (past and present) is that it’s “under powered” or “has low power”. What this typically means is the study doesn’t have many participants (typically between 5 and 40) and so has low statistical power for most effect sizes in psychology . But something being “under powered” only makes sense when…

Should you analyse ordinal data like interval or ratio data?
A couple of months ago, I wrote a summary of a recent paper arguing you shouldn’t analyse ordinal data like interval or ratio. If you do so, there’s a risk of inflated Type I and Type II error rates, as well as reduced power [zotpressInText item=”{VD8XETGZ}”][note]Open access version here[/note]. In response, Helen Wauck wrote a…

Should you calculate a pvalue when there isn’t randomisation?
The thought behind this question was prompted by reading [zotpressInText item=”{TIBTBKWD}” format=”%a% (%d%, %p%)”], which argues against frequentist inferential statistics. One of the arguments refers to an underlying assumption required to compute pvalues; they need random sampling. Without this, a pvalue is meaningless. But this is rare in social science research [zotpressInText item=”{VRZPC486}” format=”%a% (%d%,…

You can’t assume a normal distribution for your data with N>30
The central limit theorem (CLT) is one of the most foundational concepts in all probability (Daly, 2013). It is commonly understood as: when the means of a variable with a suitable number of observations is plotted on a graph, it can form a normal distribution. When the data comes from many independent and random events, the sum…

Essential R packages for education researchers
The intention behind this is to create an updating well of R resources for scientists who focus on education research. If you have any more suggestions please write a comment below or contact me on social media and I’ll add them. Data manipulation and visualization Tidyverse (Sam Parsons) A collection of packages which allow for comprehensive data…

Why you should think of statistical power as a curve
Statistical power is often defined as “the probability of correctly rejecting H0 when a true association is present” where H0 is the null hypothesis, often an association or effect size of zero (Sham & Purcell, 2014). It is determined by the the effect size you want to detect1, the size of your sample (N), and…