RStudio Video 7 - Nuanced Interpretation of Normality, supported by Shapiro-Wilk Testing in RStudio

The interpretation of whether or not the data meet the normal distribution is not always apparent and how to best interpret normal assumptions (or appreciable departures remains a subject of debate); however, for small data sets, the Shapiro-Wilk test with a p-value approach has had support, with p-values below certain thresholds (like less than 0.05 or even less than 0.15) being indicative of nonparametric distributions. For larger data sets, especially those over 2,000 observations, approaches more reliant on the Shapiro-Wilk Test w-value (above 0.99 being treated as not too distant from normal) coupled with a unimodal bell curve histogram together suggesting meeting the normal assumption. R Studio (rstudio.cloud) provides the Shapiro-Wilk Test (with supporting W-values and p-values) and the ability to produce histograms. The nuance of normal distribution assessment is roughly discussed and how to use these normal distribution diagnostic approaches are described and demonstrated.