A more effective way to check the plausibility of a normal model is to construct a special graph, called a normal-score plot, of the sample data. In order to describe this method, we will first explain the meaning of normal score, indicate how the plot is constructed, and the explain how to interpret to plot. For an easy explanation of the ideas, we work with the small sample size. In practical applications, at least 15 or 20 observations are needed to detect a meaningful pattern in the plot.