Assessing Normality
Step One: Make a Graph
Step One: Make a Graph
Can be a stemplot or a histogram. Stop if skewed; normal distributions should be at least relatively symmetric.
Step Two: Check the Empirical Rule
Step Two: Check the Empirical Rule
Check what percent of the data lies between the intervals x̄ ± Sx, x̄ ± 2Sx, and x̄ ± 2Sx. They should be close to 68%, 95%, and 99.7%, respectively
Step Three: Construct a Normal Probability Plot
Step Three: Construct a Normal Probability Plot
Use the calculator. Convert the data into their standardized scores in a list. Create a scatterplot using the original data as the x-axis and the standardized data as the y-axis. If the plot is linear, the distribution is normal.
Step Four: State the Conclusion in Context
Step Four: State the Conclusion in Context
For example:
"The distribution of great white sharks is normally distributed because the histogram is relatively symmetric, the data matches the empirical rule and the line of best fits in the normal probability plot is linear.