Basic statistics webcasts always draw a large crowd, new questions and interesting tips. Sam Gardner’s April 7 webcast was no exception.
Sam started with distributions. He mentioned the CTRL broadcast command, where for certain reports, holding the CTRL key, clicking the red triangle and then selecting the command applies the command to all similar reports open in that window. Sam used it to create normal quantile plots for 10 distributions he had run for sample physical fitness data. Other keyboard shortcuts are available via the JMP Help menu: Help>Books>Quick Reference.
A question about outliers led Sam to discuss how outliers display in relation to box plots. I excerpted a bit about box plots where Sam describes box plots in detail, including how to interpret the markings for quantiles median, mean, confidence intervals, 50% data range and the typical range for data if they were normally distributed.
For those interested in outliers, Sam pointed to Mark Bailey’s Grubbs Outlier Test script, which you can download from the JMP File Exchange (requires free SAS login). The script iteratively (one column at a time) detects an outlier in univariate data using the method of Grubbs. The report includes a statement if the outlier is or is not detected at the significance level indicated and gives a p-value for the observed G statistic.
If you missed the live webcast or want a refresher, consider viewing a basic statistics demo that Sam recorded last year, where he uses JMP for one-sample statistical analysis, two-sample statistical analysis and simple regression.
Don’t forget to take advantage of the Statistics Index available from JMP Help>Statistics Index. Depending on the statistic you look up, it defines the statistic, lets you open a data table to use when running the statistics or gives an example of how the statistic is rendered and displayed in JMP. You can also get information on a statistic by grabbing the question mark icon from the menu bar and clicking it on a statistic value displayed in a JMP analysis.