If you did not save the full lead data set from lab1 previously, then read it into S-Plus now. We will need the full data set, not just the subset we used in lab1. If you did save it, make sure that lead is the active dataframe; go to the Data menu and choose Select Data... Select Existing Data, and then in the Existing Data Name field, select lead.
Create a side-by-side boxplot of Verbal IQ (Iqv) for the three groups. Review the steps used in lab1 to create a side-by-side boxplot if necessary. To add a confidence interval to the boxplot, go to the Other Specs "tab", and click to check the box for "Draw Conf Bounds". The confidence interval will appear in a different color. Clean up any labels in the plot, add a title and subtitles that provide documentation for the plot. i.e. source, meaning of labels, etc. Repeat for IQ performance (Iqv) and full-scale IQ (Iqf).
Are assumptions for calculating confidence intervals OK?
Why is the confidence interval so much narrower than the main "box" in the boxplot?
To use S-Plus to find the mean, standard deviation, standard error of the mean, and a confidence interval for the mean, repeat the Data Summaries. Go to the Statistics menu and select Data Summaries, and then Summary Statistics. Select Iqv, Iqp, and Iqf for the Variables field, and select Lead.type for the Group Variables in the "Summaries by Group" box. Click on the Statistics tab to bring up other options. Click on the box for Std Error of Mean, and Conf Limits for Mean; by default the boxes for mean and standard deviation should already be checked off. (if not click to check). Unselect the quantiles. Click on OK. A table of the output will appear in the report window.
Turn in a concise report (1 page max, typed) on your findings with interpretations of the boxplots and confidence intervals.
The following JAVA applet is designed to illustrate the Central Limit Theorem. Try it!
Click on the following link to bring up a JAVA Applet that simulates normal data and constructs confidence intervals. Each line in the plot represents a confidence interval for the mean based on simulated normal data with the population mean set to 0. Not all intervals will contain the population mean. Play with the simulator to see how the number of intervals that do not contain the mean changes with alpha. How do the lengths of the intervals change with alpha?
Bring up JAVA Confidence Interval Simulator