Computing confidence intervals for the difference between means with paired data

This tutorial covers the steps for computing confidence intervals for the mean difference of paired data in StatCrunch. To begin, load the Weight Loss Program data set, which will be used throughout this tutorial. The data set is for an fictitious weight loss program. Each of the 15 rows represent a participant in the program. The Weight Before column is the individualâ€™s weight (in pounds) before undergoing the weight loss program, and the Weight After column is the same individualâ€™s weight (in pounds) after undergoing the weight loss program. This tutorial will cover using paired T methods for comparing the mean difference in weight before and after the program with this raw data set.

Calculating a confidence interval for mean difference

StatCrunch can create a confidence interval for the mean difference with paired data. Choose the Stat > T Stats > Paired menu option. Select the Weight After column for Sample 1 in and select the Weight Before column for Sample 2 in. Under Perform, choose Confidence interval for μ_{D} = μ_{1} - μ_{2}. By default, StatCrunch has a value of 0.95 for the Level input which will produce a 95% confidence interval for the difference between the two means. Enter 0.90 for this input to produce a 90% confidence interval instead and click Compute!. The results below show a 90% confidence interval for the mean weight loss associated with the program where "L. Limit" represents the lower limit and "U. Limit" represents the upper limit of this confidence interval.

Saving differences to data table

In certain situations, it is convenient to use the differences for each pair in other StatCrunch routines. Choose Options > Edit to reopen the paired T dialog window. Turn on the Differences option by checking the associated box under Save and click Compute!. The differences for each pair is then stored in a new data column that can be used for subsequent calculations.

Always Learning
Pearson