Confidence Intervals (CI)

The class codes which produce resampling based CI's are classified into two.

The first type is the CC which gives the sorted statistic values. The user will have the option of choosing the number of trials (bootstraps) to be used. The student then has to pick the appropriate percentiles to obtain the CI. For example if the sorted statistic values are (100 resamples)

Sorted Medians 
--------------------------- 

21.0    22.0    22.0    22.0    22.0    22.0    22.0    22.0    22.0    22.0    
22.0    22.0    22.0    22.0    22.0    22.0    22.0    23.0    23.0    23.0    
23.0    23.0    23.0    23.0    23.0    23.0    23.0    23.0    23.0    23.0    
23.0    23.0    23.0    23.0    32.0    32.0    32.0    32.0    32.0    32.0    
32.0    32.0    32.0    32.0    32.0    32.0    32.0    32.0    32.0    32.0    
32.0    32.0    32.0    34.0    34.0    34.0    34.0    34.0    34.0    34.0    
34.0    34.0    34.0    34.0    34.0    34.0    34.0    34.0    34.0    34.0    
34.0    34.0    34.0    34.0    34.0    34.0    34.0    34.0    34.0    34.0    
34.0    34.0    34.0    34.0    34.0    35.0    35.0    35.0    35.0    35.0    
35.0    35.0    35.0    35.0    35.0    37.0    42.0    42.0    42.0    44.0

A 95% CI for the population median is found by picking the 3rd and 98th values : (22,42). Class codes of this type are given below.

• One-Sample bootstrap means and medians (Sorted)
• Two-Sample bootstrap Wilcoxon statistic (Sorted)

The second type is the CC which gives the 95% CI directly. In this case the uses does not have the option of choosing the number of trials. This value is prechosen to be 1000. The following is one of the CC's which do this.

CC's which give the 95% CI directly are given below.

One-Sample bootstrap confidence intervals for the population mean and median

• Two-Sample hypothesis test and confidence interval for the location parameter based on the
• mean
• median
• Wilcoxon
• Paired Sample Analysis (Wilcoxon)

2000-08-21