Procedures

BootCI includes three procedures. For all, you can specify the number of repetitions (i.e., the number of samples drawn) as 1,000, 10,000, 50,000, or 100,000; the CI as 90%, 95%, or 99%; and the output format as either 0–7 digits after the decimal point or general format instead.

Semipartial correlation coefficient squared (sr2).  For this procedure, check the variables for the base model (labeled Base V’s) and the ones to add (labeled IV), and select outcome variable (labeled DV).  Then select the wizard icon (at right on toolbar) or Run > CI for s2 (on menu).  BootCI displays the computed Rsquared for both base and full model and the delta R squared (labeled stat) along with the bootstrapped median and CIs (labeled CI sampled).  For comparison, CIs based on large sample estimation formulas are also given (labeled CI ~est).  For the base and full model R squared, the formula given by Wishart (1931; cited in Olkin & Finn, 1995) is used and for delta R squared, the formula given by Alf and Graf (1999).

Correlations (rs).  For this procedure, check the variables (labeled Base V’s) to be used (the All button checks all; off unchecks them).  Then select the officer icon (middle on toolbar) or Run > CIs for rs (on menu).  BootCI displays the computed rs for all pairs of checked variables along with their bootstrapped medians and CIs (labeled CI sampled).  For comparison, CIs computed with the standard estimation formula (labeled CI ~est) are also given.

Means.  For this procedure, check the variables (labeled Base V’s) to be used (the All button checks all; offunchecks them).  Then select the workman icon (at left on toolbar) or Run > CIs for means (on menu).  BootCI displays the computed means for all checked variables along with the bootstrapped medians and CIs (labeled CI sampled).  For comparison, CIs based on the standard estimation formula (labeled CI ~est) are also given.

Output

The output displayed in the BootCI window is tab-delimited.  If columns don’t lineup as you wish (and to preserve your output), simply right click and Select all, right click and select Copy, then Paste into a spreadsheet program like Excel and adjust the margin widths as you wish.