Create more reliable models and generate more accurate results

IBM® SPSS® Bootstrapping is an efficient way to ensure that analytical models are reliable and will produce accurate results. It can be used to test the stability of analytical models and procedures found throughout the SPSS Statistics product family, including descriptive, means, crosstabs, correlations, regression and many others.

SPSS Bootstrapping enables you to:

Quickly and easily estimate the sampling distribution of an estimator by re-sampling with replacement from the original sample.
Create thousands of alternate versions of a data set for a more accurate view of what is likely to exist in the population.
Reduce the impact of outliers and anomalies, helping to ensure the stability and reliability of your models.
Estimate the standard errors and confidence intervals of a population parameter such as the mean, median, proportion, odds ratio, correlation coefficient, regression coefficient and more.

Close [x]

Descriptives table

The descriptives table contains statistics and bootstrap confidence intervals for those statistics. The bootstrap confidence interval for the mean (86.39, 105.20) is similar to the parametric confidence interval (86.42, 105.30) and suggests that the "typical" employee has roughly 7-9 years of previous experience. However, Previous Experience (months) has a skewed distribution, which makes the mean a less desirable indicator of “typical” current salary than the median.

Descriptives table

Close [x]

Confidence interval for proportions – Statistic column

The Statistic column shows the values usually produced by Frequencies, using the original dataset. The Bootstrap columns are produced by the bootstrapping algorithms.

Bias is the difference between the average value of this statistic across the bootstrap samples and the value in the Statistic column. In this case, the mean value of Churn within last month is computed for all 1000 bootstrap samples, and the average of these means is then computed.
Std. Error is the standard error of the mean value of Churn within last month across the 1000 bootstrap samples.
The lower bound of the 95% bootstrap confidence interval is an interpolation of the 25th and 26th mean values of Churn within last month, if the 1000 bootstrap samples are sorted in ascending order. The upper bound is an interpolation of the 975th and 976th mean values.

Confidence interval for proportions – Statistic column

Close [x]

Confidence interval for proportions – Frequency table

The Frequency table shows confidence intervals for the percentages (proportion × 100%) for each category, and are thus available for all categorical variables.

Confidence interval for proportions – Frequency table

Close [x]

Bootstrap for proportions

In the Std. Error column, you see that the parametric standard errors for some coefficients, like the intercept, are too small compared to the bootstrap estimates, and thus the confidence intervals are wider. For some coefficients, like [minority=0], the parametric standard errors were too large, while the significance value of 0.006 reported by the bootstrap results, which is less than 0.05, shows that the observed difference in salary increases between employees who are and are not minorities is not due to chance.

Bootstrap for proportions