Chapter 2 Monte Carlo Simulation and IBM SPSS Bootstrapping
In an ideal world, we would have the time and funds to obtain information from an entire population we’re interested in, and then we could draw our conclusions; in the real world, this is rarely ever the case. Instead we have to rely on samples, and because of this, we want to ensure that our samples are representative of the population. In addition, because we are using samples rather than a whole population, the statistical techniques we use have assumptions that should be met so that these techniques are performing at their optimal level. However, there are occasions when traditional assumptions either do not hold or there is uncertainty in the sample values. To help alleviate these problems, IBM SPSS Statistics added two advanced statistical techniques that allow users to estimate statistics (like the mean, standard deviation, and so on): Bootstrapping, in version 18, and Monte Carlo Simulation, in version 21. SPSS Bootstrapping is a module, but Simulation is available to all users of SPSS Base.
The basic idea behind bootstrapping is that instead of obtaining additional samples from the population, we create additional samples by resampling data (with replacement) from the original sample. Each of the created samples will follow the same data distribution of the original sample, which in turn, follows the population. Bootstrapping also pertains to situations where the exact sampling distribution of the statistics ...
Get SPSS Statistics for Data Analysis and Visualization now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.