| Session Title: Master Teacher Series: Preparing Data for Their Next Big Thing: Statistical Analysis |
| Demonstration Session 755 to be held in Pacific C on Friday, Nov 4, 4:30 PM to 6:00 PM |
| Sponsored by the Quantitative Methods: Theory and Design TIG |
| Presenter(s): |
| Dale Berger, Claremont Graduate University, dale.berger@cgu.edu |
| Abstract: It is well known that statistical analyses and inferences can be quite inaccurate if there is a mismatch between data and statistical models. Yet we all have been tempted to run statistical analyses before we have completed a critical first step -- data screening. To avoid costly errors and embarrassment, careful attention must be given to identify and deal with problematic data before we apply our statistical tools for decision making. This demonstration will include a step-by-step application of a checklist of issues to be addressed as we prepare for data analysis. We will discuss diagnostics and remedies, along with principles that can guide our choices. Participants will be given a checklist for data screening and examples of diagnostic and remedial applications with SPSS syntax and output. Examples include univariate, bivariate, and multivariate applications. |