|
|
Bewildered by the array of statistical methods available in today's world? Need a guide through the maze of approaches to data analysis? The Computing Center Documents Room library stocks a handy series of booklets and manuals on specific aspects of statistical analysis that can help you choose the right method for your needs.
Here are some typical subject areas that are addressed in the Documents Room statistics collection:
A common mistake among beginners is to use continuous data analysis methods (such as Pearson's correlation, linear regression, or t-tests), when they would be better served by choosing categorical data analysis techniques. Almost any statistic can be calculated with a set of numbers; it's the interpretation of these numbers that is critical. Categorical data analysis is designed to interpret discrete data that continuous methods lack.
To learn more about working with categorical data (especially for data collected from a survey), you might want to check out "Categorical Data Analysis Using the SAS System" by Maura Stokes, et al. Although it's written with the SAS user in mind, the book presents clear descriptions of categorical data analysis problems and offers detailed solutions. Another good resource is the Professional and Advanced Statistics manuals for SPSS v.6.1. A complete set of SPSS manuals is available for review or check-out in the Documents Room.
EXCEL works well for some very basic statistics, such as linear regression and graphical displays. On the other hand, EXCEL does not work well if you have large data sets, or need to use repeated measures, categorical, or any type of multivariate problem. In such cases, it would be better to transfer your data files to darkwing or oregon as comma-delimited ASCII text files and run SAS or SPSS to analyze them. A new Documents Room acquisition, "Data Analysis with Microsoft EXCEL" by Kenneth Berk and Patrick Carey, explains the whys and wherefores of using EXCEL for data analysis. This book also includes a special statistics add-in module for Windows users.
If you're pondering which technique to use, the Documents Room also has a collection of recently-published manuals and guidebooks covering a wide range of statistical approaches. You'll find information on understanding linear models, longitudinal data analysis, multivariate analysis, factor analysis, graphics, survival analysis, time series, and much more. For helpful summaries of specific subjects, take a look at the list of topics covered by the Sage publication series. These short booklets contain easy-to-read descriptions of commonly-used data analysis techniques.
For more information on data analysis techniques, contact the Academic Users consulting service (phone: 346-1758; email: consult@oregon). Or, check the Statistical Resources web page at http://cc.uoregon.edu/~robinh