#### Statistics Module

DADiSP/Stats is a menu-driven module for DADiSP that computes statistical measures, calculates probability values based on standard distributions, displays summary reports and graphics, and performs hypothesis tests on sample data.

#### Industry Standard Measurements

DADiSP/Stats uses DADiSP's familiar graphical Worksheet and pull-down menus. All of the DADiSP analysis and graphics features are available to operate on original input data and on the results of statistical operations. The statistics module offers an additional dimension to DADiSP, providing scientists and engineers with accepted industry statistical measurements to analyze quantitative data.

#### Descriptive Statistics

Sample size, mean, median, variance, standard deviation, standard error, maximum, minimum, and range values are immediately avaliable at the click of a button. Any computation can be used as a stand alone value for further computation or presented as a group in a summary report. Statistical computations can be performed on single datasets or tables with row or column orientation.

#### Graphical Statistical Analysis

Mean, median, variance, error bar and box plots provide quick visual summaries of common statistical properites. Xbar, Range and combined Xbar / Range plots are appropriate for quality control and process management applications. Histograms, scatter plots and residual plots further complement the statistical visualization features of DADiSP/Stats.

#### Regression Analysis

Linear and Polynomial fitting offer straightforward statistical modeling capabilities. Sophisticated multiple regression techniques model data with a linear combination of almost any user defined terms. Regression statistics, residual and comparison plots aid in the interpretation of regression results. Chi square goodness of fit statistics provide standard acceptance probabilities and significance levels for regression coefficients.

#### ANOVA

DADiSP/Stats includes powerful Analysis of Variance routines. Variation, sum of squares, degrees of freedom and mean squared statistics are computed for average, between and within sample groups. F ratio, F probability and significance levels for hypothesis testing are automatically displayed.

#### Data Comparison

T test sample comparison for pooled and unequal variances compute significance levels for variance testing. One and Two Tailed T tests calculate statistical measurements for mean testing. F Ratio test, cross-correlation, R and R^2 correlation coefficents provide additional tools for statistical data comparison.

#### Probability Generation

Chi squared, Normal, T Test and F probability density functions are easily generated to provide common standards for several statistical measurements. Lookup functionality is optimized to produce fast and accurate results.

#### Data Management

DADiSP can access data from external binary or ASCII files, from other software packages, or directly from the keyboard. The importing and data management features allow you to read single or multi-column tables. A dataset may be stored as a multi-variate table or as a set of individual series. Data can be read into DADiSP for a one-time data processing session or stored the data in DADiSP permanently.

#### Integrated Results

Each menu option in the statistics menus pops up a dialog box that provides a description of the menu feature and prompts the location of the sample data. The results of each statistical test are displayed automatically on the screen. Each menu option generates a presentation quality screen report or graph that contains the statistical results and text summaries. Each report or graph can be incorporated into a document or sent directly to the printer.

#### Verification and Automation

The underlying statistical macros and menus can be viewed to verify the statistical algorithms or adapt the individual routines to particular applications. DADiSP/Stats was developed using DADiSP's custom menu and macro extension language. The menu and macro files are simple ASCII text files. The statistical functions are fully documented, including formulas and source references. While DADiSP/Stats is completely menu-driven, any of the new statistics features can be accessed from the command line or from one of DADiSP's automated session scripts.