ViSta is a freely available visual statisics system which features visual tools for revealing structure in data; for showing the results of statistical analyses; for guiding the analyst through an analysis; and for structuring the analyst's data analysis process.
ViSta is available for Microsoft Windows, MacOS, and for Unix with X11. ViSta and its documentation are available from our WWW and FTP servers at:
* http://forrest.psych.unc.edu/research/vista.html
* ftp://www.psych.unc.edu/pub/forrest/vista/
ViSta is also mirrored in Australia, Austria, United Kingdom, and USA at
* http://ssda.anu.edu.au/mirrors/ViSta.html (Australia)
* http://sunsite.univie.ac.at/ViSta (Austria)
* http://micros.hensa.ac.uk/mirrors/vista (UK - ac.uk only 8AM-8PM GMT)
* http://www.psych.unc.edu/fwy/research (USA)
The new release may not be available in Australia for a few days.
* Guidance is available for students and other novices.
* A structured graphical user interface is available for all users.
* A menu interface is available for users who don't need guidance.
* A command line interface is available for sophisticated users and those who don't like graphical interfaces.
* Multivariate data analysis procedures are available for students in multivariate analysis.
* The complete Lisp-Stat (Tierney, 1990) programming environment is available to researchers, graduate students and programmers who wish to extend ViSta's capabilities.
* GuideTools are available for teachers and other experts to create guidance for students and novices.
* Applets can be written for web-based distance learning.
ViSta is being used to teach introductory statistics; introductory and advanced multivariate statistics; statistical graphics; and computational statistics. ViSta is also being used for research and development in statistical graphics and computational statistics.
ViSta is designed for an audience of users having a very wide range of data analysis sophistication, ranging from novice to expert. ViSta provides seamlessly integrated data analysis environments specifically tailored to the user's level of expertise. Guidance is available for students and novices, and tools are available for teachers and other experts to create guidance for these novices. A structured graphical user interface is available for competent users, and a command line interface is available for sophisticated users. The complete Lisp-Stat (Tierney, 1990) programming environment is available to researchers, programmers and graduate students who wish to extend ViSta's capabilities.
ViSta performs univariate and multivariate statistical data analysis and visualization. ViSta's data analysis and visualization capabilities include: exploratory graphics and descriptive statistics; univariate tests and visualizations; analysis of variance; regression analysis; principal components analysis; multidimensional scaling and correspondence analysis. All analysis methods have extensive visualization features.
> DYNAMIC GRAPHICS
Histograms, Boxplots, Diamond Plots, Dotplots, Scatterplots, Biplots, Spinplots, Scatterplot Matrices. Plots support brushing and labeling, and are dynamically linked.
> DESCRIPTIVE STATISTICS
Means, Standard Deviations, Variances, Ranges, Quartiles, Medians, Correlations, Covariances, Distances
>UNIVARIATE ANALYSIS
T- and Z-tests and confidence intervals for single sample, paired samples and two independent samples data, with Wilcoxon Signed-Rank and Mann-Whitney tests in appropriate situations.
> ANOVA
Univariate analysis of variance of orthogonal one- or multi-way data. Model may or may not include two-way interactions. The model visualization is a spreadplot comprised of boxplot, diamond plot, quantile plot, quantile-quantile plot and effects plot.
> REGRESSION
Regression analysis, including univariate, multivariate, robust and monotonic regression, as well as redundancy analysis. Univariate regression visualization includes regression, influence, added-variable, and residual plots . Multivariate regression visualization includes a biplot, spinplot, histogram and scatterplot-matrix.
> PRINCIPAL COMPONENTS
Analysis of correlations or covariances. The model visualization is a spreadplot comprised of a biplot, spin-plot, scree-plot and scatterplot-matrix.
> MULTIDIMENSIONAL SCALING
Analysis of one or more symmetric or asymmetric matrices. The model visualization is a spreadplot comprised of a scatterplot, spin-plot, scree-plot and scatterplot-matrix. The spreadplot supports graphical re-estimation of model parameters.
> CORRESPONDENCE ANALYSIS
Analysis of two-way contingency tables. The model visualization is a spreadplot comprised of a biplot, spinplot, residuals plot and scree-plot. The spreadplot supports graphical re-estimation of model parameters.
Maps that visualize the overall structure of an on-going data analysis session. These maps are created by ViSta as the data analysis session progresses. They aid novice and competent analysts in remembering their analysis session, and permit the analyst to return to earlier steps of the analysis.
> GUIDEMAPS
Maps that visually guide students and other novice data analysts through their data analysis sessions. These maps are created by experts using ViSta's Authoring tools.
> SPREADPLOTS
Multi-window dynamic statistical graphics that visualize data structure and analysis results
> MODEL RE-VISION
Dynamic interactive graphical tools that permit the user to visually re-estimate model parameters
> ViDAL
ViSta's Data Analysis Language. ViDAL is an extension of the underlying Lisp-Stat language that can be used by sophisticated data analysts for data analysis and by programmers to enhance ViSta's capabilities.
> DATASHEET
A graphical datasheet for displaying and editing data.
> AUTHORING TOOLS
Visual tools for authoring new GuideMaps.
NEW FEATURES FOR ALL PLATFORMS:
1) Data can be simulated by sampling from any of 10 hypothetical population distributions.
2) The data simulation feature includes an optional visualization which interactively demonstrates the central limit theorem. A click of the mouse generates one (or several) new samples whose means and standard deviations are added to histograms of their sampling distributions.
3) Part of the code is provided as compiled byte-code for MSWindows and MacOS to improve efficiency.
NEW FEATURES IN ViSta FOR UNIX:
ViSta for Unix now includes all of the features previously available only in the MSWindows and MacOS versions. These include:
1) A new OLS, robust and monotonic regression module.
2) Redundancy analysis (new option of the multivariate regression module).
3) Extensive visual regression diagnostics.
4) A visualization method for Univariate Analysis.
5) A completed visualization method for Analysis of Variance.
6) An improved visualization method for Multivariate Regression.
7) A multivariate regression test.
Minor changes made in ViSta for Unix which were already made in ViSta for MacOS and MSWindows, include:
1) Import Data works with strings as well as symbols and numbers.
2) Create-Data can remove rows of multivariate data with missing values.
3) Datasheet menu items now allow adding several obs/vars simultaneously.
4) Datasheet shows matrix names for matrix data.
5) Datasheet window scrolling and sizing improved.
In addition, window and dialog layout, input and output file dialog boxes, graphic fonts, and the exit process have been improved in the Unix version.
Professor, Psychometrics
CB# 3270 Davie Hall email: forrest@unc.edu
UNC Psychometric Lab www: http://forrest.psych.unc.edu
Chapel Hill, NC 27599-3270 USA voice: 1-919-962-5038
ViSta Software and Documentation site:
http://forrest.psych.unc.edu/research/ViSta.html
ViSta Mirrors:
http://ssda.anu.edu.au/mirrors/ViSta.html Australia
http://sunsite.univie.ac.at/ViSta Austria
http://micros.hensa.ac.uk/mirrors/vista UK (ac.uk only 8AM-8PM GMT)
http://www.psych.unc.edu/fwy/research USA