
Statistiksoftware  Programmpakete 





SPSS von SPSS
Get indepth data access, data preparation, analytical reporting, segmentation and predictive modeling using SPSS. Easily work with large datasets  faster  as you put your data to work across your entire organization using the optional server. Add more analytical power  as you need it  with optional SPSS modules. Combine SPSS with any of several related products and take your analysis even further.



NCSS von NCSS
NCSS 2000 is a comprehensive, easytouse, statistical program. It includes over 200 procedures and graphics. It is Windows 9x/NT compatible, imports/exports major spreadsheet, database, and statistical file formats. The output mixes text and graphics and is easily transferred to word processors. NCSS is capable of processing large data files (over 1000 variables and 200000 rows).



SAS von SAS
Base SAS software provides tools essential for mastering the four datadriven tasks common to virtually any application: data access, management, analysis and presentation. All within a powerful applications development environment.



SYSTAT von SYSTAT Software
SYSTAT offers more scientific and technical graphing options than any other desktop statistics package. Robust algorithms from leading statisticians give meaningful results  even with extreme data. Fit regression models in nested, 2level data when assumptions of independent observations are violated and the estimates from standard least squares general linear model (GLM) are suspect.



Stata 9 von StataCorp
Stata 9 is a fast and powerful statistical package for research professionals of all disciplines, from biostatisticians to political scientists.



CLEAVE von Human Cognitive Neurophysiology Lab
CLEAVE ("C Language Exploratory Analysis of Variance with Enhancements") is designed for the analysis of very large data sets of the sort obtained in experiments using EEG and fMRI. CLEAVE runs on any operating system and is distributed with executables for Windows and Linux as well as with source code for compilation on other systems. CLEAVE incorporates easytouse data specifications and adds a number of feature enhancements including: corrections of degrees of freedom for nonindependent factors, effect size estimations, a convenient ranking of factors based upon effect size and significance levels, posthoc significance tests, posthoc power table, and random factors.



Statistik  Modelltestung 





LISREL und HLM von Scientific Software International (SSI)
LISREL is a program for structural equation modeling. A free student version of LISREL with reduced capabilities is available.
HLM is a program for hierarchical data analysis. A free student version of HLM with reduced capabilities is available.



AMOS von SPSS
Amos provides you with powerful and easytouse structural equation modeling (SEM) software. Create more realistic models than if you used standard multivariate statistics or multiple regression models alone. Using Amos, you specify, estimate, assess, and present your model in an intuitive path diagram to show hypothesized relationships among variables. This enables you to test and confirm the validity of claims such as "value drives loyalty" in minutes, not hours.



EQS von Multivariate Software
EQS provides researchers and statisticians with a simple method for conducting the full range of structural equation models including multiple regression, multivariate regression, confirmatory factor analysis, structured means analysis, path analysis, and multiple population comparisons. With EQS, no knowledge of matrix algebra is required!



Statistik  Nonparametrik 





Resampling Statistics von P. Good
This program will help you to perform many of the bootstrap and permutation tests described in 'Resampling Methods' by P. Good (1997).



BLOSSOM Statistical Software von Fort Collins Science Center
Blossom includes routines for Multiresponse Permutation Procedures (MRPP, MRBP, and PTMP), Multiresponse Sequence Procedures (MRSP), Least Absolute Deviation and Quantile Regression analysis (LAD). Routines added in this update include a permutation version of ordinary least squares (OLS) regression that parallels the least absolute deviation (LAD) regression permutation test; a permutation and asymptotic chisquare approximation of Pvalues for a rankscore statistic for regression quantiles; empirical coverage tests for univariate goodnessoffit and gsample comparisons that are extensions of the KolmogorovSmirnov family of statistics for comparing cumulative distribution functions, including an option for testing goodnessoffit for a random uniform distribution on a circle; a second option for standardizing multiple dependent variables in multiresponse permutation procedures (MRPP) based on the variance/covariance matrix (Hotelling’s commensuration); computing exact probabilities by complete enumeration of all possible combinations for small block and treatment designs in multiresponse randomized block permutation procedures (MRBP); a Monte Carlo resampling approximation alternative for all the MRPP family of statistics (MRPP /NPERM); and multivariate medians and distance quantiles (MEDQ) to be used as descriptive statistics with MRPP analyses.



PERMUSTAT von A. F. Hayes am Darmouth College
PERMUSTAT for Macintosh conducts approximate and exact randomization tests of a variety of null hypotheses. Rather than relying on normaltheory estimates derived from theoretical sampling distributions for generating probabilities, PERMUSTAT computes probabilities by creating a "permutation distribution" of the obtained outcome, conditioned on the data available, and comparing the obtained test statistic to the permutation distribution.



RESAMPLING STATS von RESAMPLING STATS
RESAMPLING STATS lets you take advantage of the past decade's revolution in statistics using simulation and resampling for hypothesis testing, confidence intervals, assessing standard errors and more. Whether you are an expert or a casual user, a resea rcher, analyst or teacher, you can benefit from the new resampling approach to statistics. No formulas or tables. Every step perfectly understood. RESAMPLING STATS offers an easytouse, powerful tool to conduct repeated simulations (including the bootstr ap) with your data, calculate test statistics, and analyze and view the results.



NPSTAT und NPFACT von R. B. May, M. Hunter und M. Masson
NPSTAT and NPFACT are a subset of programs developed to carry out computer intensive analysis (such as randomization tests but not bootstrapping). They provide nonparametric methods of null hypothesis testing.



StatXact und LogXact von Cytel Software
StatXact 4 is the most complete, uptodate software package for exact nonparametric statistical inference on continuous or categorical data in the world. With StatXact we've covered every mainstream nonparametric procedure, using a fullfledged GUI implementation under Windows. We have provided at least one solution for every important nonparametric statistical problem involving either binary, categorical or continuous data. And, as is our trademark, every single one of these solutions is exact. That is, it remains valid for small, sparse and unbalanced data sets. All in all you can access over 80 exact procedures through the StatXact pulldown menus. In addition, we've provided the tools for you to construct your own exact procedures, or to combine two or more procedures in an optimal fashion through a suitable choice of scores.
LogXact is an logistic regression software that can analyze small, sparse, or imbalanced binary data.



Statistical Calculator (SC) von Mole Software
SC is a statistical calculator that runs under MSDOS and Unix. It can be used very simply, like a handheld calculator: for example, you can evaluate scalar or vector/matrix expressions, or type Tprob(2.145,23) to find an area in the right tail of Student t with 23 degrees of freedom. At this level, sc differs from handheld calculators in what is built in: you can look up noncentral distributions, perform Fourier analysis, carry out robust smoothing, do bootstrap and jackknife calculations, fit many types of regression model, minimize or integrate a funtion, and so on, in each case with a single instruction. But SC is also a structured programming language, and it comes with over 400 builtin routines (compiled C) and a library of over 400 external routines (written in the sc language and loaded from disk as required). This gives instant access to a large range of statistical methods.



CISM
CISM is a freeware application developed to help in statistical problems by means of permutation (randomization) and bootstrap techniques (PC, Win95 and higher versions). The program was created and tested by means of PC with K62 350 MHz CPU and 64 MB (later also 128 MB) of RAM as well as Win98 operating system aboard.



Weitere Überblicksseiten:
Register of Package Programs for Exact Statistical Inference



Buchempfehlungen:
Bortz, J., Lienert, G. A. & Boehnke, K. (2000). Verteilungsfreie Methoden in der Biostatistik (2. Aufl.). Berlin: Springer.
Good, P. (2000). Permutation tests: a practical guide to resampling methods for testing hypotheses (2nd ed.). New York: Springer.



Statistik  Teststärke/Metaanalyse 





GPower von A. Buchner, E. Erdfelder und F. Faul
GPower is a general power analysis program that comes in two essentially equivalent versions: one runs under the Macintosh OS and the other was designed for MSDOS. GPower performs highprecision statistical power analyses for the most common statistical tests in behavioral research, that is, ttests (independent samples, correlations, and any other ttest), Ftests (ANOVAS, multiple correlation and regression, and any other Ftest), and Chi²tests (goodness of fit and contingency tables). GPower computes power values for given sample sizes, effect sizes, and alpha levels (post hoc power analyses), sample sizes for given effect sizes, alpha levels, and power values (a priori power analyses), and alpha and beta values for given sample sizes, effect sizes, and beta/alpha ratios (compromise power analyses).



PASS von NCSS
PASS performs power analysis and calculates sample sizes. Use it before you begin a study to calculate an appropriate sample size (it meets the requirements of government agencies that demand technical justification of the sample size you have used). Use it after a study to determine if your sample size was large enough. PASS lets you solve for power, sample size, effect size, and alpha level. It automatically displays charts and graphs along with numeric tables and text summaries in a portable format that is cut and paste compatible with all word processors so you can easily include the results in your proposal.



MetaAnalysis von Dr. R. Schwarzer an der Freien Universität Berlin
MetaAnalysis is a program package that runs under DOS and performs a number of metaanalyses. It has become very popular after version 5.3 was released in 1989 and is probably still the most frequently used metaanalysis software in the world. It can be used in entire statistics classrooms to promote an understanding of this method. Several metaanalytic methods have become popular in recent years, but many researchers have been reluctant to adopt them because no convenient microcomputer program was available. In the meantime, a comprehensive software package has been developed that solves most of the problems encountered in metaanalytic research. Three approaches can be selected: (1) the analysis of exact probabilities (p values) as suggested by Rosenthal, (2) the analysis of effect sizes d, following the methods of either Glass or of Hedges and Olkin, and (3) the analysis of effect sizes r as recommended by Hunter, Schmidt, and Jackson.



Comprehensive MetaAnalysis (CMA) von MetaAnalysis
Comprehensive MetaAnalysis allows you to work with a spreadsheet interface, compute the treatment effect (or effect size) automatically, perform the metaanalysis quickly and accurately, create a highresolution forest plot with a single click, perform a cumulative metaanalysis, perform a sensitivity analysis, assess the impact of moderator variables, work with multiple subgroups or outcomes within studies, assess the potential impact of publication bias, work with subsets of the data.



MetaWin von M. S. Rosenberg
MetaWinallows the user to easily carry out all common metaanalysis calculations on independent studies from a wide variety of disciplines, including medicine, ecology, and the social sciences, compare mean effect sizes and confidence limits of different classes of studies with the click of a mouse, and examine the results of within and betweengroup heterogeneity using parametric methods as well as resampling tests. The program allows users to refine their analyses by choosing which specific groups and individual studies from their data sets they wish to analyze or eliminate. In addition, Version 2.0 of MetaWin allows users to calculate any of six different effect sizes from their data, and can be used for a variety of statistical metaanalytic models. MetaWin runs on IBMcompatible computers running Windows 95/98/NT.


