CONTENTS

General Statistics

The UCLA Department of Statistics maintains a World of Statistics with many links to departments of statistics, software companies, data sources, etc. It also offers Free Electronic Consulting! (POSTED: May 15, 2003)

StatiBot: Your expert system for inferential statistics helps you choose the appropriate analysis for your data set. It will even implement many of them for you! [Of course, caveat emptor!] (POSTED: February 7, 2004)

UCLA Academic Technologgy Services has extensive Statistical Computing Resources. Special strengths include: SAS, Stata, Multilevel Modeling, and Survey Data Analysis. (POSTED: May 13, 2003)

The University of Texas has an extensive Statistics FAQ collection. There are separate sections for AMOS, HLM, Lisrel, MPlus, SAS, SPSS. (POSTED: May 13, 2003)

York University's Statistical Consulting Service and Psychology Department have a nice site for Statistics and Statistical Graphics Resources, with over 580 links. Topics include Graphics and Data Visualization, Psychology and Psychometrics, Categorical Data Analysis as well as the following programs: SAS, SPSS, LispStat, S Plus, Mathematica . (POSTED: May 25, 2003)

The CENTER FOR STATISTICS AND THE SOCIAL SCIENCES, STATISTICAL CONSULTING of the University of Washington has a useful Links for Some Statistical Methods web site, as well as a site with Working Papers. (POSTED: November 10, 2003)

StatLib is "a system for distributing statistical software, datasets, and information by electronic mail, FTP and WWW". (POSTED: December 3, 2003)

Return to Contents

Online Textbooks

Not sure which online textbook is right for you? Here is a review of three major sites (the next three resources), by Amy Fluet. Its a bit old, from 1999, but still relevant: Statistically Significant Sites (POSTED: December 29, 2003)

An interesting and extensive site that tries to explain statistical things clearly: A New View of Statistics. (POSTED: December 28, 2003)

David Lane of Rice University has the HyperStat Online Textbook. A good site for basic statistics, along with links to 195 other online stat courses and other resources. winner of a number of awards. Hyperstat can also be used by instructors for teaching elementary statistyics. This is part of the Rice Virtual Lab in Statistics , which also has Simulations/Demonstrations, Case Studies, and an Analysis Lab, which allows online analysis of basic statistics, such as correlations and t-tests. Paste in your data, specify the analysois, and voila! (POSTED: May 24, 2003)

StatSoft, the makers of Statistica have an Online Electronic Statistics Textbook that seems to be of good quality. (POSTED: May 12, 2003)

David Garson of North Carolina State University has a very extensive set of notes on statistical and methodological topics, Statnotes: An Online Textbook including detailed material on advanced topics, including, (among many others) Structural Equation Modeling, Path Analysis, Logistic Regression, Discriminant Function Analysis and less detailed material on dozens of other statistical topics (e.g., Correspondence Analysis, Canonical Correlation, Latent Class Analysis, Multilevel Modeling/Hierarchical Regression, and Partial Least Squares), as well as such non-statistical research topics as Focus Group Research, Participant Observation, Content Analysis, Sociometry and Network Analysis, Ethnographic Research, Narrative Analysis, and Theoretical Frameworks. (POSTED: May 25, 2003)

Here is a set of brief notes on many statistical topics designed to explain them to practitioners (doctors is their target audience). These come highly recommended. British Medical Journal: Statistics Notes (POSTED: July ,16 2003)

Vassar College has created a web-based book Concepts and Applications of Inferential Statistic which is connected to the VassarStats web site for statistical computation, with Java applets to perform many analyses. (POSTED: January 5, 2004)

Yet another online textbook, requiring more mathematical background, but also providing more insight. Goes from basic concepts through multivariate statistics, by David J. Krus of Arizona State University : Visual Statistics with Multimedia (POSTED: May 24, 2003)

Craig A. Wendorf of the Department of Psychology at the University of Wisconsins - Stevens Point has Web-based Manuals for Univariate and Multivariate Statistics. In addition to basic topics, they cover Multiple Regression, ANOVA/ANCOVA, MANOVA, Cannonical Variate Analysis, Linear Discriminate Function Analysis, and Log-Linear. Descriptions are brief, but provide a beginning. (POSTED: May 23, 2003)

Steve Simon has a web site with an extensive "collection of handouts that I use in my job as a statistical consultant." STATS - STeve's Attempt to Teach Statistics. It includes such important topics as How to Read a Medical Journal Article, Planning your research study Managing Your Research Data. (POSTED: January 13, 2004)

Gerard E. Dallal, Chief of the Biostatistics Unit at the Jean Mayer USDA Human Nutrition Research Center on Aging at Tufts University maintains a nice site: The Little Handbook of Statistical Practice, not surprisingly, with a biostatistical orientation. (POSTED: July 16, 2003)

The Exploring Data "website provides curriculum support materials for teachers of introductory statistics" with an emphasis on graphical exploratory analysis: boxplots, (PUBLISHED and POSTED: July , 2003)

A complete online text with good coverage of multiple regression, cluster analysis, etc. by David W. Stockburger. MULTIVARIATE STATISTICS: CONCEPTS, MODELS, AND APPLICATIONS See also his INTRODUCTORY STATISTICS: CONCEPTS, MODELS, AND APPLICATIONS (through t-tests, correlation, and ANOVA) (POSTED: August 3, 2003)

An online handbook, with an engineering orientation is the: NIST/Sematech: Engineering Statistics Handbook. (POSTED: November 10, 2003)

Here are materials for an Advanced Biometry class. In addition to the usual statistical topics, includes Ordination; The Bootstrap, Jackknife, and Randomization; Spatial Statistics; & Bayesian Estimation. (POSTED: November 10, 2003)

Julian L. Simon's book which teaches basic statistics from the resampling perspective Resampling: The New Statistics is now available online. (POSTED: January 5, 2004)

Return to Contents

Online Journals

An online statistics journal is InterStat. Another online journal is the Journal of Statistics Education put out by the American Statistical Association (ASA). They also maintain a Data Archive of data sets from published articles. (POSTED: October 18, 2003)

Online: The International Journal of Biostatistics. (POSTED: March 2, 2006)

STATISTICA SINICA makes papers over one year old available online. (POSTED: December 10, 2003)

Return to Contents

General Issues in Data Analysis

Clay Helberg has written a valuable discussion of pitfalls to avoid in statistical data analysis: Pitfalls of Data Analysis (or How to Avoid Lies and Damned Lies) (POSTED: May 25, 2003)

Return to Contents

Exploratory Factor Analysis

Exploratory Factor Analysis A book manuscript by Ledyard Tucker and Robert MacCallum (POSTED: 4/22/03)

Kristopher J. Preacher and Robert C. MacCallum REPAIRING TOM SWIFT'S FACTOR ANALYSIS MACHINE POSTED: 4/22/03)

Paul Barrett's Preprints [Incl: Can a single item replace an entire scale ]
Also, program for Factor Similarity Analysis (POSTED: 4/22/03)

Michael Friendly has a detailed: Annotated Factor Analysis Bibliography. (POSTED: April 6, 2004)

Return to Contents

Structural Equation Modeling

Ed Rigdon's SEM FAQ POSTED: 4/22/03)

G. David Garson has extensive class notes on Structural Equation Modeling, with a good introduction and detailed information on using LISREL and AMOS. (POSTED: May 25, 2004)

Scott M. Hofer has an extensive reading list for his course Multivariate Study of Change and Human Development with a stong emphasis on SEM approaches. (POSTED: May 14, 2003)

David Rogosa's song Ballad of the Causal Modeller mp3 or rm format. Lyrics (POSTED: May 14, 2003)

The University of Texas has an Structural Equation Modeling using AMOS: An Introduction. (POSTED: November 10, 2003)

MPlus is a program for complex SEM models, including latent mixture models and models having both multilevel and latent variable features. Unlike many SEM programs, MPlus can deal with discrete outcomes. Their web site now has an extensive collection of downloadable papers illustrating uses of the program. The UCLA Academic Technology Services has available a number of viewable movies of a number of seminars on use of MPlus to conduct various types of latent variable models: ATS Seminars, Spring 2004. Also available for viewing is a course: Statistical Analysis with Latent Variables by Professor Bengt Muthén, developer of MPlus. (POSTED: March 17 & May 19, 2004)

An alternative approach to SEM-type analyses is partial least squares (PLS), which models variances rather than covariances, and may work with smaller sample sizes and/or more indicator variables than traditional SEM. While information on PLS is rather scarce, here are a few places to get started: The StatSoft Textbook: Partial Least Squares (PLS) is a good beginning. Wynne Chin has an: Overview of the PLS Method and: Frequently Asked Questions – Partial Least Squares & PLS-Graph. His <PLS-Graph software that may be available for testing: PLS-Graph Beta Testers Site. [If using PLS-Graph, see: PLS-Graph Tutorial by Paulus Insap Santosa]. See also Chin's references, including basic PLS refs: http://disc-nt.cba.uh.edu/chin/plsref.htm and his paper: Partial Least Squares is to LISREL as Principal Components is to Common Factor Analysis. Other sources are Herv´e Abdi's: Partial Least Squares (PLS) Regression from the Encyclopedia of Social Sciences Research Methods. Another free PLS program is: LVPLS, an old DOS program. (POSTED: March 4, 2005)

Return to Contents

Bayesian Analysis

Simon Jackman has excellent introductory resources on Bayesian methods: Estimation and Inference via Markov chain Monte Carlo: a resource for social scientists. See especially his lecture notes: ICPSR 1999 and ICPSR 2001. Also his three days of lecture notes: Day I Bayesian Analysis in the Social Sciences via Markov chain Monte Carlo: a practical introduction; Day II: Social and Political Measurement via Item Response Models; and Day III: Hierarchical Models. (POSTED: November 24, 2004)

Brani Vidakovic has a collection of: Bayesian Statistics: Handouts. (POSTED: November 24, 2004)

BUGS (Bayesian inference Using Gibbs Sampling) is the free software for Bayesian modeling.Also comes in a Windows flavor: WinBUGS. (POSTED: November 24, 2004)

B-Course "is a web-based data analysis tool for Bayesian modeling, in particular dependence and classification modeling. It can also be used as an interactive tutorial which provides you with data sets that have been prepared in advance." (POSTED: November 26, 2003)

The International Society for Bayesian Analysis (ISBA) promotes "the development and application of Bayesian statistical theory and methods useful in the solution of theoretical and applied problems in science, industry and government." (PUBLISHED and POSTED: January , 2004)

The MCMC Preprint Service covers Markov Chain Monte Carlo, an increasingly popular Bayesian approach. In addition to preprints, it includes links to software and other useful goodies. (POSTED: January 2, 2004)

Return to Contents

MDS and Cluster Analysis

The Nonmetric Multidimensional Scaling (NMDS) and Cluster Analysis page of Neuro Scholar: Knowledge Management of the Neuroscientific literature (POSTED: April 22, 2003) includes discussion of Procrustes rotation, among others.

Return to Contents

Survival Analysis

Stephen Jenkins has made available his extensive lecture notes [essentially a book] on: Survival Analysis with Stata. (POSTED: July 30, 2004)

Return to Contents

Longitudinal Analysis
(See also Multilevel Models, Structural Equation Modeling, and Latent Growth Curve Models)

A useful discussion of the sphericity assumption for longitudinal data by Thom Baguley: An introduction to sphericity. [See also his: A glossary for experimental design and analysis ] (POSTED: 19, 2003)

SLEIPNER is a collection of 16 modules "implementing methods of analysis that were carefully chosen to form a self-contained and empirically grounded toolbox for handling longitudinal data within the person-oriented paradigm." It was developed at the Stockholm Laboratory for Developmental Science by Lars R. Bergman and Bassam Michel El-Khouri. (POSTED: March 23, 2004)

The version of SLEIPNER presented here (version 2.1) now comprises 16 distinct modules forming a workbench for pattern-oriented analysis in the context of the person-oriented approach, structurally formulated to address classificatory questions that arise against a background of cluster analytic oriented investigations of primarily longitudinal, but also cross-sectional data. The release of this version coincides with the release of our book entitled Studying Individual Development in an Interindividual Context which is authored by Bergman, Magnusson, & El-Khouri and published by Lawrence Erlbaum Associates, Inc., 2002.
Return to Contents

Multilevel Models

Epidemiological Bulletin has published an A Glossary for Multilevel Analysis in three parts: Part I; Part II; and: Part III. (POSTED: October 29, 2005)

Multilevel Project You can subscribe to the Multilevel Modelling Newsletters here, as well get back issues and many other resources. (POSTED: May 13, 2003)

Kelvyn Jones, Myles Gould and SV Subramanian have an excellent: Guide to Web-Based resources for Multilevel Modeling and Analysis (POSTED: November 12, 2004)

Tom Snijders' Multilevel Analysis page. (POSTED: May 13, 2003). Includes PINT, a program for power analysis in two-level models. Also includes TWOMOK:

for non-parametric scale analysis of two-level data. The nesting structure is defined by "subjects" (level-one units) being nested in "objects" (level-two units). Each subject provides responses on a set of m dichotomous items with regard to the object in which this subject is nested. The objects are to be scaled, the subjects may be considered to provide 'parallel tests' for the objects.... Examples are: pupils (subjects) who reply to questionnaires about their teachers (objects), and employees (subjects) who \ fill in questionnaires about their departments (objects).

The theoretical sections of the HLM 5 manual are available online. (PUBLISHED and POSTED: November , 2003)

Stephen Raudenbush & Xiao-Feng Liu have made available Optimal Design a "free Windows-based computer program to calculate optimal sample designs and power for a variety of organizational effect and growth study designs." (POSTED: October 13, 2003)

Joob Hox's home page has many publications available for download, including his 1995 book: Applied Multilevel Analysis (pdf). (POSTED: December 23, 2003)

Robert Gibbons & Donald Hedeker have an internet edition of their paper: Applications of Mixed-Effect Models in Biostatistics (pdf) available. Hedeker, et al. Random-effects regression analysis of correlated grouped-time survival data is also available online. Finally, Donald Hedeker & Robert Gibbons' MIXOR, MIXREG, MIXNO, & MIXPREG suite of multilevel programs for DOS and MS Windows are also available for free download. (POSTED: December 28, 2003)

Jan de Leeuw & Ita Kreft have a 2001 survey paper on: Software for Multilevel Analysis (pdf). (POSTED: December 28, 2003)

A generalized model, combining features of multilevel models and structural equation modeling is Generalized Linear Latent And Mixed Models (GLLAMMS). The gllamm.org site has information, and a suite of programs to run the models in Stata. See also the Primer on gllamm: gllamm for complex problems by Stas Kolenikov" (POSTED: June 10, 2005)

Return to Contents

Latent Growth Curve Models

An FAQ: Fitting a latent growth curve model with AMOS (PUBLISHED 29 February 2000 POSTED: May 13, 2003)

A Bengt Muthen Technical Report on Latent Variable Modeling of Longitudinal and Multilevel Data (pdf) (POSTED: May 13, 2003)

Return to Contents

Spatial Statistics

Peter J. Diggle has a site with information on Spatial Statistics. There is also some material on longitudinal data analysis. Paulo J. Ribeiro Jr. and Peter J. Diggle are developing geostatistical software for the R software: geoR. (POSTED: July 16 & December 3, 2003)

Luc Anselin directs the Spatial Analysis Laboratory (SAL), which also makes available the GeoDa software for spatial data analysis. There are also a collection of Luc Anselin's Current Working Papers. The SAL site also has tutorials, links to course notes, and links to other spatial analysis resources. (POSTED: May 25, 2004)

The Center for Spatially Integrated Social Science (CSISS) "seeks to develop unrestricted access to tools and perspectives that will advance the spatial analytic capabilities of researchers throughout the social sciences. CSISS is funded by the National Science Foundation." (POSTED: May 25, 2004)

AI-GEOSTATS describes itself as "the central information server for Geostatistics and Spatial Statistics." (POSTED: July 9, 2004)

There are a number of spatial analysis packages available for the free R language: R spatial projects. See also: . Among particular packages available are: Gstat "for multivariable geostatistical modeling, prediction and simulation"; geoR: A package for geostatistical data analysis using the R software and and its add-on: geoRglm: Software for generalised linear spatial models using R. (POSTED: July 14, 2004)

GeoBUGS is a module for the free program: WinBUGS that performs spatial analysis by Bayesian modeling, using Markov chain Monte Carlo (MCMC) techniques. (POSTED: July 9, 2004)

Space-Time Analysis of Regional Systems (STARS) "is an open source package designed for the analysis of areal data measured over time. STARS brings together a number of recently developed methods of space-time analysis into a user-friendly graphical environment offering an array of dynamically linked graphical views." (POSTED: August 9, 2004)

A. Stewart Fotheringham, Martin Charlton, and Chris Brunsdon have a web site devoted to: Geographically Weighted Regression which investigates spatialy varying relationships between variables. (POSTED: November 12, 2004)

The Spatial Systems Group provides training and TA on spatial analysis. Their site has introductory materials available, including: Introduction to GIS and Geostatistics (pdf), Spatial Data Analysis (pdf), and Spatial Analysis and the Spatial Statistical System (pdf). (POSTED: July 9, 2004)

National Institute of Justice (NIJ) has a site devoted to crime mapping: Mapping and Analysis for Public Safety (MAPS). In addition to many mongraphs, gides, etc., available there is: CrimeStat is a free Spatial Statistics Program from the National archive of Criminal Justice Data. It interacts with GIS software. (POSTED: July 9, 2004)

The University of Virginia has: Geostat (the Geospatial and Statistical Data Center) which has access to a number of data sets, including the Uniform Crime Report data. (POSTED: May 25, 2004)

An introduction: Exploratory Spatial Data Analysis: Emerging tools and concepts. (POSTED: October 27, 2004)

Return to Contents

Nonlinear Multivariate Analysis

Gifi System Homepage George Michailides and Jan de Leeuw paper: The Gifi System for Nonlinear Multivariate Analysis (POSTED: 4/22/03)

Return to Contents

Ordination (in Ecology)

Ecologists use a lot of interesting exploratory and classification techiques, including correspondence analysis, multidimensional scaling, cannonical correlation, and the bilot. See: Ordination Methods for Ecologists Also, Michael Palmer has a A Glossary Of Ordination-Related Terms and Ordination-related links (POSTED: 4/23/03)

Return to Contents

Forecasting

One purpose of data analysis is forecasting. Here is a site devoted to: Forecasting Principles, by J. Scott Armstrong of the Marketing Department of The Wharton School, University of Pennsylvania. (POSTED: June 2, 2004)

Return to Contents

Missing Data

Here is a site with information and resources on one of the most popular ways of dealing with missing data: Multiple Imputation Online, maintained by Stef van Buure of TNO Prevention and Health in Leiden. (POSTED: May 27, 2003)

Programs and other information for Multiple Imputation by Joe Schafer and John Graham are available from The Methodolog Center at Penn State: Free Missing Data Software. (POSTED: May 27, 2003)

The University of Texas Statistical Services FAQ on missing data approaches is a nice introduction; FAQ #25: Handling missing or incomplete data (POSTED: 28, 2003)

Gary King proposes a better algorithm for multiple imputation for missing data: Analyzing Incomplete Political Science Data: An Alternative Algorithm for Multiple Imputation. (POSTED: November 11, 2003)

Return to Contents

Power Analysis

The Power Project has resources on power analysis, including Other Useful Websites. (POSTED: December 28, 2003)

The HyperStat Online Textbook has good power analysis resources. (POSTED: December 28, 2003)

The UCLA Department of Statistics has an online Power Calculator for simple models with normal, exponential, binomial distributions, and correlation coefficients. (POSTED: June 12, 2003)

The Java applets for power and sample size page also allows calculating power for many designs, including t-tests and ANOVA. (POSTED: December 28, 2003)

See Steve Simon's Selecting an appropriate sample size. (POSTED: January 13, 2004)

For those using logistic regression, power can be calculated at Power Calculations for Logistic Regression with Exposure Measurement Error. (POSTED: November 10, 2003)

Russell Lenth has a paper Some Practical Guidelines for Effective Sample-Size Determination (pdf) available, as well as Two Sample Size Practices That I Don't Recommend (pdf). (POSTED: January 13, 2004)

Return to Contents

Meta-Analysis

An online site devoted to The Meta Analysis of Research Studies. Its the place to start. (POSTED: October 18, 2003)

Gene V Glass has an essay: Meta-Analysis at 25 (POSTED: October 18, 2003)

Meta - Analysis: Methods of Accumulating Results Across Research Domains is an online textbook. (POSTED: October 18, 2003)

David Kenny has a computer program for basic meta-analysis, including effect size calculation and testing homogeneity. Outputs data to stat program: META (POSTED: May 14, 2003)

Return to Contents

Artificial Neural Networks

Philip H. Goodman, and Frank E. Harrell, have a paper discussing the advantages and disadvantages of ANNs for regression and biostatistical modeling: Neural Networks: Advantages and Limitations For Biostatistical Modeling The NevProp Artificial Neural Network Software with Cross-Validation and Boostrapped Confidence Intervals software is also freely available. from the Goodman Brain Computation Lab at the University of Nevada. (PUBLISHED and POSTED: July , 2003)

Return to Contents

Psychometrics

Methods of Psychological Research is an online journals published by the Methods Section in the German Psychology Society (DGPs. (POSTED: January 19, 2004)

Practical Assessment, Research & Evaluation (PARE) is an online journal with many methodological articles and an educational assessment orientation. (PUBLISHED and POSTED: October , 2003)

One of the major issues in many studies using raters is how to assess the degree of agreement between raters. This site is a great introduction: Statistical Methods for Rater Agreement. (POSTED: February 7, 2004)

The Psychometric Society publishes their flagship journal Psychometika. All issues from 1936-2000 are available to members on CD at a reasonable price. [See the Index 1936-2000 (pdf)] Many classical statistical techniques were first presented here. (POSTED: October 19, 2003)

The Japanese journal (in English) Behaviormetrika makes recent articles available online. Wai Chan's paper Analyzing Ipsative Data in Psychological Research is a very useful summary. (POSTED: November 10, 2003)

The Penn State Schreyer Institute for Teaching Excellence has brief introductions to psychometrics in the context of academic testing, including Test Design and Construction, Classical Test Theory and a brief introduction to Item Response Theory. (POSTED: December 30, 2003)

An introductory site on Item Response Theory including a 180-page introductory book The Basics of Item Response Theory (pdf) by Frank Baker. (POSTED: December 29, 2003)

John S. Uebersax has a site devoted to Latent Class Analysis, as well as an introductory site on Latent Trait Analysis and Item Response Theory (IRT) Models . See also the Open Letter on Latent Class Analysis and Psychiatry Research, raising issues on how these techniques have been used to identify psychiatric subgroups. (POSTED: February 7, 2004)

For specialized IRT models: Resources For Item Response Theory And Its Applications including both Cumulative Item Response Theory Models (still under construction) and Item Response Theory Models For Unfolding. (POSTED: December 29, 2003)

The University of Illinois IRT Modeling Lab has an IRT Tutorial and free software available. (POSTED: December 29, 2003)

For Rasch models, see: Rasch Measurement Transactions. See also the Institute for Objective Measurement, Inc.. (POSTED: December 29, 2003)

The Rasch software, WinSteps is for sale, but older DOS versions are available for free, as is a student/evaluation version. For IRT, there is a free IRT Command Language (ICL) program, by Brad Hanson "that can perform single- or multiple-group estimation of the 1-, 2-, and 3-parameter logistic item response models for dichotomous items, and the partial credit model and generalized partial credit model for polytomous items." It was built using the Estimation Toolkit for Item Response Models (ETIRM) set of "C++ classes and functions for parameter estimation of item response theory (IRT) models." Brad Hanson also has a number of IRT and psychometric papers available online. (POSTED: December 30, 2003)

Another software for IRT & Rasch models is IRT-Lab, which "an be obtained by providing a full mailing address to Randy Penfield at penfield@coe.ufl.edu" (PUBLISHED and POSTED: January , 2004)

"IRT-Lab performs a variety of functions that can be used by researchers, teachers, and students of Rasch measurement and item response theory (IRT) to facilitate the understanding of these nonlinear measurement models.
  • Graph item characteristic curves associated with the dichotomous Rasch and IRT models, the partial credit model, the generalized partial credit model, and the graded response model.
  • View the log-likelihood functions associated with person and item parameter estimation
  • View the information functions and corresponding standard error functions associated with person parameter estimation.
  • Simulate data for tests having items following a variety of Rasch and IRT models.

Q methodology is a technique for the quantitative analysis of repeated intra-individual measurements. It is used in the study of individual psychotherapy cases, as well as other material (e.g., Luborsky's Symptom-Context Method). It forms a bridge between the quantitative and qualitative research traditions. Here is a web page devoted to it: The QMethod Page (POSTED: August 2, 2003)

Return to Contents

Biostatistics

Much biostatistical (and other) data have multiple perspectives, e.g., children, parents, and teacher. This site is devoted to state-of-the-art approaches to such data: Analysis of multiple informant data. (POSTED: February 7, 2004)

Online: The International Journal of Biostatistics. (POSTED: March 2, 2006)

Return to Contents

Econometrics

The Econometrics Journal Online has many resources and links. (POSTED: October 19, 2003)

RePEc and the associated Ideas are part of a "a collaborative effort of over 100 volunteers in 41 countries to enhance the dissemination of research in economics. The heart of the project is a decentralized database of working papers, journal articles and software components. All RePEc material is freely available." The system contains Working Papers, Articles, Chapters, Books, and Software Components. IDEAS is a service providing information about working papers and published research to the economics profession. Much econometric material, including several journals and a lot of software, is available in RepEc. (POSTED: January 9, 2004)

Return to Contents

Sociological Statistics

Adrian E. Raftery discusses Statistics in Sociology, 1950{2000: A Selective Review (POSTED: November 10, 2003)

Return to Contents

Probability

The Probability Web " is a collection of probability resources on the World Wide Web." (POSTED: June 7, 2003)

Charles M. Grinstead & J. Laurie Snell have an online Introduction To Probability in the form of downloadable pdf files. From basic concepts to Markov Chains and Random Walks. (POSTED: December 29, 2003)

For more advanced probability theory (advanced undergraduate to graduate level) see: Probability Theory: The Logic of Science (postscript, use Ghostview to view) by E. T. Jaynes. This book views probability as the key to correct logical inference. (POSTED: December 30, 2003)

Return to Contents

Statistical Conundrums & Paradoxes

Here is a very nice illustration of Simpson's Paradox, that inference from subgroups of a total population may not be the same as inference about that population. This article shows that the Paradox can have policy-relevant implications, as an analysis shows that US population educational test scores have stayed flat (in some cases) while scores for every subgroup increased! The right has used the flat scores to claim that public education is in crisis. Read this to understand what's going on: Those Misleading SAT and NAEP Trends: Simpson's Paradox at Work (POSTED: July 16, 2003)

Return to Contents

Individual Statisticians/Data Analysts
(Especially those with multiple interests that don't fit in one section)

David A. Kenny Interest include: SEM, Causality, Regression Artifacts, Dyadic Analysis, Social Relations Model, Meta-Analysis (POSTED: May 14, 2003)

David A. Freedman -- Many downloadable papers on causality, ecological inference, and census data analysis (POSTED: May 14, 2003)

John Willett has many papers on longitudinal analysis, including discrete time survival analysis, latent growth modeling, and general issues in the measurement of change. See also the site for his colleague: Judith D. Singer. (POSTED: January 4, 2005)

Return to Contents

Technical Reports & Working Papers

Collection of Biostatistics Research Archive (COBRA) features working/research/technical papers from professors and researchers at Berkeley, Harvard, Penn, John Hopkins, Michigan, Washington and other major institutions. COBRA is an open-access website. (POSTED: March 2, 2006)

Carnegie Mellon Department of Statistics Technical Reports (POSTED: December 3, 2003)

A source of many papers by faculty: UCLA Statistics Research [Inc: Gifi Home Page] (POSTED: May 7, 2003)

The CENTER FOR STATISTICS AND THE SOCIAL SCIENCES, STATISTICAL CONSULTING of the University of Washington Working Papers. (POSTED: November 10, 2003)

Return to Contents

Major Software Packages
SAS

SAS Webring "This Ring is a collection of web sites containing FAQ's, Hints and Tips dedicated to SAS® software coding techniques." (POSTED: May 14, 2003)

Stata

Stata, my favorite statistics package, has extensive resources, including an excellent set of FAQs. You can learn much about data analysis from them. (POSTED: May 15, 2003)

The Statistical Consulting Group of UCLA Academic Technology Services has significantly updated their Resources for Learning and Using Stata, including downloadable movies and FAQ pages and teaching tools updated for Stata 8. Also available there is their: Stata Portal with links to Stata resources all over the web. Especially important are Introductory Tutorials, Course Lecture Notes using Stata, and Sample Programs for statistical texts, using Stata to conduct the analyses. (POSTED: December 3, 2003 & July 20, 2004)

Svend Juul of the Department of Epidemiology and Social Medicine, University of Aarhus, Denmark has a 72-page: Introduction to Stata 8, including material on the new graphics. (POSTED: May 15, 2003)

And for Stata users, the SSC Archives at Boston College is the site for most user add-on modules. Go here to see what's been added in the last 30 days. For instructions on submitting and obtaining SSC modules go to : Submitting and retrieving materials from the SSC Archive. (PUBLISHED and POSTED: January 9 & 10, 2004)

Richard Williams has available handouts on the use of Stata for many common [and not so common] analyses: Stata Highlights. Topics from Using Stata for Confidence Intervals to Ordered Logit & Multinomial Logit Models. (POSTED: November 21, 2004)

For Stata users, here are: Some papers and presentations by Roger Newson, with material especially on rank order statistics, splines, and his parmest suite for creating plots and tables after estimation commands. (POSTED: June 3, 2003)

Stephen Jenkins has made available his extensive lecture notes [essentially a book] on: Survival Analysis with Stata. (POSTED: July 30 & December 13, 2004)

SPSS

The Statistical Consulting Group of UCLA Academic Technology Services have significant SPSS resources available, including their SPSS Portal links to SPSS resources across the web. Especially important are Introductory Tutorials, Course Lecture Notes using SPSS, and Sample Programs for statistical texts, using SPSS to conduct the analyses. (POSTED: July 20, 2004)

For the absolute beginner, see SPSS for Windows: Getting Started for instructions on how to enter, modify, and organize data in SPSS. It also shows how to perform basic analyses using the menu system. It contains many screen shots to show you exactly what to do. (POSTED: December 28, 2003)

For macros, code snippets,and other resources, see: Raymond's SPSS Tools from Raynald Levesque. (POSTED: June 10, 2005)

Wolfgang Ludwig-Mayerhofer has an Internet Guide to SPSS for Windows. [NOTE: this guide teaches how to use SPSS Syntax, not, the graphical user interface (GUI) or the analysis menus. For serious analysis, this is the only way to go. As Ludwig-Mayerhofer points out: "Even though many people think otherwise (because programmers make them believe so): Working with syntax files often is much easier." However, the links page also lists several tutorials that teach the GUI and analysis menus.] This tutorial goes from an introduction to entering data, through data transformation, to advanced analyses, such as multilevel modeling. (POSTED: December 28, 2003)

The 445 page book SPSS for Beginners is downloadable in pdf, but without the ability to print from it. If one want to be able to print, one can buy it in downloadable or printed form. The book contains SPSS screen shots on virtually every page. The down side, from my perspective, is that it teaches menu based analysis, which is problematic for serious research. (POSTED: January 5, 2004)

For more advanced users, the downloadable 390 page book: SPSS Programming and Data Management by Raynald Levesque is must have. (POSTED: June 10, 2005)

R/S+

R is a free complete statistics and graphing package, based on the S language, and offered as Free Software under the GNU General Public License. There are versions for most operating systems, including Windows and Linux. There are also numerous add-on packages for many advanced statistical and data processing applications. The down side is that it requires some investment of time to learn. Available at: The R Project for Statistical Computing (POSTED: June 3, 2003)

An invaluable source of ideas for solving R problems: RRips. The same thing can also be found at RTips. This may get replaced by RwikiHome (POSTED: June 3, 2003)

Frank Harrell maintains an extensive page of S-Plus & R materials: S-Plus and R Libraries, Functions, and Documentation (POSTED: November 11, 2003)

The Statistical Consulting Group of UCLA Academic Technology Services has significant S-Plus/R resources available, including their S-Plus/R Portal links to S-Plus/R resources across the web. Especially important are Introductory Tutorials, Course Lecture Notes using S-Plus/R, and Sample Programs for statistical texts, using S-Plus/R to conduct the analyses. (POSTED: July 20, 2004)

A useful resource of example programs for various tasks: R by example by Ajay Shah. (POSTED: August 5, 2004)

A nice web site: R for economists, created by Ajay Shah. See also: A Brief Guide to R for Beginners in Econometrics by Mahmood Arai . (POSTED: August 5, 2004)

Return to Contents

Free General-Purpose Statistical Software

Here is a site devoted solely to: Free Statistical Software. Another site listing: NEWS On Free Statistical Software, such as new versions, etc. It is linked to pages allowing access and downloading of the Free Statistical Software and Free Mathematics Software. (POSTED: October 18, 2003 & September 26, 2004)

This site organizes access to free, web-based software so as to constitute a free general statistics package: Free Web-Based Software. (POSTED: September 17, 2004)

OpenStat is a comprehensive, menu-driven statistical package developed by retired professor Bill Miller. It's menu's are similar to those of SPSS, so experience using it should be transferable. Comes with 350+ page manual. (POSTED: July 30, 2004)

MicrOsiris is a formerly commercial program that is now available for free (though a $10 contribution is requested). In addition to a compressive suite of statistics, it has special procedures for ordination, and data mining. (POSTED: July ,30 2004)

UNESCO has developed WinIDAMS, a general-purpose package for "the validation, manipulation and statistical analysis of data." It contains quite a number of fairly advanced techniques, including factor analysis, correspondence analysis, cluster analysis, multidimensional scaling, multiple classification analysis, and partial order scoring. (PUBLISHED and POSTED: October , 2003)

OpenEpi is "a project to create open source software for Epidemiologic statistics in JavaScript and HTML. OpenEpi can be run from a web site or downloaded and run without a web connection." (POSTED: June 8, 2005)

Scilab is a free scientific computation package, with much of the functionality of Matlab. (POSTED: June 7, 2003)

Scilab includes hundreds of mathematical functions with the possibility to add interactively programs from various languages (C, Fortran...). It has sophisticated data structures (including lists, polynomials, rational functions, linear systems...), an interpreter and a high level programming language.

Return to Contents

Specialized Statistical Software

Markov chain Monte Carlo (MCMC) analyses for Bayesian inference Using Gibbs Sampling can be conducted using software from The BUGS Project (POSTED: May 27, 2003)

M C Neale has developed Mx, Mx is a matrix algebra interpreter and numerical optimizer for structural equation modeling and other types of statistical modeling of data. Mx is available with a GUI for Windows. (POSTED: May 25, 2003)

Ecological Data Analysis: Exploratory and Euclidean methods in Environmental Sciences is freely available for PC and PowerMac. It contains many exploratory techniques, including PCA, PLS,correspondence analysis,spatial data analysis, etc. Analyses can be conducted online, or th program can be dowloaded. ADE-4 There is also an ADE-4 package for R(POSTED: 4/23/03)

For exploratory analyses, by Eugen Horber, EDA Software [Incl: Correspondence analysis, Procrustes rotation, biplot] Free demo. (POSTED: 4/23/03)

Political scientist Gary King has a number of specialized programs. These include programs for missing data; logistic regression with rare events; survey analysis with anchoring vignettes for cross-cultural comparison; programs for clearer presentation of statistical findings; and, of course, ecological inference. He also has an extensive set of papers available. (POSTED: November 11, 2003)

For Social Network Analysis, UCINET 6 can be tried for 30 days for free. See also the 151 page: Introduction to Social Network Methods by Robert A. Hanneman. For more links to Social Network Analysis sites and software, go to my BGSP Research Resources in the Social Science section. [You may have to schroll down.] (POSTED: August 6, 2004)

Return to Contents

Mathematics & Mathematical Software

An excellent site for mathematics: MathWorld. (POSTED: August 5, 2004)

Maxima a computer algebra system that is a descendant of DOE Macsyma is available free under the GPL. (PUBLISHED and POSTED: September 26, 2004)

Of use when one is doing serious calculations are Computer Algebra Systems (a CAS is a program that can be used to do symbolic manipulation of mathematical expressions). The most famous of these is Mathematica, which is fairly expensive, unless one is a student. A free such system is Yacas (Yet Another Computer Algebra System). While nowhere near as developed as Mathematica, the price is right. (POSTED: December 27, 2003)

For numerical calculation, GNU Octave "is a high-level language, primarily intended for numerical computations. It provides a convenient command line interface for solving linear and nonlinear problems numerically, and for performing other numerical experiments using a language that is mostly compatible with Matlab. It may also be used as a batch-oriented language." User contributed Octave scripts are available at Octave codes, part of the Ideas system at Boston College. (POSTED: December 27, 2003)

Calcugator is an online or downloadable program for mat, plotting, and programming. It has over 200 functions. The authors claim "As a calculator it excels in simplicity of use and rivals programs like MATLAB." (POSTED: December 28, 2003)

Return to Contents

Web-Based Calculation

There are now extensive Web Pages Performing Statistical Calculations. Here is a list of hundreds of them: Web Pages that Perform Statistical Calculations!. It also includes Materials to help you Select the Right Statistical Test, Expression Evaluators, Scientific Calculators, programs for Symbolic Calculation, and Function and Data Plotters. WARNING: There is no guarantee that the calculations are correct! Use at your own risk! If you don't know what you are doing, beware! (POSTED: May 23, 2003)

Return to Contents

Statistical Tables (incl. Power Tables)

Victor L. Bissonnette has a number of statistical tables available for download. They are in pdf format. these include power tables for t-tests, correlations, ANOVA, etc.. (POSTED: December 29, 2003)

Return to Contents

Data Sources

The Data and Story Library "is an online library of datafiles and stories that illustrate the use of basic statistics methods." (POSTED: November 10, 2003)

A Casebook for a First Course in Statistics and Data Analysis by Samprit Chatterjee, Mark S. Handcock, &Jeffrey S. Simonoff contains "examples drawn from many fields, where statistical analysis is needed to answer a particular question." It includes 38 data files in various formats. (POSTED: December 30, 2003)

The University of Massachusetts, Amherst has a page with datasets illustrating many statistical techniques, from ANOVA to survival, factor, cluster, & time-series analysis. These are an excellent resource for teaching statistics: Statistical Datasets. "To download a dataset using a text only browser, such as Lynx, use the PRINT and SAVE command. Using Netscape or IE, from the File menu select SAVE AS. On the Save As dialog select File Type Text. If that is not a choice, choose All (*.*) and be sure the file extensoin is .dat. Some versions of IE go directly to the Save dialog." (POSTED: November 11, 2003)

An extensive collection of Datasets are available here, with an emphasis on S+ & R. Also contains links to other statistical data collections. Here is another page with extensive links to data sets, with an emphasis on Canadian data, but many general links.. (POSTED: November 11, 2003)

The Harvard-MIT Virtual Data Center is creating free software "to enable the sharing of quantitative research data, and the development of distributed virtual collections of data and documentation." This software is currently in use at the Harvard MIT Data Center VDC which makes an extensive collection of data sets available online. You can perform search for topics. A search for tobacco yielded 10 data sets. Some of the data sets from the Murray Research Center collection of longitudinal data are available here. Others are available by request from the Center. (POSTED: November 11, 2003)

Return to Contents

Statistics & Social Issues

The British Radical Statistics Group has a web page: Radical Statistics, who "believe that statistics can be used to support radical campaigns for progressive social change. Statistics should inform, not drive policies. Social problems should not be disguised by technical language." They also have an online journal: Radical Statistics: The Journal going back many years. (POSTED: November 9, 2004)

Return to Contents

Statistical Humor

Gary C. Ramseyer has created the: First Internet Gallery of Statistics Jokes. (POSTED: November 9, 2005)

A new, invaluable resource: Journal of Obnoxious Statistics. (POSTED: January 24, 2006)

Return to Contents

Complied by Stephen Soldz
Boston Graduate School of Psychoanalysis
1581 Beacon St.
Brookline, MA 02446


Also check out my other web sites:
BGSP Research Resources
Resources for the Study of Violence
& Psychoanalytic Resources Online

If you find this information useful,
or if you have other ideas for what should be included,
please let us know by sending us e-mail

doteasy.com - free web hosting. Free hosting with no banners.