WINPEPI updated: computer programs for epidemiologists, and their teaching potential.
Abramson JH. · 2011
Read WINPEPI updated: computer programs for epidemiologists, and their teaching potential. by Abramson JH. (2011) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/europmc-pmc3041648.txt.
Introduction
Abramson described in 2011 an updated set of free statistics programs built for epidemiologists, the specialists who track disease in populations. The suite, called WINPEPI, runs on ordinary computers and can be downloaded at no cost. It has grown steadily and now holds seven programs with well over a hundred modules between them.
Programs cover describing data, comparing two groups, matched comparisons, and two kinds of regression for rates and odds. A finder tool points users to the right module for each task, and manuals give formulas, limits, and references. Because the tools are portable and easy to run, the author stresses their value in teaching. Students can learn core study design and analysis ideas by doing real calculations, not just reading about them. Practicing health workers gain the same benefit in the field.
Catalog record
| CREATOR | Abramson JH. |
|---|---|
| YEAR | 2011 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 34 KB |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
WINPEPI updated: computer programs for epidemiologists, and their teaching potential By Joseph H Abramson Published in: Epidemiologic Perspectives & Innovations : EP+I DOI: 10.1186/1742-5573-8-1 ABSTRACT Background The WINPEPI computer programs for epidemiologists are designed for use in practice and research in the health field and as learning or teaching aids. The programs are free, and can be downloaded from the Internet. Numerous additions have been made in recent years. Implementation There are now seven WINPEPI programs: DESCRIBE, for use in descriptive epidemiology; COMPARE2, for use in comparisons of two independent groups or samples; PAIRSetc, for use in comparisons of paired and other matched observations; LOGISTIC, for logistic regression analysis; POISSON, for Poisson regression analysis; WHATIS, a "ready reckoner" utility program; and ETCETERA, for miscellaneous other procedures. The programs now contain 122 modules, each of which provides a number, sometimes a large number, of statistical procedures. The programs are accompanied by a Finder that indicates which modules are appropriate for different purposes. The manuals explain the uses, limitations and applicability of the procedures, and furnish formulae and references. Conclusions WINPEPI is a handy resource for a wide variety of statistical routines used by epidemiologists. Because of its ready availability, portability, ease of use, and versatility, WINPEPI has a considerable potential as a learning and teaching aid, both with respect to practical procedures in the planning and analysis of epidemiological studies, and with respect to important epidemiological concepts. It can also be used as an aid in the teaching of general basic statistics. INTRODUCTION Numerous procedures have been added to the WINPEPI freeware computer programs for epidemiologists since the publication of a description of the package in 2004 [1]. The number of programs (now all 32-bit) has grown from four to seven, and the total number of modules (most of which provide multiple procedures) from 75 to 122. This paper describes the current version (11.0) of the package, with emphasis on recent enhancements and the package's potential as a learning or teaching aid. WINPEPI is the Windows version of the DOS-based PEPI (an acronym for Programs for EPIdemiologists) package [2], which grew from a set of programs for programmable pocket calculators published in 1983 [3]. It is not a complete compendium of statistical routines for epidemiologists. It does not, for example, provide Cox regression, procedures for the study of disease clustering, or propensity-score methods. Nor does it provide data-management facilities. But it is a handy resource for many commonly-used routines and for others that are not very commonly used or not very easily found (or not easily found in a user-friendly format). Reviewers have repeatedly called it a "Swiss army knife" of utilities for epidemiologists and biomedical researchers. THE PROGRAMS The programs are DESCRIBE, COMPARE2, PAIRSetc, WHATIS, and the recently-added LOGISTIC, POISSON, and ETCETERA. They are accompanied by detailed manuals (occupying over 500 pages in all), which explain the uses and limitations of specific procedures, and furnish formulae and references. Describe DESCRIBE now contains 21 modules for use in descriptive epidemiology, including the appraisal of separate samples in comparative studies. It can appraise rates or proportions or categorical data (dichotomous, nominal or ordinal) or numerical data (including survival data), examine a sequence of rates or other values (including the appraisal of seasonal variation), perform direct and indirect standardization, estimate prevalence from a cluster or stratified sample or from pooled samples or by the capture-recapture method, determine required sample sizes, and appraise screening or diagnostic tests (including meta-analytic procedures). Recent enhancements include plotting of epidemic curves using data or a model; probit analysis and a nonparametric counterpart; and improved procedures concerning diagnostic and screening tests - appraisals, comparisons, combination, meta-analyses, chance-corrected measures of validity, stratum-specific likelihood ratios, determination of optimal cutpoints, and determination of required sample size. Compare2 COMPARE2 (now with 30 modules) is for use in comparisons of two independent groups or samples. It can handle both categorical data (dichotomous, nominal or ordinal; including clustered binomial data) and numerical data (including survival times), proportions or odds, risks, and rates, and determine power and sample size for a variety of tests. It can deal with stratified data, permitting the control of confounding by the stratifying variables, and the assessment of heterogeneity indicative of effect modification. It may be used for analyses and meta-analyses of cross-sectional, cohort and case-control studies and trials. Recent enhancements include the drawing of forest plots, and two tests for a skewed funnel plot (for use in meta-analysis); added or modified procedures for comparing trimmed means, for comparing ordered categories, for appraising dose-response relationships and other trends, for comparing variances, for comparing the changes in two groups, for appraising equivalence, for appraising differences in one tail of a distribution, and for analyzing data based on inverse sampling; ridit analysis of a 2 × k table with ordered categories; pairwise comparisons of categories in a 2 by k table; additional measures of effect (omega-squared, eta-squared, Cohen's d and w, and standardized mean differences and ratios of means); confidence intervals for the generalized odds ratio for ordered categories; pooled results for stratified data in survival analyses, and directly-standardized survival proportions to control for confounding; sample sizes for cluster-randomization trials; and the calculation of required sample size from the results of a previous comparison. Pairsetc PAIRSetc (now with 34 modules) performs comparisons of paired and other matched observations, such as matched-control trials and cohort studies, matched case-control studies, and before-after studies, and reliability studies that compare two or more replicate observations or methods of measurement. The "etc" in its name indicates its ability to deal with matched sets larger than pairs. It may be used for analyses and meta-analyses of cross-sectional, cohort or case-control studies, and trials. Like COMPARE2, it can handle dichotomous, nominal, ordinal, or numerical variables (including survival times), and can analyze stratified data. It can appraise the effect of misclassification, and determine power and sample size for a variety of tests. The many recent enhancements include two-way analysis of variance by ranks, analysis of covariance, monotonic regression analysis, the analysis of paired survival data, assessment of regression to the mean, tests that adjust for this regression, measures of effect (omega-squared, Cohen's f, generalized odds ratio for ordered categories), analysis of clustered matched-pair data, analysis of marginal homogeneity, procedures for using incompletely paired data when comparing proportions, analysis of data obtained by inverse sampling, and the determination of sample sizes for estimating kappa and for McNemar tests. For analyses of agreement, there are added or modified procedures with respect to the analysis of variable as well as fixed numbers of replicate measurements; weighted kappa for multiple ratings; the concurrent appraisal of interrater and intrarater reliability; intraclass correlation coefficients; coefficients of individual agreement; indices of proportionate positive and negative agreement; confidence intervals for percentages of positive and negative agreement; chance-corrected (delta-based) measures of agreement; Bayesian confidence intervals for indices of positive and negative agreement; 95% limits of agreement between methods of measurement; tetrachoric correlation coefficients; correlation when some data are misssing; pairwise comparisons of multiple sets of data; and comparisons of intrarater reliability coefficients. There is also a measure of disagreement between numerical observations. Whatis WHATIS is a "ready reckoner" utility program with four modules. It provides a calculator (expression evaluator) that