Popular Statistical Software Packages Research Articles

Efficient theory development and factor retention criteria: Abandon the ‘eigenvalue greater than one’ criterion

Other things being equal, a theory with fewer constructs is preferable over others. In exploratory factor analysis, a common method used in theory development, the most popular factor retention criterion used in marketing is the eigenvalue greater than one rule. Its use often results in over extraction, which leads to the development of less than parsimonious theories. Even the use of confirmatory factor analysis fails to detect the presence of these superfluous constructs. Although several more accurate criteria exist, they are not discussed in major marketing research texts, journals, and popular statistical software packages. In this paper, we appraise popular factor retention practices in marketing, demonstrate how they may lead to the development of inefficient theories, draw attention to a number of resources for choosing appropriate retention criteria, and develop an easy-to-use Web-based engine to effortlessly implement one such method, parallel analysis.

SPSS and SAS programs for generalizability theory analyses

The identification and reduction of measurement errors is a major challenge in psychological testing. Most investigators rely solely on classical test theory for assessing reliability, whereas most experts have long recommended using generalizability theory instead. One reason for the common neglect of generalizability theory is the absence of analytic facilities for this purpose in popular statistical software packages. This article provides a brief introduction to generalizability theory, describes easy to use SPSS, SAS, and MATLAB programs for conducting the recommended analyses, and provides an illustrative example, using data (N = 329) for the Rosenberg Self-Esteem Scale. Program output includes variance components, relative and absolute errors and generalizability coefficients, coefficients for D studies, and graphs of D study results.

Joint Modelling of Repeated Transitions in Follow-up Data - A Case Study on Breast Cancer Data

In longitudinal studies where time to a final event is the ultimate outcome often information is available about intermediate events the individuals may experience during the observation period. Even though many extensions of the Cox proportional hazards model have been proposed to model such multivariate time-to-event data these approaches are still very rarely applied to real datasets. The aim of this paper is to illustrate the application of extended Cox models for multiple time-to-event data and to show their implementation in popular statistical software packages. We demonstrate a systematic way of jointly modelling similar or repeated transitions in follow-up data by analysing an event-history dataset consisting of 270 breast cancer patients, that were followed-up for different clinical events during treatment in metastatic disease. First, we show how this methodology can also be applied to non Markovian stochastic processes by representing these processes as "conditional" Markov processes. Secondly, we compare the application of different Cox-related approaches to the breast cancer data by varying their key model components (i.e. analysis time scale, risk set and baseline hazard function). Our study showed that extended Cox models are a powerful tool for analysing complex event history datasets since the approach can address many dynamic data features such as multiple time scales, dynamic risk sets, time-varying covariates, transition by covariate interactions, autoregressive dependence or intra-subject correlation.

A Review of Software Packages for Data Mining

We present to the statistical community an overview of five data mining packages with the intent of leaving the reader with a sense of the different capabilities, the ease or difficulty of use, and the user interface of each package. We are not attempting to perform a controlled comparison of the algorithms in each package to decide which has the strongest predictive power, but instead hope to give an idea of the approach to predictive modeling used in each of them. The packages are compared in the areas of descriptive statistics and graphics, predictive models, and association (market basket) analysis.As expected, the packages affiliated with the most popular statistical software packages (SAS and SPSS) provide the broadest range of features with remarkably similar modeling and interface approaches, whereas the other packages all have their special sets of features and specific target audiences whom we believe each of the packages will serve well. It is essential that an organization considering the purchase of a data mining package carefully evaluate the available options and choose the one that provides the best fit with its particular needs.

A review of ordinal regression models applied on health-related quality of life assessments.

There has been increasing emphasis in medical research on the design and analysis of quality of life scales. Many quality of life scales are ordinal and statistical methods such as ordinal regression models have been reviewed on a number of occasions. However, when such models are applied, the way the data have been generated is often overlooked. In this paper we illustrate the use of ordinal regression models, in particular the proportional odds model, the partial proportional odds model and the stereotype model in the MRC Cognitive Function and Ageing Study (MRC CFAS). The partial proportional odds and the stereotype models are often under-utilized, perhaps due to the lack of available software. However, in this paper, analysis based on these models has been carried out using the popular statistical software package SAS and macros devised in SAS. Furthermore, bootstrapping techniques have been applied to obtain valid estimates of the standard errors of the parameters in the stereotype model. Strikingly different results were obtained using the different ordinal regression models. We conclude that the way the data have been generated is particularly important for the analysis of quality of life assessments. Different methods of generating scores yield data with different properties. It is now possible to fit a variety of ordinal regression models and so select the appropriate one that correctly models the data.

SPSS and SAS programs for determining the number of components using parallel analysis and velicer's MAP test.

Popular statistical software packages do not have the proper procedures for determining the number of components in factor and principal components analyses. Parallel analysis and Velicer's minimum average partial (MAP) test are validated procedures, recommended widely by statisticians. However, many researchers continue to use alternative, simpler, but flawed procedures, such as the eigenvalues-greater-than-one rule. Use of the proper procedures might be increased if these procedures could be conducted within familiar software environments. This paper describes brief and efficient programs for using SPSS and SAS to conduct parallel analyses and the MAP test.

Should the Median Test be Retired from General Use?

Although several authors have indicated that the median test has low power in small samples, it continues to be presented in many statistical textbooks, included in a number of popular statistical software packages, and used in a variety of application areas. We present results of a power simulation study that shows that the median test has noticeably lower power, even for the double exponential distribution for which it is asymptotically most powerful, than other readily available rank tests. We suggest that the median test be “retired” from routine use and recommend alternative rank tests that have superior power over a relatively large family of symmetric distributions.

Computing estimates of incidence, including lifetime risk: Alzheimer's disease in the Framingham Study. The Practical Incidence Estimators (PIE) macro.

The incidence of disease is estimated in medical and public health applications using various different techniques presented in the statistical and epidemiologic literature. Many of these methods have not yet made their way to popular statistical software packages and their application requires custom programming. We present a macro written in the SAS macro language that produces several estimates of disease incidence for use in the analysis of prospective cohort data. The development of the Practical Incidence Estimators (PIE) Macro was motivated by research in Alzheimer's Disease (AD) in the Framingham Study in which the development of AD has been prospectively assessed over an observation period of 24 years. The PIE Macro produces crude and age-specific incidence rates, overall and stratified by the levels of a grouping variable. In addition, it produces age-adjusted rates using direct standardization to the combined group. The user specifies the width of the age groups and the number of levels of the grouping variable. The PIE macro produces estimates of future risk for user-defined time periods and the remaining lifetime risk conditional on survival event-free to user-specified ages. This allows the user to investigate the impact of increasing age on the estimate of remaining lifetime risk of disease. In each case, the macro provides estimates based on traditional unadjusted cumulative incidence, and on cumulative incidence adjusted for the competing risk of death. These estimates and their respective standard errors, are provided in table form and in an output data set for graphing. The macro is designed for use with survival age as the time variable, and with age at entry into the study as the left-truncation variable; however, calendar time can be substituted for the survival time variable and the left-truncation variable can simply be set to zero. We illustrate the use of the PIE macro using Alzheimer's Disease incidence data collected in the Framingham Study.

Regression analysis of data with repeated measurements using the method of successive differences

This paper discusses the first-order differencing approach to longitudinal data with measurements on individual subjects taken at irregular time points. Unlike the two-stage approach to this problem, which first fits a line or curve to each individual's growth pattern and then regresses the fitted line or curve on the covariates of interest, subjects with only two-observations can still contribute to the estimation of parameters. Another major advantage is that any time trend of the response or any covariate effects on the response over a time interval within the total study period can be detected. Such an analysis can be very useful in clinical trials studies, especially in trials involving testing a new drug or therapy. The therapeutic effect of a drug or therapy may decrease with time and it is important to know the time period in which the treatment is most effective so that the shortest treatment duration can be determined to reduce the risks of toxic side-effects. Since estimation and inference of model parameters are performed under the framework of the multiple regression model, the procedures are easily implemented using popular statistical software packages. The estimators are asymptotically normal even when the assumed correlation structure is not met by the underlying data. The methods are illustrated with data from a study of primary anaplastic brain tumors.

Covariance Analysis for Split-Plot and Split-Block Designs

Abstract Commentaries are informative essays dealing with viewpoints of statistical practice, statistical education, and other topics considered to be of general interest to the broad readership of The American Statistician. Commentaries are similar in spirit to Letters to the Editor, but they involve longer discussions of background, issues, and perspectives. All commentaries will be refereed for their merit and compatibility with these criteria. Proper methodology for the analysis of covariance for experiments designed in a split-plot or split-block design is not found in the statistical literature. Analyses for these designs are often performed incompletely or even incorrectly. This is especially true when popular statistical computer software packages are used for the analysis of these designs. This article provides several appropriate models, ANOVA tables, and standard errors for comparisons from experiments arranged in a standard split-plot, split–split-plot, or split-block design where a covariate has been measured on the smallest size experimental unit.

Teaching spatial statistics to geographers using MINITAB

Abstract This paper surveys computer software for spatial statistics, emphasising its educational advantages. Initially this software was written in FORTRAN code accessing the International Mathematical and Statistics Library (IMSL) package of commercial subroutines (a counterpart to the NAG package in the United Kingdom). Accompanying programming skills proved too demanding for many geography students and staff. Consequently, spatial statistics analysis was approached within the context of standard commercial software packages. MINITAB was explored, first in its PC and then in its mainframe versions. Unfamiliarity of North American geographers with this package led to parallel code development for the popular statistical software package, SAS. Practices and pitfalls of these endeavours with regard to classroom situations are reported. A demonstration of MINITAB‐PC 7.2 code is presented. Real world examples are drawn from experiences of postgraduate‐level spatial statistics classes both in studying the co...