Jay Magidson

• Statistical Innovations

The practice of latent class (LC) modeling using a bias-adjusted three-step approach has become widely popular. However, the current three-step approach has one important drawback – its key assumption of conditional independence between external variables and latent class indicators is often violated in practice, such as when a (nominal) covariate...

Purpose: To quantify the influence of spectral domain optical coherence tomography (SDOCT) on decision-making in patients with suspected glaucoma. Methods: A prospective cross-sectional study involving 40 eyes of 20 patients referred by community optometrists due to suspected glaucoma. All patients had disc photographs and standard automated perim...

Purpose: To quantify the influence of spectral domain optical coherence tomography (SDOCT) on decision-making in patients with suspected glaucoma. Methods: A prospective cross-sectional study involving 40 eyes of 20 patients referred by community optometrists due to suspected glaucoma. All patients had disc photographs and standard automated perime...

Background: Data characterising long-term survivors (LTS) with human epidermal growth factor receptor 2 (HER2)–positive metastatic breast cancer (MBC) are limited. This analysis describes LTS using registHER observational study data. Methods: A latent class modelling (LCM) approach was used to identify distinct homogenous patient groups (or classe...

Tremelimumab (Ticilimumab, Pfizer), is a monoclonal antibody (mAb) targeting cytotoxic T lymphocyte-associated antigen-4 (CTLA-4). Ipilimumab (Yervoy, BMS), another anti-CTLA-4 antibody is approved by the Federal Drug Administration (FDA). Biomarkers are needed to identify the subset of patients who will achieve tumor control with CTLA-4 blockade E...

We introduce a new regression method—called Correlated Component Regression (ccr)—which provides reliable predictions even with near multicollinear data. Near multicollinearity occurs when a large number of correlated predictors and relatively small sample size exists as well as situations involving a relatively small number of correlated predictor...

9080 Background: Tremelimumab (Ticilimumumab, Pfizer), a monoclonal antibody targeting CTLA-4, a T cell inhibitory molecule, has shown activity in metastatic melanoma. Ipilimumab (Yervoy, BMS), another antibody targettingCTLA-4, improves survival relative to a peptide vaccine and is now FDA approved. A minority of patients will achieve durable tumo...

Background: Survival for patients with castration-resistant prostate cancer is highly variable. We assessed the effectiveness of a whole-blood RNA transcript-based model as a prognostic biomarker in castration-resistant prostate cancer. Methods: Peripheral blood was prospectively collected from 62 men with castration-resistant prostate cancer on...

4516 Background: Survival for patients with castration resistant prostate cancer (CRPC) is highly variable. We developed a whole blood RNA transcript-based model as a prognostic biomarker in CRPC. Methods: Peripheral blood was collected from 62 men with CRPC in a training set and from 140 patients with CRPC in a validation set on various treatment...

Identification of differentially expressed genes from 4 melanoma patients (stage IV, with no treatment, M1–M4) and 6 healthy control individuals (C1–C6). A. Heatmap of 63 selected transcripts with 1.8-fold or greater significant differences between healthy controls and melanomas and a high score by comparison replicates. Colored spots indicate sign...

Pathway analysis predicted IL-4 as one of the differentially expressed signaling pathways between normal controls and melanomas. Significant pathways (p<0.05) were analyzed using differentially expressed transcripts (p<0.05, fold >1.8). Analysis was performed using pathway analysis tool from GeneSpringGX 10.0. The figure represents IL-4 signaling p...

Differentially expressed gene identified by microarray analysis. (DOC)

Developing analytical methodologies to identify biomarkers in easily accessible body fluids is highly valuable for the early diagnosis and management of cancer patients. Peripheral whole blood is a "nucleic acid-rich" and "inflammatory cell-rich" information reservoir and represents systemic processes altered by the presence of cancer cells. We con...

A new ensemble dimension reduction regression technique, called Correlated Component Regression (CCR), is proposed that predicts the dependent variable based on K correlated components. For K = 1, CCR is equivalent to the corresponding Naïve Bayes solution, and for K = P, CCR is equivalent to traditional regression with P predictors. An optional st...

This appendix lists a number of books, chapters, and articles that may prove useful to those who wish to learn more about latent class analysis. Although the listing is by no means exhaustive, an effort has been made to include many of the most widely known sources. We have selected a set of general headings with which to separate the various works...

5052 Background: Screening for CaP with PSA testing is limited by a high number of false postives, particularly in the setting of benign prostatic hypertrophy (BPH). The goal of this study was to develop whole blood RNA transcript-based diagnostic tests that improve the diagnosis of CaP over PSA alone. Methods: From August 2006 to October 2008, thr...

Latent Markov modeling is used as an alternative to the Current Population Survey (Census, 2002) reinterviewing methodology for estimating the measure- ment error in the recorded employment status. This alternative methodology, which is implemented in the syntax version the Latent GOLD program, turns out to be a promising new approach for estimatin...

A property of MaxDiff (Maximum Difference Scaling) is that only relative judgments are made regarding the items, which implies that items are placed on separate relative scales for each segment. One can directly compare item preference for a given segment, but comparisons between respondents in different segments may be problematic. In this paper w...

Latent class regression models may be used to identify segments that differ with respect to the contribution of product attributes on their ratings of the associated products. However, such solutions tend be dominated by the overall liking (or the respondents’ response tendency) rather than differences in the liking of the presented products. In th...

This article discusses a modeling framework that links two well-known statistical methods: structural equation modeling (SEM) and latent class or finite mixture modeling. Mixture SEM can either be viewed as a refinement of multivariate normal (MVN) mixtures, where the within-class covariance matrices are smoothed according to a postulated SEM struc...

The views of Eric Bradlow on the issues and a 'wish list' for conjoint analysis are presented. The important issue of stability of the estimated partworths was also listed. Incorporating respondent heterogeneity is the single most important way to assure that the partworth accurately reflects the individual consumer's values, as opposed to just som...

Many constructs that are of interest to social scientists can not be ob-served directly. Examples are preferences, attitudes, behavioral intentions, and personality traits. Such constructs can only be measured indirectly by means of observable indicators, such as questionnaire items designed to elicit responses related to an attitude or preference....

Background and Purpose: A pediatric long-bone fracture classification proposal was developed following the principle of the Mller AO classification long bones for adults and assessed for reliability and accuracy in a series of four pilot agreement studies (classification sessions). Material and Methods: Six surgeons independently classified 136 rad...

A hierarchical extension of the finite mixture model is presented that can be used for the analysis of nested data structures. The model permits a simulta- neous model-based clustering of lower- and higher-level units. Lower-level observa- tions within higher-level units are assumed to be mutually independent given cluster membership of the higher-...

The CHAID algorithm has proven to be an effective approach for ob- taining a quick but meaningful segmentation where segments are defined in terms of demographic or other variables that are predictive of a single categorical crite- rion (dependent) variable. However, response data may contain ratings or purchase history on several products, or, in...

A major goal of data mining is to extract a small number of meaningful "factors" from a larger number of variables available on a database. While traditional factor analysis (FA) offers such a data reduction capability, it is severely limited in prac- tice because it requires all variables to be continuous, and it uses the assumption of multivariat...

An overview is provided of recent developments in the use of latent class (LC) and other types of finite mixture models for classification purposes. Several extensions of existing models are presented. Two basic types of LC models for classification are defined: supervised and unsupervised structures. Their most important special cases are presente...

Discrete choice models have proven to be good methods for predicting market shares for new products based on consumers' expressed preferences between choice alternatives. However, the standard aggregate model fails to take into account the fact that preferences (utilities) differ from one respondent to another (or at least from one segment to anoth...

We propose an alternative method of conducting exploratory latent class analysis that utilizes latent class factor models, and compare it to the more traditional approach based on latent class cluster models. We show that when formulated in terms of R mutually independent, dichotomous latent factors, the LC factor model has the same number of disti...

This chapter presents a way to visualize the effects (odds ratios) in the analysis of categorical outcome data through powerful graphic displays. It presents new graphic representations for logit models and illustrates these graphs using several datasets. Traditional tabular results obtained from the estimation of these models are often complex and...

The multivariable techniques most prevalent in direct marketing research today for modeling response to a mailing (responders versus non-responders) are multiple regression, discriminant analysis, and AID (abbreviations are explained in the Introduction). This article shows why these traditional approaches may provide erroneous and misleading resul...

Examples of some common pitfalls in the analysis of categorical data are discussed in the context of causal interpretation of the results. Though no statistical technique can replace theory, the author shows that log-linear modeling and chi square automatic interaction detection can provide researchers with powerful tools for gaining valuable causa...

Examples of some common pitfalls in the analysis of categorical data are discussed in the context of causal interpretation of the results. Though no statistical technique can replace theory, the author shows that log-linear modeling and chi square automatic interaction detection can provide researchers with powerful tools for gaining valuable causa...

The recent literature on-log-linear models incorrectly implies that the Iterative Proportional Fitting (IPF) algorithm and associated computer programs such as ECTA can only be used to estimate hierarchical (not nonhierarchical) log-linear models. While ECTA and similar programs are designed for the estimation of hierarchical models, it is shown he...

In this paper we derive a measure of dispersion for a nominal variable having k ⩾ 2 categories and compare it with ordinary quantitative variance and with entropy. We then develop two qualitative analogs to the R2 statistic, one based on qualitative variance, the other on entropy. For concreteness, data from the study of The American Soldier by S....

The traditional approach for partialing out the effects of confounding factors in a non equivalent control group situation is to calculate a partial correlation coefficient or a (partial) regression coefficient controlling for one or more "covariates." Since the covariates generally are imperfect measures of the factors, this procedure will typical...

An overview is provided of recent developments in the use of latent class (LC) and other types of %nite mixture models for classi%cation purposes. Several extensions of existing models are presented. Two basic types of LC models for classi%cation are de%ned: supervised and unsupervised structures. Their most important special cases are presented an...

The most important predictor in a regression model may be a suppressor variable which does not predict the outcome variable directly but improves the overall prediction by enhancing the effects of other predictors in the model. The most important gene in a 9-gene model for early detection of prostate cancer is the gene SP1, whose mean is not signif...

Typescript. Thesis (M.S.)--University of Wisconsin--Madison, 1971. Includes bibliographical references (leaves 51-53).

Incluye índice