Specifying the functional form of a covariate is a fundamental part of developing a regression model. The choice to include a variable as continuous, categorical, or as a spline can be determined by model fit. This paper offers an efficient and user-friendly SAS^® macro (%SPECI) to help analysts determine how best to specify the appropriate functional form of a covariate in a linear, logistic, and survival analysis model. For each model, our macro provides a graphical and statistical single-page comparison report of the covariate as a continuous, categorical, and restricted cubic spline variable so that users can easily compare and contrast results. The report includes the residual plot and distribution of the covariate. You can also include other covariates in the model for multivariable adjustment. The output displays the likelihood ratio statistic, the Akaike Information Criterion (AIC), as well as other model-specific statistics. The %SPECI macro is demonstrated using an example data set. The macro includes the PROC REG, PROC LOGISTIC, PROC PHREG, PROC REPORT, PROC SGPLOT, and more procedures in SAS^® 9.4.

Read the paper (PDF) | Download the data file (ZIP)

Duplicates in a clinical trial or survey database could jeopardize data quality and integrity, and they can induce biased analysis results. These complications often happen in clinical trials, meta analyses, and registry and observational studies. Common practice to identify possible duplicates involves sensitive personal information, such as name, Social Security number (SSN), date of birth, address, telephone number, etc. However, access to this sensitive information is limited. Sometimes, it is even restricted. As a measure of data quality control, a SAS^® program was developed to identify duplicated individuals using non-sensitive information, such as age, gender, race, medical history, vital signs, and laboratory measurements. A probabilistic approach was used by calculating weights for data elements used to identify duplicates based on two probabilities (probability of agreement for an element among matched pairs and probability of agreement purely by chance among non-matched pairs). For elements with categorical values, agreement was defined as matching pairs sharing the same value. For elements with interval values, agreement was defined as matching values within 1% of measurement precision range. Probabilities used to compute matching element weights were estimated using an expectation-maximization (EM) algorithm. The method was then tested on a survey and clinical trial data from hypertension studies.

SAS/ACCESS^® software grants access to data in third-party database management systems (DBMS), but how do you access data in DBMS not supported by SAS/ACCESS products? The introduction of the GROOVY procedure in SAS^® 9.3 lets you retrieve this formerly inaccessible data through a JDBC connection. Groovy is an object-oriented, dynamic programming language executed on the Java Virtual Machine (JVM). Using Microsoft Azure HDInsight as an example, this paper demonstrates how to access and read data into a SAS data set using PROC GROOVY and a JDBC connection.

Hierarchical models, also known as random-effects models, are widely used for data that consist of collections of units and are hierarchically structured. Bayesian methods offer flexibility in modeling assumptions that enable you to develop models that capture the complex nature of real-world data. These flexible modeling techniques include choice of likelihood functions or prior distributions, regression structure, multiple levels of observational units, and so on. This paper shows how you can fit these complex, multilevel hierarchical models by using the MCMC procedure in SAS/STAT^® software. PROC MCMC easily handles models that go beyond the single-level random-effects model, which typically assumes the normal distribution for the random effects and estimates regression coefficients. This paper shows how you can use PROC MCMC to fit hierarchical models that have varying degrees of complexity, from frequently encountered conditional independent models to more involved cases of modeling intricate interdependence. Examples include multilevel models for single and multiple outcomes, nested and non-nested models, autoregressive models, and Cox regression models with frailty. Also discussed are repeated measurement models, latent class models, spatial models, and models with nonnormal random-effects prior distributions.

Data is generated every second. The term big data refers to the volume, variety, and velocity of data that is being produced. Now woven into every sector, its size and complexity has left organizations faced with difficulties in being able to create, manipulate, and manage big data. This research identifies and reviews a range of big data techniques within SAS^®, highlighting the fundamental opportunities that SAS provides for overcoming a variety of business challenges. Insurance is a data-dependent industry. This research focuses on understanding what SAS can offer to insurance companies and how it could interact with existing customer databases and online, user-generated content. A range of data sources have been identified for this purpose. The research demonstrates how models can be built based on existing relationships found in past data and then used to identify prospective customers. Principal component analysis, cluster analysis, and neural networks are all considered. You will learn how these techniques can be used to help capture valuable insight, create firm relationships, and support customer feedback. Whether it is prescriptive, predictive, descriptive, or diagnostic analytics, harnessing big data can add background and depth, providing insurance companies with a more complete story. You will see that you can reduce the complexity and dimensionality of data, provide actionable intelligence, and essentially make more informed business decisions.

As you know, real world data (RWD) provides highly valuable and practical insights. But as valuable as RWD is, it still has limitations. It is encounter-based, and we are largely blind to what happens between encounters in the health-care system. The encounters generally occur in a clinical setting that might not reflect actual patient experience. Many of the encounters are subjective interviews, observations, or self-reports rather than objective data. Information flow can be slow (even real time is not fast enough in health care anymore). And some data that could be transformative cannot be captured currently. Select Internet of Things (IoT) data can fill the gaps in our current RWD for certain key conditions and provide missing components that are key to conducting Analytics of Healthcare Things (AoHT), such as direct, objective measurements; data collected in usual patient settings rather than artificial clinical settings; data collected continuously in a patient s setting; insights that carry greater weight in Regulatory and Payer decision-making; and insights that lead to greater commercial value. Teradata has partnered with an IoT company whose technology generates unique data for conditions impacted by mobility or activity. This data can fill important gaps and provide new insights that can help distinguish your value in your marketplace. Join us to hear details of successful pilots that have been conducted as well as ongoing case studies.

Correlated data is extensively used across disciplines when modeling data with any type of correlation that might exist among observations due to clustering or repeated measurements. When modeling clustered data, hierarchical linear modeling (HLM) is a popular multilevel modeling technique that is widely used in different fields such as education and health studies (Gibson and Olejnik, 2003). A typical example of multilevel data involves students nested within classrooms that behave similarly due to shared situational factors. Ignoring their correlation might result in underestimated standard errors and inflated type-I error (Raudenbush and Bryk, 2002). When modeling longitudinal data, many studies have been conducted on continuous outcomes. However, fewer studies on discrete responses over time have been completed. These studies require models within conditional, transitional, and marginal models (Fitzmaurice et al., 2009). Examples of such models that enable researchers to account for the autocorrelation among repeated observations include generalized linear mixed model (GLMM), generalized estimating equations (GEE), alternating logistic regression (ALR), and fixed effects with conditional logit analysis. This study explores the aforementioned methods as well as several other correlated modeling options for longitudinal and hierarchical data within SAS^® 9.4 using real data sets. These procedures include PROC GLIMMIX, PROC GENMOD, PROC NLMIXED, PROC GEE, PROC PHREG, and PROC MIXED.

It's essential that SAS^® users enhance their skills to implement best-practice programming techniques when using Base SAS^® software. This presentation illustrates core concepts with examples to ensure that code is readable, clearly written, understandable, structured, portable, and maintainable. Attendees learn how to apply good programming techniques including implementing naming conventions for data sets, variables, programs, and libraries; code appearance and structure using modular design, logic scenarios, controlled loops, subroutines and embedded control flow; code compatibility and portability across applications and operating platforms; developing readable code and program documentation; applying statements, options, and definitions to achieve the greatest advantage in the program environment; and implementing program generality into code to enable its continued operation with little or no modifications.

A/B testing is a form of statistical hypothesis testing on two business options (A and B) to determine which is more effective in the modern Internet age. The challenge for startups or new product businesses leveraging A/B testing are two-fold: a small number of customers and poor understanding of their responses. This paper shows you how to use the IML and POWER procedures to deal with the reassessment of sample size for adaptive multiple business stage designs based on conditional power arguments, using the data observed at the previous business stage.

This presentation discusses the options for including continuous covariates in regression models. In his book, 'Clinical Prediction Models,' Ewout Steyerberg presents a hierarchy of procedures for continuous predictors, starting with dichotomizing the variable and moving to modeling the variable using restricted cubic splines or using a fractional polynomial model. This presentation discusses all of the choices, with a focus on the last two. Restricted cubic splines express the relationship between the continuous covariate and the outcome using a set of cubic polynomials, which are constrained to meet at pre-specified points, called knots. Between the knots, each curve can take on the shape that best describes the data. A fractional polynomial model is another flexible method for modeling a relationship that is possibly nonlinear. In this model, polynomials with noninteger and negative powers are considered, along with the more conventional square and cubic polynomials, and the small subset of powers that best fits the data is selected. The presentation describes and illustrates these methods at an introductory level intended to be useful to anyone who is familiar with regression analyses.

Until recently, psychometric analyses of test data within the Item Response Theory (IRT) framework were conducted using specialized, commercial software. However, with the inclusion of the IRT procedure in the suite of SAS^® statistical tools, SAS users can explore the psychometric properties of test items using modern test theory or IRT. Considering the item as the unit of analysis, the relationship between test items and the constructs they measure can be modeled as a function of an unobservable or latent variable. This latent variable or trait (for example, ability or proficiency), vary in the population. However, when examinees having the same trait level do not have the same probability to answering correctly or endorsing an item, we said that such an item might be functioning differently or exhibiting differential item functioning or DIF (Thissen, Steinberg, and Wainer, 2012). This study introduces the implementation of PROC IRT for conducting a DIF analysis for graded responses, using Samejima's graded response model (GRM; Samejima, 1969, 2010). The effectiveness of PROC IRT for evaluation of DIF items is assessed in terms of the Type I error and statistical power of the likelihood ratio test for testing DIF in graded responses.

Randomized control trials have long been considered the gold standard for establishing causal treatment effects. Can causal effects be reasonably estimated from observational data too? In observational studies, you observe treatment T and outcome Y without controlling confounding variables that might explain the observed associations between T and Y. Estimating the causal effect of treatment T therefore requires adjustments that remove the effects of the confounding variables. The new CAUSALTRT (causal-treat) procedure in SAS/STAT^® 14.2 enables you to estimate the causal effect of a treatment decision by modeling either the treatment assignment T or the outcome Y, or both. Specifically, modeling the treatment leads to the inverse probability weighting methods, and modeling the outcome leads to the regression methods. Combined modeling of the treatment and outcome leads to doubly robust methods that can provide unbiased estimates for the treatment effect even if one of the models is misspecified. This paper reviews the statistical methods that are implemented in the CAUSALTRT procedure and includes examples of how you can use this procedure to estimate causal effects from observational data. This paper also illustrates some other important features of the CAUSALTRT procedure, including bootstrap resampling, covariate balance diagnostics, and statistical graphics.

Model validation is an important step in the model building process because it provides opportunities to assess the reliability of models before their deployment. Predictive accuracy measures the ability of the models to predict future risks, and significant developments have been made in recent years in the evaluation of survival models. SAS/STAT^® 14.2 includes updates to the PHREG procedure with a variety of techniques to calculate overall concordance statistics and time-dependent receiver operator characteristic (ROC) curves for right-censored data. This paper describes how to use these criteria to validate and compare fitted survival models and presents examples to illustrate these applications.

SAS/STAT^® software has several procedures that estimate parameters from generalized linear models designed for both continuous and discrete response data (including proportions and counts). Procedures such as LOGISTIC, GENMOD, GLIMMIX, and FMM, among others, offer a flexible range of analysis options to work with data from a variety of distributions and also with correlated or clustered data. SAS^® procedures can also model zero-inflated and truncated distributions. This paper demonstrates how statements from PROC NLMIXED can be written to match the output results from these procedures, including the LS-means. Situations arise where the flexible programming statements of PROC NLMIXED are needed for other situations such as zero-inflated or hurdle models, truncated counts, or proportions (including legitimate zeros) that have random effects, and also for probability distributions not available elsewhere. A useful application of these coding techniques is that programming statements from NLMIXED can often be directly transferred into PROC MCMC with little or no modification to perform analyses from a Bayesian perspective with these various types of complex models.

Cumulative logistic regression models are used to predict an ordinal response. They have the assumption of proportional odds. Proportional odds means that the coefficients for each predictor category must be consistent or have parallel slopes across all levels of the response. This paper uses a sample data set to demonstrate how to test the proportional odds assumption. It shows how to use the UNEQUALSLOPES option when the assumption is violated. A cumulative logistic regression model is built, and then the performance of the model on a test set is compared to the performance of a generalized multinomial model. This shows the utility and necessity of the UNEQUALSLOPES option when building a cumulative logistic regression model. The procedures shown are produced using SAS^® Enterprise Guide^® 7.1.

Longitudinal count data arise when a subject's outcomes are measured repeatedly over time. Repeated measures count data have an inherent within subject correlation that is commonly modeled with random effects in the standard Poisson regression. A Poisson regression model with random effects is easily fit in SAS^® using existing options in the NLMIXED procedure. This model allows for overdispersion via the nature of the repeated measures; however, departures from equidispersion can also exist due to the underlying count process mechanism. We present an extension of the cross-sectional COM-Poisson (CMP) regression model established by Sellers and Shmueli (2010) (a generalized regression model for count data in light of inherent data dispersion) to incorporate random effects for analysis of longitudinal count data. We detail how to fit the CMP longitudinal model via a user-defined log-likelihood function in PROC NLMIXED. We demonstrate the model flexibility of the CMP longitudinal model via simulated and real data examples.

The increasing complexity of data in research and business analytics requires versatile, robust, and scalable methods of building explanatory and predictive statistical models. Quantile regression meets these requirements by fitting conditional quantiles of the response with a general linear model that assumes no parametric form for the conditional distribution of the response; it gives you information that you would not obtain directly from standard regression methods. Quantile regression yields valuable insights in applications such as risk management, where answers to important questions lie in modeling the tails of the conditional distribution. Furthermore, quantile regression is capable of modeling the entire conditional distribution; this is essential for applications such as ranking the performance of students on standardized exams. This expository paper explains the concepts and benefits of quantile regression, and it introduces you to the appropriate procedures in SAS/STAT^® software.

The analysis of longitudinal data requires a model that correctly accounts for both the inherent correlation amongst the responses as a result of the repeated measurements, as well as the feedback between the responses and predictors at different time points. Lalonde, Wilson, and Yin (2013) developed an approach based on generalized method of moments (GMM) for identifying and using valid moment conditions to account for time-dependent covariates in longitudinal data with binary outcomes. However, the model developed using this approach does not provide information about the specific relationships that exist across time points. We present a SAS^® macro that extends the work of Lalonde, Wilson, and Yin by using valid moment conditions to estimate and evaluate the relationships between the response and predictors at different time periods. The performance of this method is compared to previously established results.

An important component of insurance pricing is the insured location and the associated riskiness of that location. Recently, we have experienced a large increase in the availability of external risk classification variables and associated risk factors by geospatial location. As additional geospatial data becomes available, it is prudent for insurers to take advantage of the new information to better match price to risk. Generalized additive models using penalized likelihood (GAMPL) have been explored as a way to incorporate new location-based information. This type of model can leverage the new geospatial information and incorporate it with traditional insurance rating variables in a regression-based model for rating. In our method, we propose a local regression model in conjunction with our GAMPL model. Our discussion demonstrates the use of the LOESS procedure as well as the GAMPL procedure in a combined solution. Both procedures are in SAS/STAT^® software. We discuss in detail how we built a local regression model and used the predictions from this model as an offset into a generalized additive model. We compare the results of the combined approach to results of each model individually.

When creating statistical models that include multiple covariates (for example, Cox proportional hazards models or multiple linear regression), it is important to address which variables are categorical and continuous for proper analysis and interpretation in SAS^®. Categorical variables, regardless of SAS data type, should be added in the MODEL statement with an additional CLASS statement. In larger models containing many continuous or categorical variables, it is easy to overlook variables that should be added to the CLASS statement. To solve this problem, we have created a macro that uses simple input from the model variables, with PROC CONTENTS and additional logic checks, to create the necessary CLASS statement and to run the desired model. With this macro, variables are evaluated on multiple conditions to see whether they should be considered class variables. Then, they are added automatically to the CLASS statement.

Read the paper (PDF) | View the e-poster or slides (PDF)

Machine Learning algorithms have been available in SAS software since 1979. This session provides practical examples of machine learning applications. The evolution of machine learning at SAS is illustrated with examples of nearest-neighbor discriminant analysis in SAS/STAT PROC DISCRIM to advanced predictive modeling in SAS Enterprise Miner. Machine learning techniques addressed include memory based reasoning, decision trees, neural networks, and gradient boosting algorithms.

In this presentation you will learn the basics of working with nested data, such as students within classes, customers within households, or patients within clinics through the use of multilevel models. Multilevel models can accommodate correlation among nested units through random intercepts and slopes, and generalize easily to 2, 3, or more levels of nesting. These models represent a statistically efficient and powerful way to test your key hypotheses while accounting for the hierarchical nesting of the design. The GLIMMIX procedure is used to demonstrate analyses in SAS.

Because many SAS^® users either work for or own companies that house big data, the threat that malicious software poses becomes even more extreme. Malicious software, often abbreviated as malware, includes many different classifications, ways of infection, and methods of attack. This E-Poster highlights the types of malware, detection strategies, and removal methods. It provides guidelines to secure essential assets and prevent future malware breaches.

Read the paper (PDF) | View the e-poster or slides (PDF)

This workshop provides hands-on experience performing statistical analysis with the Statistics tasks in SAS Studio. Workshop participants will learn to perform statistical analyses using tasks, evaluate which tasks are ideal for different kinds of analyses, edit the generated code, and customize a task.

Another year implementing, validating, securing, optimizing, migrating, and adopting the Hadoop platform. What have been the top 10 accomplishments with Hadoop seen over the last year? We also review issues, concerns, and resolutions from the past year as well. We discuss where implementations are and some best practices for moving forward with Hadoop and SAS^® releases.

Investors usually trade stocks or exchange-traded funds (ETFs) based on a methodology, such as a theory, a model, or a specific chart pattern. There are more than 10,000 securities listed on the US stock market. Picking the right one based on a methodology from so many candidates is usually a big challenge. This paper presents the methodology based on the CANSLIM1 theorem and momentum trading (MT) theorem. We often hear of the cup and handle shape (C&H), double bottoms and multiple bottoms (MB), support and resistance lines (SRL), market direction (MD), fundamental analyses (FA), and technical analyses (TA). Those are all covered in CANSLIM theorem. MT is a trading theorem based on stock moving direction or momentum. Both theorems are easy to learn but difficult to apply without an appropriate tool. The brokers' application system usually cannot provide such filtering due to its complexity. For example, for C&H, where is the handle located? For the MB, where is the last bottom you should trade at? Now, the challenging task can be fulfilled through SAS^®. This paper presents the methods on how to apply the logic and graphically present them though SAS. All SAS users, especially those who work directly on capital market business, can benefit from reading this document to achieve their investment goals. Much of the programming logic can also be adopted in SAS finance packages for clients.

In the increasingly competitive environment for banks and credit unions, every potential advantage should be pursued. One of these advantages is to market additional products to your existing customers rather than to new customers, since your existing customers already know (and hopefully trust) you, and you have so much data on them. But how can this best be done? How can you market the right products to the right customers at the right time? Predictive analytics can do this by forecasting which customers have the highest chance of purchasing a given financial product. This paper provides a step-by-step overview of a relatively simple but comprehensive approach to maximize cross-sell opportunities among your customers. We first prepare the data for a statistical analysis. With some basic predictive analytics techniques, we can then identify those members who have the highest chance of buying a financial product. For each of these members, we can also gain insight into why they would purchase, thus suggesting the best way to market to them. We then make suggestions to improve the model for better accuracy.

Meta-analysis is a method for combining multiple independent studies on the same subject or question, producing a single large study with increased accuracy and enhanced ability to detect overall trends and smaller effects. This is done by treating the results of each study as a single observation and performing analysis on the set, while controlling for differences between individual studies. These differences can be treated as either fixed or random effects, depending on context. This paper demonstrates the process and techniques used in meta-analysis using human trafficking studies. This problem has seen increasing interest in the past few years, and there are now a number of localized studies for one state or a metropolitan area. This meta-analysis combines these to begin development of a comprehensive analytic understanding of human trafficking across the United States. Both fixed and random effects are described. All elements of this analysis were performed using SAS^® University Edition.

A microservice architecture prescribes the design of your software application as suites of independently deployable services. In this paper, we detail how you can design your SAS^® 9.4 programs so that they adhere to a microservice architecture. We also describe how you can leverage Many-Task Computing (MTC) in your SAS^® programs to gain a high level of parallelism. Under these paradigms, your SAS code will gain encapsulation, robustness, reusability, and performance. The design principles discussed in this paper are implemented in the SAS^® Infrastructure for Risk Management (IRM) solution. Readers with an intermediate knowledge of Base SAS^® and the SAS macro language will understand how to design their SAS code so that it follows these principles and reaps the benefits of a microservice architecture.

Dynamic social networks can be used to monitor the constantly changing nature of interactions and relationships between people and groups. The size and complexity of modern dynamic networks can make this task extremely challenging. Using the combination of SAS/IML^®, SAS/QC^®, and R, we propose a fast approach to monitor dynamic social networks. A discrepancy score at edge level was developed to measure the unusualness of the observed social network. Then, multivariate and univariate change-point detection methods were applied on the aggregated discrepancy score to identify the edges and vertices that have experienced changes. Stochastic block model (SBM) networks were simulated to demonstrate this method using SAS/IML and R. PROC SHEWHART and PROC CUSUM in SAS/QC and PROC SGRENDER heat maps were applied on the aggregated discrepancy score to monitor the dynamic social network. The combination of SAS/IML, SAS/QC, and R make it an ideal tool to monitor dynamic social networks.

The SAS/IML^® language excels in handling matrices and performing matrix computations. A new feature in SAS/IML 14.2 is support for nonmatrix data structures such as tables and lists. In a matrix, all elements are of the same type: numeric or character. Furthermore, all rows have the same length. In contrast, SAS/IML 14.2 enables you to create a structure that contains many objects of different types and sizes. For example, you can create an array of matrices in which each matrix has a different dimension. You can create a table, which is an in-memory version of a data set. You can create a list that contains matrices, tables, and other lists. This paper describes the new data structures and shows how you can use them to emulate other structures such as stacks, associative arrays, and trees. It also presents examples of how you can use collections of objects as data structures in statistical algorithms.

In item response theory (IRT), the distribution of examinees' abilities is needed to estimate item parameters. However, specifying the ability distribution is difficult, if not impossible, because examinees' abilities are latent variables. Therefore, IRT estimation programs typically assume that abilities follow a standard normal distribution. When estimating item parameters using two separate computer runs, one problem with this approach is that it causes item parameter estimates obtained from two groups that differ in ability level to be on different scales. There are several methods that can be used to place the item parameter estimates on a common scale, one of which is multi-group calibration. This method is also called concurrent calibration because all items are calibrated concurrently with a single computer run. There are two ways to implement multi-group calibration in SAS^®: 1) Using PROC IRT. 2) Writing an algorithm from scratch using SAS/IML^®. The purpose of this study is threefold. First, the accuracy of the item parameter estimates are evaluated using a simulation study. Second, the item parameter estimates are compared to those produced by an item calibration program flexMIRT. Finally, the advantages and disadvantages of using these two approaches to conduct multi-group calibration are discussed.

Multicollinearity can be briefly described as the phenomenon in which two or more identified predictor variables in a multiple regression model are highly correlated. The presence of this phenomenon can have a negative impact on the analysis as a whole and can severely limit the conclusions of the research study. This paper reviews and provides examples of the different ways in which multicollinearity can affect a research project, and tells how to detect multicollinearity and how to reduce it once it is found. In order to demonstrate the effects of multicollinearity and how to combat it, this paper explores the proposed techniques by using the Behavioral Risk Factor Surveillance System data set. This paper is intended for any level of SAS^® user. This paper is also written to an audience with a background in behavioral science or statistics.

A new ODS destination for creating Microsoft Excel workbooks is available starting in the third maintenance release for SAS^® 9.4. This destination creates native Microsoft Excel XLSX files, supports graphic images, and offers other advantages over the older ExcelXP tagset. In this presentation, you learn step-by-step techniques for quickly and easily creating attractive multi-sheet Excel workbooks that contain your SAS^® output. The techniques can be used regardless of the platform on which SAS software is installed. You can even use them on a mainframe! Creating and delivering your workbooks on demand and in real time using SAS server technology is discussed. Using earlier versions of SAS to create multi-sheet workbooks is also discussed. Although the title is similar to previous presentations by this author, this presentation contains new and revised material not previously presented.

Read the paper (PDF) | Download the data file (ZIP)

Multicategory logit models extend the techniques of logistic regression to response variables with three or more categories. For ordinal response variables, a cumulative logit model assumes that the effect of an explanatory variable is identical for all modeled logits (known as the assumption of proportional odds). Past research supports the finding that as the sample size and number of predictors increase, it is unlikely that proportional odds can be assumed across all predictors. An emerging method to effectively model this relationship uses a partial proportional odds model, fit with unique parameter estimates at each level of the modeled relationship only for the predictors in which proportionality cannot be assumed. First used in SAS/STAT^® 12.1, PROC LOGISTIC in SAS^® 9.4 now extends this functionality for variable selection methods in a manner in which all equal and unequal slope parameters are available for effect selection. Previously, the statistician was required to assess predictor non-proportionality a priori through likelihood tests or subjectively through graphical diagnostics. Following a review of statistical methods and limitations of other commercially available software to model data exhibiting non-proportional odds, a public-use data set is used to examine the new functionality in PROC LOGISTIC using stepwise variable selection methods. Model diagnostics and the improvement in prediction compared to a general cumulative model are noted.

Read the paper (PDF) | Download the data file (ZIP) | View the e-poster or slides (PDF)

This paper explores the utilization of medical services, which has a characteristic exponential distribution. Because of this characteristic, a variable generalized linear model can be applied to it to obtain self-managed health plan rates. This approach is different from what is generally used to set the rates of health plans. This new methodology is characterized by capturing qualitative elements of exposed participants that old rate-making methods are not able to capture. Moreover, this paper also uses generalized linear models to estimate the number of days that individuals remain hospitalized. The method is expanded in a project in SAS^® Enterprise Guide^®, in which the utilization of medical services by the base during the years 2012, 2013, 2014, and 2015 (the last year of the base) is compared with the Hospital Cost Index of Variation. The results show that, among the variables chosen for the model, the income variable has an inverse relationship with the risk of health care expenses. Individuals with higher earnings tend to use fewer services offered by the health plan. Male individuals have a higher expenditure than female individuals, and this is reflected in the rate statistically determined. Finally, the model is able to generate tables with rates that can be charged to plan participants for health plans that cover all average risks.

Bayesian inference has become ubiquitous in applied science because of its flexibility in modeling data and advances in computation that allow special methods of simulation to obtain sound estimates when more mathematical approaches are intractable. However, when the sample size is small, the choice of a prior distribution becomes difficult. Computationally convenient choices for prior distributions can overstate prior beliefs and bias the estimates. We propose a simple form of prior distribution, a mixture of two uniform distributions, that is weakly informative, in that the prior distribution has a relatively large standard deviation. This choice leads to closed-form expressions for the posterior distribution if the observed data follow a normal, binomial, or Poisson distribution. The explicit formulas are easily encoded in SAS^®. For a small sample size of 10, we illustrate how to elicit the mixture prior and indicate that the resulting posterior distribution is insensitive to minor misspecification of input values. Weakly informative prior distributions suitable for small sample sizes are easy to specify and appear to provide robust inference.

In a randomized study, subjects are randomly assigned to either a treated group or a control group. Random assignment ensures that the distribution of the covariates is the same in both groups and that the treatment effect can be estimated by directly comparing the outcomes for the subjects in the two groups. In contrast, subjects in an observational study are not randomly assigned. In order to establish causal interpretations of the treatment effects in observational studies, special statistical approaches that adjust for the covariate confounding are required to obtain unbiased estimation of causal treatment effects. One strategy for correctly estimating the treatment effect is based on the propensity score, which is the conditional probability of the treatment assignment given the observed covariates. Prior to the analysis, you use propensity scores to adjust the data by weighting observations, stratifying subjects that have similar propensity scores, or matching treated subjects to control subjects. This paper reviews propensity score methods for causal inference and introduces the PSMATCH procedure, which is new in SAS/STAT^® 14.2. The procedure provides methods of weighting, stratification, and matching. Matching methods include greedy matching, matching with replacement, and optimal matching. The procedure assesses covariate balance by comparing distributions between the adjusted treated and control groups.

Predictive analytics has been evolving in property and casualty insurance for the past two decades. This paper first provides a high-level overview of predictive analytics in each of the following core business operations in the property and casualty (P&C) insurance industry: marketing, underwriting, actuarial pricing, actuarial reserving, and claims. Then, a common P&C insurance predictive modeling technical process in SAS^® dealing with large data sets is introduced. The steps of this process include data acquisition, data preparation, variable creation, variable selection, model building (also known as model fitting), model validation, model testing, and so on. Finally, some successful models are introduced. Base SAS^®, SAS/STAT^® software, SAS^® Enterprise Guide^®, and SAS^® Enterprise Miner are presented as the main tools for this process. This predictive modeling process could be tweaked or directly used in many other industries as the statistical foundations of predictive analytics have large overlaps across P&C insurance, health care, life insurance, banking, pharmaceutical, genetics industries, and so on. This paper is intended for any level of SAS^® user or business people from different industries who are interested in learning about general predictive analytics.

We live in a world of data; small data, big data, and data in every conceivable size between small and big. In today's world, data finds its way into our lives wherever we are. We talk about data, create data, read data, transmit data, receive data, and save data constantly during any given hour in a day, and we still want and need more. So, we collect even more data at work, in meetings, at home, on our smartphones, in emails, in voice messages, sifting through financial reports, analyzing profits and losses, watching streaming videos, playing computer games, comparing sports teams and favorite players, and countless other ways. Data is growing and being collected at such astounding rates, all in the hope of being able to better understand the world around us. As SAS^® professionals, the world of data offers many new and exciting opportunities, but it also presents a frightening realization that data sources might very well contain a host of integrity issues that need to be resolved first. This presentation describes the available methods to remove duplicate observations (or rows) from data sets (or tables) based on the row's values and keys using SAS.

The purpose of this paper is to show a SAS^® macro named %SURVEYGENMOD developed in a SAS/IML^® procedure as an upgrade of macro %SURVEYGLM developed by Silva and Silva (2014) to deal with complex survey design in generalized linear models (GLMs). The new capabilities are the inclusion of negative binomial distribution, zero-inflated Poisson (ZIP) model, zero-inflated negative binomial (ZINB) model, and the possibility to get estimates for domains. The R function svyglm (Lumley, 2004) and Stata software were used as background, and the results showed that estimates generated by the %SURVEYGENMOD macro are close to the R function and Stata software.

Has the rapid pace of SAS/STAT^® releases left you unaware of powerful enhancements that could make a difference in your work? Are you still using PROC REG rather than PROC GLMSELECT to build regression models? Do you understand how the GENMOD procedure compares with the newer GEE and HPGENSELECT procedures? Have you grasped the distinction between PROC PHREG and PROC ICPHREG? This paper will increase your awareness of modern alternatives to well-established tools in SAS/STAT by using succinct, high-level comparisons rather than detailed descriptions to explain the relative benefits of procedures and methods. The paper focuses on alternatives in the areas of regression modeling, mixed models, generalized linear models, and survival analysis. When you see the advantages of these newer tools, you will want to put them into practice. This paper points you to helpful resources for getting started.

There is an industry-wide push toward making workflows seamless and reproducible. Incorporating reproducibility into the workflow has many benefits; among them are increased transparency, time savings, and accuracy. We walk through how to seamlessly integrate SAS^®, LaTeX, and R into a single reproducible document. We also discuss best practices for general principles such as literate programming and version control.

Every visualization tells a story. The effectiveness of showing data through visualization becomes clear as these visualizations will tell stories about differences in US mortality using the National Longitudinal Mortality Study (NLMS) data, using the Public-Use Microdata Samples (PUMS) of 1.2 million cases and 122 thousand records of mortality. SAS^® Visual Analytics is a versatile and flexible tool that easily displays the simple effects of differences in mortality rates between age groups, genders, races, places of birth (native or foreign), education and income levels, and so on. Sophisticated analyses including logistical regression (with interactions), decision trees, and neural networks that are displayed in a clear, concise manner help describe more interesting relationships among variables that influence mortality. Some of the most compelling examples are: Males who live alone have a higher mortality rate than females. White men have higher rates of suicide than black men.

Read the paper (PDF) | View the e-poster or slides (PDF)

Does your job require you to create reports in Microsoft Excel on a quarterly, monthly, or even weekly basis? Are you creating all or part of these reports by hand, referencing another sheet containing rows and rows and rows of data? If so, stop! There is a better way! The new ODS destination for Excel enables you to create native Excel files directly from SAS^®. Now you can include just the data you need, create great-looking tabular output, and do it all in a fraction of the time! This paper shows you how to use the REPORT procedure to create polished tables that contain formulas, colored cells, and other customized formatting. Also presented in the paper are the destination options used to create various workbook structures, such as multiple tables per worksheet. Using these techniques to automate the creation of your Excel reports will save you hours of time and frustration, enabling you to pursue other endeavors.

One of the research goals in public health is to estimate the burden of diseases on the US population. We describe burden of disease by analyzing the statistical association of various diseases with hospitalizations, emergency department (ED) visits, ambulatory/outpatient (doctors' offices) visits, and deaths. In this short paper, we discuss the use of large, nationally representative databases, such as those offered by the National Center for Health Statistics (NCHS) or the Agency for Healthcare Research and Quality (AHRQ), to produce reliable estimates of diseases for studies. In this example, we use SAS^® and SUDAAN to analyze the Nationwide Emergency Department Sample (NEDS), offered by AHRQ, to estimate ED visits for hand, foot, and mouth disease (HFMD) in children less than five years old.

Read the paper (PDF) | View the e-poster or slides (PDF)

In many healthcare settings, patients are like customers they have a choice. One example is whether to participate in a procedure. In population-based screening in which the goal is to reduce deaths, the success of a program hinges on the patient's choice to accept and comply with the procedure. Like in many other industries, this not only relies on the program to attract new eligible patients to attend for the first time, but it also relies on the ability of the program to retain existing customers. The success of a new customer retention strategy within a breast screening environment is examined by applying a population averaged model (also know as marginal models), which uses generalized estimating equations (GEEs) to account for the lack of independence of the observations. Arguments for why a population average model was applied instead of a mixed effects model (or random effects model) are provided. This business case provides a great introductory session for people to better understand the difference between mixed effects and marginal models, and illustrates how to implement a population average model within SAS^® by using the GENMOD procedure.

Bivariate Cox proportional models are used when we test the association between a single covariate and the outcome. The test repeats for each covariate of interest. SAS^® uses the last category as the default reference. This raises problems when we want to keep using 0 as our reference for each covariate. The reference group can be changed in the CLASS statement. But, if a format is associated with a covariate, we have to use the corresponding format instead of raw numeric data. This problem becomes even worse when we have to repeat the test and manually enter the reference every single time. This presentation demonstrates one way of fixing the problem using the MACRO function and SYMPUT function.

Read the paper (PDF) | Download the data file (ZIP)