SAS Codes Research Articles

Predicting the body weight of indigenous goat breeds from morphological measurements using the classification and regression tree (CART) data mining algorithm

Classification and regression tree (CART) is a tree-based data mining algorithm that develops a model to predict an outcome. This study purposed to create a model to predict the body weight (BWT) of Red Sokoto (RS), Sahel (SH), and West African Dwarf (WAD) goats using morphological measurements (such as body length, BL; head girth, HG; head width, HDW; face length, FAL; height at wither, HTW; rump length, RL; shoulder width, SW; rump width, RW; and rump height, RH). In total, 600 goats were used for this study (200 each of RS, SH, and WAD goats). Pearson?s Moment Correlation was used to evaluate the degree of association between BWT and each morphological measurement. Concomitantly, CART analysis was performed to estimate which independent variable (morphological measurements) played a considerable role in the BWT (dependent variable) prediction. In RS and WAD goats, a positive and statistically significant (p < 0.0001) correlation was observed between BWT and each morphological measurement. However, in SH goats, both positive and negative statistically significant correlations were observed between BWT and morphological measurements. The CART analysis indicated that in RS and WAD goats, HG played a considerable role in BWT prediction, while, in SH goats, BL was considered the most critical independent variable in BWT prediction. Therefore, this study suggests that HG can be used as a prognostic index for BWT estimation in Red Sokoto and West African Dwarf, while BL can be used for Sahel goats. The SAS codes used are available via a GitHub repository (https://github.com/Soullevram/CART).

Application of weighted adjustments of dropout rates in sensitivity analysis of medical repeated measurements data and implementation with SAS

Objective: To explore the application of weighted adjustments of dropout rates in sensitivity analysis of medically repeated measurements data and the implementation with SAS 9.4 software. Methods: By compiling SAS codes, mixed-effects models for repeated measures were used to conduct the covariance analysis of multivariable repeated measurements data. Meanwhile, the overall dropout rate and the dropout rates of each group were used to make weighted adjustments by applying pattern-mixture models, which was considered to be a sensitivity analysis to validate the stability of results. Results: The dropout rates of placebo group, low-dose and high-dose groups were 8.77%, 11.79% and 16.15%, respectively, the differences were significant (P=0.025). The results of mixed-effects models for repeated measures showed the differences of curative effect indicators changes from baselines of between high-dose, low-dose groups and placebo group were significant (P=0.008 and P=0.002). The results of pattern-mixture models considering weighted adjustments of the respective groups' dropout rates were consistent with those of mixed-effects models for repeated measures. Conclusions: The pattern-mixture models considering weighted adjustments of dropout rates can be used in the sensitivity analysis of repeated measurements data. The SAS codes can provide a practical basis for the popularization and application of weighted adjustments of dropout rates in the sensitivity analysis of repeated measurements data.

Improved strategy to collect sensitive data by using negative binomial and negative hypergeometric distribution as randomization devices

In this paper, we improve upon the negative binomial randomized response model by implementing a two-stage randomization process. Using the modified estimator of the proportion of members of a population possessing a sensitive attribute, we claim to achieve both better efficiency and better protection. The findings are verified based on extensive simulation study. Then we improvise the technique of use of negative hypergeometric distribution by introducing two-stage randomization response process. Using our new technique, we also assess the efficiency and protection of the respondents as compared to the negative hypergeometric model through extensive simulation studies. We also provide SAS codes used in the evaluations of the strategies considered.

The SEC Filing Review Process: Insights from Accounting Research

As part of its goal to monitor and enhance the quality of information available to investors, the Securities and Exchange Commission (SEC) reviews companies’ filings to ensure compliance with applicable financial reporting and disclosure requirements. Increased public interest and the substantial costs for both the SEC in reviewing and the companies in responding have led to a rapidly growing body of accounting literature that examines the filing review process. We survey and comment on 80 published and unpublished academic research papers in this literature and identify significant gaps that future research should address. We also summarize the institutional features of the filing review process, aiming to inform future academic studies. Our survey should be of interest to both academics and market participants evaluating the effectiveness and efficiency of the filing review process. To download the SAS codes, please visit https://lcunningham.utk.edu/sec-filing-review-process-sas-codes/

Overview of logistic regression model analysis and application

Logistic regression is a kind of multiple regression method to analyze the relationship between a binary outcome or categorical outcome and multiple influencing factors, including multiple logistic regression, conditional logistic regression, polytomous logistic regression, ordinal logistic regression and adjacent categorical logistic regression. This paper illustrates the basic principle, independent variable selection and assignment, applied condition, model evaluation and diagnosis for multiple logistic regression model. Moreover, the principle and application for polytomous logistic regression and ordinal logistic regression models were also introduced. By providing SAS codes and detailed explanations of the result for an example of obesity, readers could be able to better understand logistic regression model, and apply this method correctly to their research and daily work, so as to improve their capacity of the data analysis.

A new approach for fitting growth models in random environment

Nonlinear growth models are widely employed in Animal sciences for describing growth of various species ofanimals. Nonlinear estimation procedures are generally employed for estimation of parameters. However, onelimitation of these models is that they are applicable only when the data are available at equidistant epochs. Another limitation is that the fluctuations in the system cannot be satisfactorily explained simply by adding an error term to the deterministic formulation. The purpose of this article is to bring to the notice of Animal scientists the new approach of Stochastic differential equation modelling, which is capable of incorporating both the above aspects. The methodology is discussed by considering Gompertz growth model. Relevant SAS codes for fitting the model are developed. Finally, the methodology is illustrated on secondary monthly pig weight data, collected at thepiggery farm of Indian Veterinary Research Institute, Izatnagar, Bareilly, India.

Categorical Omega With Small Sample Sizes via Bayesian Estimation: An Alternative to Frequentist Estimators.

When item scores are ordered categorical, categorical omega can be computed based on the parameter estimates from a factor analysis model using frequentist estimators such as diagonally weighted least squares. When the sample size is relatively small and thresholds are different across items, using diagonally weighted least squares can yield a substantially biased estimate of categorical omega. In this study, we applied Bayesian estimation methods for computing categorical omega. The simulation study investigated the performance of categorical omega under a variety of conditions through manipulating the scale length, number of response categories, distributions of the categorical variable, heterogeneities of thresholds across items, and prior distributions for model parameters. The Bayes estimator appears to be a promising method for estimating categorical omega. Mplus and SAS codes for computing categorical omega were provided.

Gender Differences in Determinants of Growth From 1st to 8th Grade

Research has shown that overweight tracks from childhood to adulthood with home environmental factors believed to play a role. However, these relationships have not been studied during early childhood or with detailed longitudinal analyses and the method of evaluating growth may influence results. PURPOSE: The purpose of the study was to examine gender differences in the influence of parent obesity, education, and income during childhood (1st grade) on BMI during adolescence (8th grade), using BMI percentile (BMI%) and BMI50, an alternative indicator that accounts for differences in size of growth centiles. METHODS: First grade students were recruited from 37 elementary schools. Body mass was measured to the nearest 0.1kg and height to the nearest 0.25cm by trained nurses using a standardized protocol. Parents completed a questionnaire capturing mother and father’s height and weight, education level and household income. Mother and father BMI were each categorized into one of three levels (1: BMI <25, 2: BMI 25-29.9 and 3: BMI ≥30) and multiplied to create a ParentRisk score. BMI% was calculated using standard SAS codes and BMI50 was calculated as the difference (kg/m2) between measured BMI and the 50th percentile for the child’s gender and age. BMI was also assessed in 8th grade. General linear models were used to examine the influence of predictors on BMI in 8th grade, using BMI% and BMI50. RESULTS: Complete data were available for 422 students (205 male, 217 female). BMI in 1st grade was a strong predictor of BMI in 8th grade in both genders. However, the influence of other factors differed by gender and by assessment method. For males, father’s education approached significance in the BMI% model (F = 2.35, p = 0.056) but was not significant in the BMI50 model (F = 0.52, p = 0.72). For females, ParentRisk (F = 3.88, p < 0.01) and mother’s education (F = 3.15, p = 0.01) were significant predictors of BMI% but family income was not. ParentRisk was also a significant predictor of BMI50 (F = 3.06, p = 0.01) but mother’s education was not (F = 2.07, p = 0.08). CONCLUSIONS: Parent and home environmental factors appear to have differential effects by gender with parent weight playing a stronger role in girls. However, the method of evaluating BMI change may be important. Future research should examine these gender differences over longer time spans.

Impact of The Lunch is in the Bag Intervention on the Dietary Quality of Preschooler's Sack Lunches Measured with the HEI‐2010

ObjectiveTo measure the impact of Lunch is in the Bag (LIITB), a multi‐level behavioral nutrition intervention, on the dietary quality of the lunches that parents packed for their preschool children using Healthy Eating Index (HEI) 2010 total and component scores at 6‐week, 22‐week and 28‐week follow‐up periods.BackgroundThe low dietary quality of U.S. children's diets contribute to the development of childhood obesity and comorbidities. Up to 61% of U.S. children under the age of 6 receive care outside the home. The LIITB efficacy trial conducted in 30 Early Care and Education (ECE) centers in central Texas to evaluate parent‐packing and child‐eating of bag lunches provided a unique opportunity to use the HEI‐2010 to measure the intervention's impact on dietary quality of the packed lunches over time.MethodsParent‐child dyads (n=628) were recruited from 30 Early Care and Education centers in 3 cities of central Texas. Centers were randomized to control or intervention. Two food records were collected on 2 non‐consecutive days per measurement period. Dietary data were collected at baseline, after the 5‐week intervention (6 week follow‐up), after a fall out period, before the 1‐week booster (22‐week follow‐up), and post‐booster (28‐week follow‐up). Dietary records were analyzed using the Food Intake Analysis System and HEI‐2010 scores were computed using standard SAS codes. Time‐by‐treatment interactions in multi‐level modeling were computed for HEI‐2010 total and component scores to measure changes from baseline. Analyses were adjusted for child sex, age, Body Mass Index; and parental race, education and marital status.ResultsThe LIITB intervention produced an increase in the dietary quality of 4.5% (SE=1.1% p&lt;0.0001) at 6‐week follow‐up period, 4.7% (SE=1.1%, P&lt;0.0001) at 22‐week follow‐up period and 2.9% (SE=1.1%, P&lt;0.01) at 28‐week follow‐up compared to baseline. The lunches packed by parents in the intervention condition had Increases in HEI‐2010 Component scores for: Total Vegetables (10.4%, SE=2.9%, P&lt;0.001), Greens and Beans (4.3%, SE=2.2% P&lt;0.05), Total Fruit (8.3%, SE=2.3%, P&lt;0.05), Whole Fruit (8.3%, SE=2.6%, P&lt;0.05), Whole Grains (8.5%, SE=3.4%, P&lt;0.01), Refined Grains (10.1%, SE=3.1%, P&lt;0.01). Increase in HEI‐2010 component scores were sustained through the 22‐week follow‐up period for Whole Grains and Refined Grains. Despite these improvements, the Total HEI score at the 22‐week follow‐up was 52.74 points out of (100 SE=1.4) at intervention and 49.4 points (SE = 1.4) at control ECE centers.ConclusionsMeasuring the dietary quality using the Healthy Eating Index 2010 is a useful specific tool to provide information about the impact of the intervention on food groups and nutrients. More efforts should be done to increase the dietary quality of the meals offered to preschoolers at lunch to promote lifelong healthy eating habits.Support or Funding InformationFunding for this study was made possible by the National Cancer Institute (National Cancer Institute/NIH Grant R01CA149643) and the Michael &amp; Susan Dell Foundation through the Michael and Susan Dell Center for Healthy Living. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Cancer Institute, the National Institutes of Health, or the Michael and Susan Dell Foundation. The Clinical Trials Number is NCT001292434

ESTIMAÇÃO DOS PARÂMETROS DA DISTRIBUIÇÃO BETA-BINOMIAL: UMA APLICAÇÃO USANDO O SOFTWARE SAS

In this paper we describe the parameter estimation of the beta-binomial distribution using the procedure NLMIXED of the SAS software. The beta-binomial distribution is a discrete mixture distribution which can capture overdispersion in the data. The estimation of parameters of the beta-binomial distribution can lead to computational problems, since it does not belong to the exponential family and there are not explicit solutions for the maximum likelihood estimation. Using a real dataset, we show that the SAS software can be satisfactorily used for the estimation of the parameters. We also consider the possibility of including a covariate in the model. The SAS codes used in this paper are given in an Appendix.

Relative index of inequality and slope index of inequality: a structured regression framework for estimation.

The relative index of inequality and the slope index of inequality are the two major indices used in epidemiologic studies for the measurement of socioeconomic inequalities in health. Yet the current definitions of these indices are not adapted to their main purpose, which is to provide summary measures of the linear association between socioeconomic status and health in a way that enables valid between-population comparisons. The lack of appropriate definitions has dissuaded the application of suitable regression methods for estimating the slope index of inequality. We suggest formally defining the relative and slope indices of inequality as so-called least false parameters, or more precisely, as the parameters that provide the best approximation of the relation between socioeconomic status and the health outcome by log-linear and linear models, respectively. From this standpoint, we establish a structured regression framework for inference on these indices. Guidelines for implementation of the methods, including R and SAS codes, are provided. The new definitions yield appropriate summary measures of the linear association across the entire socioeconomic scale, suitable for comparative studies in epidemiology. Our regression-based approach for estimation of the slope index of inequality contributes to an advancement of the current methodology, which mainly consists of a heuristic formula relying on restrictive assumptions. A study of the educational inequalities in all-cause and cause-specific mortality in France is used for illustration. The proposed definitions and methods should guide the use and estimation of these indices in future studies.

A multilevel approach to network meta-analysis within a frequentist framework

Meta-analysis is a powerful tool to summarize knowledge. Pairwise or network meta-analysis may be carried out with multivariate models that account for the dependence between treatment estimates and quantify the correlation across studies. From a different perspective, meta-analysis may be viewed as a special case of multilevel analysis having a hierarchical data structure. Hence, we introduce an alternative frequentist approach, called multilevel network meta-analysis, which also allows to account for publication bias and the presence of inconsistency. We propose our approach for a three-level data structure set-up: arms within studies at the first level, studies within study designs at the second level and design configuration at the third level. This strategy differs from the traditional frequentist modeling because it works directly on an arm-based data structure. An advantage of using multilevel analysis is its flexibility, since it naturally allows to add further levels to the model and to accommodate for multiple outcome variables. Moreover, multilevel modeling may be carried out with widely available statistical programs. Finally, we compare the results from our approach with those from a Bayesian network meta-analysis on a binary endpoint which examines the effect on mortality of some anesthetics at the longest follow-up available. In addition, we compare results from the Bayesian and multilevel network meta-analysis approaches on a publicly available "Thrombolytic drugs" database. We also provide the reader with a blueprint of SAS codes for fitting the proposed models, although our approach does not rely on any specific software.

Evaluation and Interpretation of Interactions

Understanding the factors that define a given interaction is important in agricultural, agronomic, and plant breeding research, where agronomic treatments or genotypes are evaluated under several environmental conditions and where interactions usually complicate a researcher's decisions. We give examples of how interactions, in common agricultural experiments, can be examined and studied to make use of the rich information available on the interaction term of the model. Examples with different levels of interaction complexity are used to illustrate how to analyze and interpret interactions and how interaction components can be partitioned into comparisons with sensible biological interpretations. It will offer researchers a greater understanding of how to exploit interaction information beyond the standard statistical tests performed in the usual analysis of variance. Simple SAS codes for performing standard interaction contrasts and defining interaction covariables are provided.

A new inverse regression model applied to radiation biodosimetry.

Biological dosimetry based on chromosome aberration scoring in peripheral blood lymphocytes enables timely assessment of the ionizing radiation dose absorbed by an individual. Here, new Bayesian-type count data inverse regression methods are introduced for situations where responses are Poisson or two-parameter compound Poisson distributed. Our Poisson models are calculated in a closed form, by means of Hermite and negative binomial (NB) distributions. For compound Poisson responses, complete and simplified models are provided. The simplified models are also expressible in a closed form and involve the use of compound Hermite and compound NB distributions. Three examples of applications are given that demonstrate the usefulness of these methodologies in cytogenetic radiation biodosimetry and in radiotherapy. We provide R and SAS codes which reproduce these examples.

Simultaneous Optimization of Incomplete Multi-Response Experiments

This article attempts to develop a simultaneous optimization procedure of several response variables from incomplete multi-response experiments. In incomplete multi-response experiments all the responses (p) are not recorded from all the experimental units (n). Two situations of multi-response experiments considered are (i) on units all the responses are recorded while on units a subset of responses is recorded and (ii) on units all the responses (p) are recorded, on units a subset of responses is recorded and on units the remaining subset of responses is recorded. The procedure of estimation of parameters from linear multi-response models for incomplete multi-response experiments has been developed for both the situations. It has been shown that the parameter estimates are consistent and asymptotically unbiased. Using these parameter estimates, simultaneous optimization of incomplete multi-response experiments is attempted following the generalized distance criterion [1]. For the implementation of these procedures, SAS codes have been developed for both complete (k ≤ 5, p = 5) and incomplete (k ≤ 5, p1 = 2, 3 and p2 = 2, 3, where k is the number of factors) multi-response experiments. The procedure developed is illustrated with the help of a real data set.

The Sample Size Estimation and its SAS Code for Binary Response Endpoints Clinical Study- a Review

The aim of this article is to provide basic idea about sample size estimation in clinical study of binary response endpoints. Every clinical trial must be planned and it describes the objective, primary and secondary endpoint, method of collecting data, inclusion & exclusion criteria, sample size with scientific justification, statistical methodology. The calculation of sample size with sufficient levels of significance and power is an essential part of protocol which needs to be submitted to drug authorities for approval. In this review emphasis has been made to describe the different kind of design i.e. equivalence, non-inferiority, superiority and equality and their SAS codes for sample size calculation in clinical study.

On comparison of SAS codes with GLM and MIXED for the crossover studies with QT interval data

The structural complexity of crossover studies for bioequivalence test confuses analysts and leaves them a hard choice among various programs. Our study reviews PROC GLM and PROC MIXED in SAS and compares widely used SAS codes for crossover studies. PROC MIXED based on REML is more recommended since it provides best linear unbiased estimator of the random between-subject effects and its variance. Our study also considers the covariance structure within subject over period which most PK/PD studies and crossover studies ignore. The QT interval data after the administra tion of moxifloxacin for a fixed time point are analyzed for the comparison of representative SAS codes for crossover studies.

Closing in on the Truth About Recombinant Human Bone Morphogenetic Protein-2: Evidence Synthesis, Data Sharing, Peer Review, and Reproducible Research

This issue includes 2 systematic reviews that use the same data to address the same question: Compared with iliac crest bone grafting, does rhBMP-2 safely improve outcomes of spinal fusion surgery?...

A Comparison of the Polytomous Logistic Regression and Joint Cox Proportional Hazards Models for Evaluating Multiple Disease Subtypes in Prospective Cohort Studies

Polytomous logistic regression models are commonly used in case-control studies of cancer to directly compare the risks associated with an exposure variable across multiple cancer subtypes. However, the validity, accuracy, and efficiency of this approach for prospective cohort studies have not been formally evaluated. We investigated the performance of the polytomous logistic regression model and compared it with an alternative approach based on a joint Cox proportional hazards model using simulation studies. We then applied both methods to a prospective cohort study to assess whether the association of breast cancer with body size differs according to estrogen and progesterone receptor-defined subtypes. Our simulations showed that the polytomous logistic regression model but not the joint Cox regression model yielded biased results in comparing exposure and disease subtype associations when the baseline hazards for different disease subtypes are nonproportional. For this reason, an analysis of a real data set was based on the joint Cox proportional hazards model and showed that body size has a significantly greater association with estrogen- and progesterone-positive breast cancer than with other subtypes. Because of the limitations of the polytomous logistic regression model for the comparison of exposure-disease associations across disease subtypes, the joint Cox proportional hazards model is recommended over the polytomous logistic regression model in prospective cohort studies. The article will promote the use of the joint Cox model in a prospective cohort study. Examples of SAS and S-plus programming codes are provided to facilitate use by nonstatisticians.

Estimation of the Relative Excess Risk Due to Interaction and Associated Confidence Bounds

The relative excess risk due to interaction (RERI) provides a useful metric of departure from additivity of effects on a relative risk scale. In this paper, the authors show that RERI is identical to the product term in a linear odds ratio or a linear relative risk model. SAS and STATA codes are provided for fitting a linear odds ratio model that directly parameterizes RERI. In addition, this paper presents a method for obtaining likelihood-based 95% confidence bound estimates for RERI. The authors show that likelihood-based confidence intervals may differ substantially from the asymptotic confidence interval estimates advocated by previous authors. The approach presented in this paper should facilitate estimation of RERI and associated likelihood-based confidence bounds, by using standard statistical packages.