Statistical Practices of IJPT Researchers: A Review from 1993-2000.

Research articles published in the International Journal of Play Therapy (IJPT) were examined to investigate the use of selected statistical practices in quantitative inferential studies. The current article examined whether researchers (a) verified methodological assumptions of statistical analyses, (b) reported confidence intervals, (c) discussed the risk of experiment-wise Type I error, (d) preferred univariate analyses to multivariate analyses, (e) used univariate analyses as post hoc methods to detect multivariate effects, and (e) screened data and reported findings with graphical displays. Recommendations for improved statistical practice are provided. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Multivariate analysis of variance by multiple regression methods.

Discusses multivariate analysis of variance as a general case of familiar multiple regression analysis. A consequence of this approach is a unified treatment of multivariate analysis of variance which can be used by psychologists who are generally familiar with multiple regression approaches to univariate analysis of variance. It is suggested that the generality of the approach permits solutions consistent with any of the several available strategies for dealing with problems of unequal and disproportionate cell frequencies. Inherent in the multiple regression formulation is the otherwise not so obvious fact that univariate analysis of variance results are an integral part of the multivariate solution and that both are important for understanding complex data. Methods of interpreting multivariate analysis of variance results in complex factorial experimental designs are discussed. (32 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Studies on the stereoselective internal acyl migration of ketoprofen glucuronides using 13C labeling and nuclear magnetic resonance spectroscopy

BACKGROUND: To improve asthma disease management, the National Asthma Education Program (NAEP) Expert Panel published Guidelines for the Diagnosis and Management of Asthma in 1991. OBJECTIVES: To compare the current status of asthma disease management among patients in a large health maintenance organization with the NAEP guidelines and to identify the factors that may be associated with medical care (eg, emergency department visits and hospital admissions) and adherence to the guidelines. METHODS: Analyses of 1996 survey data from 5580 members with asthma (age range, 14 to 65 years) covered by a major health maintenance organization in California (Health Net). RESULTS: In general, adherence to NAEP guidelines was poor. Seventy-two percent of respondents with severe asthma reported having a steroid inhaler, and of those, only 54% used it daily. Only 26% of respondents reported having a peak flowmeter, and of those, only 16% used it daily. Age (older), duration of asthma (longer), increasing current severity of disease, and treatment by an asthma specialist correlated with daily use of inhaled steroids. Ethnicity (African American and Hispanic) correlated negatively with inhaled steroid use but positively with emergency department visits and hospital admissions for asthma. Increasing age and treatment by an asthma specialist were also identified as common factors significantly related to the daily use of a peak flowmeter and, interestingly, to overuse of beta2-agonist metered-dose inhalers. CONCLUSIONS: Although the NAEP guidelines were published 7 years ago, compliance with the guidelines was low. It was especially poor for use of preventive medication and routine peak-flow measurement. Furthermore, the results showed that asthma specialists provided more thorough care than did primary care physicians in treating patients with asthma. Combining the results of the regression analyses revealed that some of the variation in rates of emergency department visits and hospitalizations among some subpopulations can be explained by the underuse of preventive medication. This study serves the goal of documenting the quality of care and services currently provided to patients with asthma through a large health maintenance organization and provides baseline information that can be used to design and assess effective population-based asthma disease management intervention programs.     

A specification test for univariate and multivariate proportional hazards models

This paper applies White's (1982, Econometrica 50, 1-25) information matrix (IM) test for correct model specification to proportional hazards models of univariate and multivariate censored survival data. Several alternative estimators of the test statistic are presented and their size performance examined. White also suggested an estimator of the parameter covariance matrix that was robust to certain forms of model misspecification. This has been subsequently proposed by others (e.g., Royall, 1986, International Statistical Review 54, 221-226) and applied by Huster, Brookmeyer, and Self (1989, Biometrics 45, 145-156) as part of an independence working model (IWM) approach to multivariate censored survival data. We illustrate how the IM test can be used for both univariate data and as part of the IWM approach to multivariate data.     

Dyadic social interactions.

Kraemer and Jacklin (1979) proposed a method of analysis of univariate dyadic social interactions or relational data, and Mendoza and Graziano (1982) extended this method to multivariate relations. Their approach is based on an analysis-of-variance-type model that contains parameters characterizing the behavior of actors and partners and their interactions on each relation. The techniques presented in this article offer an alternative approach to the multivariate analysis of social interactions by realizing that many relations yield discrete-valued data and thus are better modeled by using methods designed for categorical data. This alternative approach is also more general because it allows more types of models to be fit. We illustrate, using the same data analyzed by the earlier methods. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

The statistical analysis of dyadic social behavior: A multivariate approach.

Extends the logic of the univariate model for studying dyadic social interaction and presents a multivariate model. The more general multivariate model is illustrated using hypothetical data and real data from a study by M. Jacklin et al (see record 1979-25785-001) on girl–girl, boy–boy, and girl–boy interactions among 45 preschoolers. The 8 variables used included passive behavior, frequency of aggression, and frequency of social withdrawal. Some advantages of this model include detection of Type I error rates and generality. (10 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

高炉铁水温度的多元时间序列建模和预测

针对高炉炉温铁水硅含量为预测对象的不确定性和高炉炉温单变量时间序列模型所含炉温输入信息量少、难以揭示各个变量之间的相互关系及变化规律的特点,以高炉铁水温度为研究对象,建立BP神经网络多元时间序列模型和T-S模糊神经网络多元时间序列模型。应用高炉实际数据做模型检验,结果表明,T-S模糊神经网络多元时间序列模型取得更好的命中率和预测精度。     

Univariate versus multivariate tests in repeated-measures experiments.

The usual univariate test for the repeated-measures effect in a 1-way design rests upon an assumption of uniform variances and covariances. The standard test of this assumption is shown to have acceptable power only when the multivariate test of the hypothesis is essentially as powerful as the univariate test. A modified form of the univariate test, not requiring the assumption of uniformity, is compared to the multivariate test with respect to power. Depending on the variance-covariance structure of the data and the alternative hypothesis, the univariate test ranges from somewhat better to much worse than the multivariate test. There are possibly interesting experimental effects which the univariate test is virtually powerless to detect. (16 ref.) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

An interactive graphics system for real-time investigation and multivariate data portrayal for complex pedigree data systems

GRIFFIN (graphics investigation of familial information), an interactive graphics system for exploratory investigation of data on individuals associated by familial relationships, was designed to provide genetic epidemiologists a flexible, rapidly responsive tool for viewing and guiding exploration of complex databases in the context of familiar pedigree structures. It graphically portrays both categorical and multivariate scalar data on individuals in those structures. The display can be inverted to show all ancestors and descendants of any individual the user designates. It provides cues to censored information when bushy pedigrees cannot be fully displayed without sacrificing legibility. These guide users on where to next invert the system. Investigators may translate/zoom the display, vary the mode of representing data, point to individuals to obtain displays of alphameric information about them, etc. Developed in Fortran using IBM's GDDM graphics subroutines for an IBM 3090 mainframe, GRIFFIN's design anticipates porting to smaller systems.     

"Prediction" of the one-minute Apgar score from fetal heart rate data

The value of any fetal monitoring technic is in its ability to predict infant outcome. In the present study, the ability of fetal heart rate (FHR) monitoring data to "predict" a measure of short-term infant outcome, the 1-minute Apgar score, was evaluated using univariate and multivariate statistical analyses. Of 61 monitored high-risk infants, 46 had high (7 to 10) and 15 had low (1 to 6) 1-minute Apgar scores. Computer analysis of FHR/intrauterine pressure (IUP) data for these 61 infants revealed that the infants with low Apgar scores had more than the expected number of late decelerations (LD). Using a threshold of ten LD and univariate analysis, 74% of the infants could be properly classified for high or low Apgar scores, but 60% of the infants with low Apgar scores were not identified. Using discriminant function (multivariate) analysis for the numbers of LD and uterine contractions, 47% of the depressed infants were appropriately identified and simple risk scoring equations were devised. Using additional observation vectors, including the number of accelerations and early decelerations, 67% of the depressed infants could be identified. The results of this study suggest that using multiple observation vectors improves the predictive capacity and, thus, the value of fetal monitoring data. Clinical experience suggests that the value of monitoring data can be further enhanced by simultaneous evaluation of other observation vectors from additional perinatal data sets using the technics of this study.     

Myelodysplastic syndromes (MDS): prognostic factors and scoring systems

OBJECTIVE: To evaluate the score systems of Cassano and Sanz and suggest a new one. DESIGN: Case series. LOCATION: Teaching hospitals: EPM UNIFESP and Faculdade de Medicina de Botucatu. PARTICIPANTS: 59 patients diagnosed from 1979 to 1992. INTERVENTION: Evaluation of clinical-laboratorial data. MEASUREMENT: Statistical comparison, uni and multivariate analysis and actuarial survival curves. RESULTS: Cassano's system divided the patients into high and low risk (p = 0.0966) while Sanz's gave high, intermediate and low risk (p = 0.0108). The univariate analysis showed hemoglobin, WBC count, E/M ratio, liver size and blast percentage in BM as statistically significant. The multivariate analysis showed blast percentage in BM (p = 0.004) and Hb (p = 0.050) as significant. Our system, considering the multivariate analysis data, divided the patients into high, intermediate and low risk (p = 0.0038). CONCLUSIONS: Sanz's system was more functional than Cassano's, while ours showed predictive survival value and ease of use in clinical practice.     

The influence of load on the lifetime distribution

This paper derives lifetime distributions for components that fail due to overloading. We introduce the notion of energy similarity. This makes it possible to use knowledge about how components behave under different loads in order to establish a probability model. To obtain parametric families, we consider two physically distinct cases: (1) components that function individually and (2) components that function as part of a system. The former leads to univariate and the latter to multivariate models. The results are particularly useful when little or no data is available.     

Configurational selectivities: Critical for development of visual perception and attention.

Argues that the most distinctive and highly developed characteristics of human visual perception lie in the ability to discriminate and extract information from configurationally complex displays. This ability is manifested early in life. Previous data on neonates and a new experiment with somewhat older infants are described that used the preferential looking technique. These experiments demonstrated a clear progression from preference for simple patterns and highly contrasting elements to more subtle variations in contrast and configuration in the 1st 5 mo of life. (French abstract) (24 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

On choosing a test statistic in multivariate analysis of variance.

Users of multivariate analysis of variance must choose which of several available test statistics to employ as a generalization of the usual univariate F statistic. A review of statistical literature concerning the power and robustness of the 4 most promising tests leads to the recommendation of K. C. Pillai's (1955) and M. S. Bartlett's (1939) trace statistic for general use. A survey of recent experimental reports revealed that psychologists have been using a 2nd best statistic and that they have frequently failed to specify their statistic to let readers judge its appropriateness. To facilitate increased use of the Pillai-Bartlett statistic, information is given concerning computation, the availability of significance tables, and a convenient F approximation. (45 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Graphs as aids to knowledge construction: Signaling techniques for guiding the process of graph comprehension.

Graphical displays are frequently used to express quantitative information in texts, but viewers are sometimes unable to comprehend and learn the relevant information. According to a cognitive analysis, graph interpretation involves (a) relatively simple pattern perception and association processes in which viewers can associate graphic patterns to quantitative referents and (b) more complex and error-prone inferential processes in which viewers must mentally transform data. Experiment 1 establishes that graphs can be redesigned to improve viewers' interpretations by minimizing the inferential processes and maximizing the pattern association processes required to interpret relevant information. In Experiments 2 and 3, the researchers isolated one important factor that affects viewers' interpretation (i.e., the perceptual organization of the information in graphs). If relevant quantitative information is perceptually grouped to form visual chunks (because relevant data points are either connected in line graphs or close together in bar graphs), then viewers describe relevant trends. If relevant information is not perceptually grouped, viewers are less likely to comprehend relevant trends. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Bayesian Updating of Parameters for a Sediment Entrainment Model via Markov Chain Monte Carlo

A Bayesian framework incorporating Markov chain Monte Carlo (MCMC) for updating the parameters of a sediment entrainment model is presented. Three subjects were pursued in this study. First, sensitivity analyses were performed via univariate MCMC. The results reveal that the posteriors resulting from two- and three-chain MCMC were not significantly different; two-chain MCMC converged faster than three chains. The proposal scale factor significantly affects the rate of convergence, but not the posteriors. The sampler outputs resulting from informed priors converged faster than those resulting from uninformed priors. The correlation coefficient of the Gram–Charlier (GC) probability density function (PDF) is a physical constraint imposed on MCMC in which a higher correlation would slow the rate of convergence. The results also indicate that the parameter uncertainty is reduced with increasing number of input data. Second, multivariate MCMC were carried out to simultaneously update the velocity coefficient C and the statistical moments of the GC PDF. For fully rough flows, the distribution of C was significantly modified via multivariate MCMC. However, for transitional regimes the posterior values of C resulting from univariate and multivariate MCMC were not significantly different. For both rough and transitional regimes, the differences between the prior and posterior distributions of the statistical moments were limited. Third, the practical effect of updated parameters on the prediction of entrainment probabilities was demonstrated. With all the parameters updated, the sediment entrainment model was able to compute more accurately and realistically the entrainment probabilities. The present work offers an alternative approach to estimating the hydraulic parameters not easily observed.     

The effect of occlusion on motion integration in infants.

Previous psychophysical studies have shown that the adult human visual system makes use of form information such as occlusion to determine whether to integrate or segregate local motion signals (J. McDermott, Y. Weiss, & E. H. Adelson, 2001). Using the displays developed by McDermott et al., these experiments examined whether occlusion and amodal completion affect motion integration in infants. After familiarizing infants with the displays, infants were tested for preference between coherent motion and local motion displays. The results indicate that 5- to 8-month-olds, but not 3-month-olds, showed a significantly greater preference for the local motion display under occlusion conditions. These results suggest that 5- to 8-month-olds perceive motion to be coherent under occlusion conditions. The results are compatible with previous data showing that amodal completion of static information emerges at around 5-6 months of age (Y. Otsuka et al., 2006a), adding that infants use amodal completion for motion integration at this same time. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Uses of discriminant analysis following MANOVA: Multivariate statistics for multivariate purposes.

When a test of multiple ANOVA is found to be significant, it must be followed by other analyses before a researcher can arrive at an accurate understanding of the data set. Two possibilities for follow-up analyses include univariate ANOVA and discriminant analysis. This article presents the results of a Monte Carlo study (N?=?450) wherein typical, but simple, multivariate data were analyzed by the 2 techniques. Results demonstrate that discriminant analysis is capable of showing the underlying dimensionality of the data as well as determining the contribution of individual variables to the underlying dimensions, whereas ANOVA is limited to specifying the contribution of each variable to group separation. It is argued that when researchers analyze multivariate data, primary goals become interpretation and understanding the data set. It is concluded that discriminant analysis is most suitable for this purpose. (29 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Bullying and Victimization in Elementary Schools: A Comparison of Bullies, Victims, Bully/Victims, and Uninvolved Preadolescents.

Research on bullying and victimization largely rests on univariate analyses and on reports from a single informant. Researchers may thus know too little about the simultaneous effects of various independent and dependent variables, and their research may be biased by shared method variance. The database for this Dutch study was large (N=1,065) and rich enough to allow multivariate analysis and multisource information. In addition, the effect of familial vulnerability for internalizing and externalizing disorders was studied. Gender, aggressiveness, isolation, and dislikability were most strongly related to bullying and victimization. Among the many findings that deviated from or enhanced the univariate knowledge base were that not only victims and bully/victims but bullies as well were disliked and that parenting was unrelated to bullying and victimization once other factors were controlled. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Numerical analysis of western blot digitalized images. The case of HIV

The present work explores the use of image digitalization of western blot (WB) aiming to extract more information about the humoral immune response of human immunodeficiency virus (HIV) infected individuals, and to analyze obtained data in a multivariate manner. The digitalization and analysis of WB images was performed on 115 sera. Images were analyzed either qualitatively: dendogram and principal component analysis (PCA) or quantitatively: PCA of the total bands, taking either the antigens, which belong to the virus, or only those which do not. Results show the feasibility of mechanical diagnosis of a large number of WB images. The dendogram and the qualitative PCA satisfactorily separated white images, images with less than four bands, and images with more complex patterns. Quantitative analysis, which keeps more information, separated the images of negative, undetermined and positive diagnosis quite precisely. It was also found that the positive images with complex patterns of antigen recognition correlate better with asymptomatic individuals. Image analysis also revealed various other bands in WB which do not seem to correspond to viral proteins and could represent autoantigens or crossed antigens between HIV and humans which may cause autoimmunity. Digital analysis of WB images is thus demonstrated to be of great usefulness in the diagnosis and of potential great interest in following the evolution and exploring the pathogenesis of AIDS.