Statistical Analysis Framework Research Articles

Business cycle measurement in the presence of structural change: international evidence

By generalizing Hamilton’s model of the US business cycle to a three-regime Markov-switching vector autoregressive model, this paper analyzes regime shifts in the stochastic process of economic growth in the US, Japan and Europe over the last four decades. Empirical evidence is established for the presence of a structural break in the expansionary GDP growth for the US and Japan based on an output-employment MS vector equilibrium correction model, and a structural break in the context of a common European business cycle. For the United States the long expansions of recent years signify basic changes in the business cycles pattern. In the case of Japan we identify long episodes of rapid economic expansions (existing until the mid 1970s) and long economic recessions (as in the 1990s). In Europe we find after an episode of catching-up in the 1970s, convergence in the business cycle pattern which suggests the notion of a European business cycle. The multi-regime Markov-switching VARs proposed are profoundly checked for their economic content and statistical congruency, and are found to provide a sound statistical framework for a comprehensive analysis of the business cycle.

Impact sound transmission through a floating floor on a concrete slab

In this paper a theoretical model to predict bending wave transmission through parallel plates connected by a resilient line is presented. The model is based on the interaction of semi-infinite plates intersecting along an infinite boundary. A full model is developed together with some approximations for the more common cases. The results of the model are then used in a statistical energy analysis (SEA) framework to predict transmission through a chipboard floating floor attached to battens which in turn are supported on a concrete floor. Acoustic transmission through the cavities is also considered. The results show that transmission through the battens is the most important path when there is no resilient layer but that this path is insignificant if any resilient layer is present. A comparison between measured and predicted results gave generally good agreement except at high frequencies when a resilient layer was present.

OPT scheduling performances: A case study

A deep insight into Optimized Production Technology (OPT) is sought for by analysing a case study regarding a leading producer of high quality acetate yarns for the textile market. The case study highlights the importance of setting proper values for OPT management parameters in order to achieve desired performances in terms of both system efficiency as well as system effectiveness. A suitable experimental design and a statistical analysis framework allowed the recognition of the impact of single OPT management parameters. Furthermore, the optimal set-points of these parameters were determined in order to cope with resource constraints in a critical manufacturing department. Scheduling performances achievable as output of 'computeraided' scheduling with OPT were confronted with traditional 'man-made' scheduling on the basis of due date performances and setup-related parameters. The results achieved allowed to avoid problems due to a scheduling procedure that underutilized OPT scheduling characteristics.

A statistical multiscale framework for Poisson inverse problems

This paper describes a statistical multiscale modeling and analysis framework for linear inverse problems involving Poisson data. The framework itself is founded upon a multiscale analysis associated with recursive partitioning of the underlying intensity, a corresponding multiscale factorization of the likelihood (induced by this analysis), and a choice of prior probability distribution made to match this factorization by modeling the "splits" in the underlying partition. The class of priors used here has the interesting feature that the "noninformative" member yields the traditional maximum-likelihood solution; other choices are made to reflect prior belief as to the smoothness of the unknown intensity. Adopting the expectation-maximization (EM) algorithm for use in computing the maximum a posteriori (MAP) estimate corresponding to our model, we find that our model permits remarkably simple, closed-form expressions for the EM update equations. The behavior of our EM algorithm is examined, and it is shown that convergence to the global MAP estimate can be guaranteed. Applications in emission computed tomography and astronomical energy spectral analysis demonstrate the potential of the new approach.

Ambient bioassays for assessing water-quality conditions in receiving streams

The purpose of this paper is to show how short-term laboratory tests, conducted according to US Environmental Protection Agency (EPA) procedures, can be used effectively to assess water-quality conditions in streams or rivers that receive pollutants from industry or municipal or agricultural areas. Standardized, short-term tests with fish, aquatic invertebrates or algae are commonly used to estimate the acute or chronic toxicity of wastewaters; this is referred to as effuent testing. The methods used for testing effluents also can be used to assess water quality in receiving streams: in this application, the procedures are referred to as ambient testing. Despite similarity in methods, the major objective of effluent testing differs importantly from that of ambient testing. In effluent testing, the key objective is to determine how toxic an effluent is; in ambient testing, the main objective usually is that of determining if the water at a site is toxic. This difference is subtle but very important: it shapes the strategy for cost-effective ambient testing, and determines the framework for effective statistical analysis and interpretation of ambient toxicity test results. Specific case-study examples are provided demonstrating the kinds of information that can be extracted from ambient toxicity testing by use of different statistical methods, including analysis of variance, contingency-table analysis, and two types of multivariate procedures (principal components analysis and logistic regression). Examples also are given supporting the idea that an effective ambient testing programme should be long-term, and contain a diagnostic-testing component analogous to the toxicity identification procedures used to supplement effluent-testing programmes. Recommendations derived as 'lessons learned' from largescale ambient toxicity testing programmes for receiving streams at Department of Energy facilities include: (1) testing more frequently with one species (preferably Ceriodaphnia) generally is more effective, in terms of information gained per dollar spent, than testing less frequently with two or more species; (2) use five or more sites per test period, plus two or more reference sites, whenever possible; (3) use four to six test periods per year; and (4) use diagnostic testing to supplement the ambient-testing programme. Various laboratory and in situ methods for environmental assessment are now under development, but these methods probably will not gain acceptance for use in regulatory situations for many years. Rapid growth in need for ecological risk assessments outstrips the rate at which new test procedures are approved for regulatory purposes. Thus, laboratory tests for estimating possible environmental impacts of toxic or disruptive pollutants are likely to be used more frequently during the next decade.

A FRAMEWORK FOR THE ANALYSIS OF HARMONISED MONITORING SCHEME DATA FOR ENGLAND AND WALES

The Harmonised Monitoring Scheme (HMS) records are the best long-term data sets for river water quality in Britain. A simple framework for statistical analysis of these data is developed which copes with the many observations below the limits of detection of the laboratory analytical methods used to measure concentrations of chemical determinands. Besides the estimates of annual mean concentrations, two products of the framework are estimates of precision of these means and a test of long-term change in annual mean concentration. The methodology is applied to concentrations of ammoniacal nitrogen, chloride, nitrite, nitrate, orthophosphate, copper and zinc at HMS river sites in England and Wales. These sites show a clear net improvement in ammoniacal nitrogen, nitrite, copper and zinc concentrations when comparing the late-1970s with the early-1990s. Conversely, between the same two periods, these sites show a clear net deterioration in chloride and orthophosphate concentrations and a small net deterioration in nitrate concentration at the HMS river sites. The simple framework is extended to include a seasonal component and a full likelihood test of long-term change. This is applied to ammoniacal nitrogen concentration at 30 HMS sites. The proportion of sites showing long-term improvement in ammoniacal nitrogen is a little higher than that obtained in the simple framework, suggesting that the simple approach is conservative in detecting change.

Pulsar timing limits on very low frequency stochastic gravitational radiation.

A low-frequency, stochastic gravitational radiation background can be detected through the irregularities it induces in pulsar arrival times. Here, we present an optimal statistical framework for analysis of timing data from a single pulsar, correcting an error in previous treatments. Observations of PSR B1855+09 yield an upper limit (95% confidence) of 1.0\ifmmode\times\else\texttimes\fi{}${10}^{\ensuremath{-}8}$ or (90% confidence) 4.8\ifmmode\times\else\texttimes\fi{}${10}^{\ensuremath{-}9}$ of the closure density at frequency 4.4\ifmmode\times\else\texttimes\fi{}${10}^{\ensuremath{-}9}$ Hz. This result probably rules out cosmological models that use cosmic strings as seeds for galaxy formation. Combined observations of four binary pulsars yield weaker limits at frequencies as low as ${10}^{\ensuremath{-}12}$ Hz.

A bayesian approach to wiggle-matching

“Wiggle-matching” is a technique developed to provide calendar dates for a series of samples of known deposition rate using the now well-established high-precision radiocarbon calibration curve. In this paper we describe how the technique may be formulated within the Bayesian framework for statistical analysis and thereby permit the incorporation of uncertainties regarding the deposition rate. The methodology is illustrated by two archaeological examples.

SEA with exterior fluid loading: An open channeling

We describe the vibration of a submerged elastic solid within a statistical energy analysis (SEA) framework. We discuss the modal structure of a submerged elastic solid, and contrast it to that assumed in traditional SEA analyses. In particular, submerged solids have higher modal density than their invacuo counterparts, much higher levels of (frequency dependent) damping, and possibly strong coupling between the fluid and solid. Further, modes of submerged structures do not form a complete set, and the energy of oscillation in a fluid/solid mode is often irretrievable, independent of the coupling in the system. These results pinpoint the shortcomings of SEA in addressing exterior fluid loading. We close by identifying research directions for the advancement of SEA toward incorporating exterior fluid loading. [Work supported by ONR.]

Sound transmission between columns and floors in framed buildings

In framed buildings joints between columns and floor structures commonly occur. Beams at the top of columns, which support the floors, increases the complexity of the joint and can greatly influence the transmission characteristics. Impedances for composite beam/plate structures are used together with wave theories to obtain predictions for vibration transmission at such joints. Predictions of sound transmission at such joints are computed within a statistical energy analysis framework and good agreement is found between measured and predicted results.

Statistical Models for Behavioral Observations

Statistical models for the Behavior Stream (renewal processes) are used to explore properties of behavioral observations. For continuous observation, statistical and psychometric properties are obtained for four types of behavioral measures: (i) empirical rates of behavior; (ii) empirical proportions, or relative frequencies, of a type of behavior; (iii) empirical prevalence (proportion of time the behavior occurs); and (iv) empirical event duration. Also, time-sampling alternatives to continuous observation are evaluated. Our formulation includes representations for three sources of unreliability: (a) finite observation time, (b) recorder errors, and (c) heterogeneity (instability) over occasions of observation. Traditional psychometric methods carried over from the analysis of responses to test items (including generalizability theory) are shown not to be applicable to behavioral observations. Our results provide a guide for design and a framework for statistical analysis in behavioral observation research.

A general framework for the statistical analysis of sequential dyadic interaction data.

A general framework for the statistical analysis of sequential dyadic interaction data.

The Radiology of Nontraumatic Coma

This article describes a logical approach to the neuroimaging evaluation of patients with coma. The decision tree analysis provides an efficient method and attempts to convey the logical sequential concepts of analysis of CCT findings in patients presenting in stupor or coma. The sequential approach tabulated serves as a good mnemonic for clinicians faced daily with decisions to be made in the evaluation and management of comatose patients. The approach also provides a framework for subsequent statistical (Bayesian or other) analysis of efficacy, cost effectiveness, and impact on clinical practice and health care delivery.

The use of time series processes to model the error structure of earnings in a longitudinal data analysis

This study presents an error structure that offers a rich statistical framework for panel data analysis. It includes as special cases most of the error specifications found in longitudinal studies of wages and earnings. A general set of procedures for choosing a specification of this error structure and estimating its parameters appears in the first part of this study. The last section applies these procedures to fit an error structure for wages and earnings of prime-age males using data from the Michigan Panel of Income Dynamics.

Thermodynamics of Small Systems

It is shown in this paper that the differential equations of macroscopic thermodynamics can be generalized in such a way that they apply as well to small (i.e., nonmacroscopic) systems. Conventional thermodynamic relations then follow from the present treatment as a limiting case (large system). As with macroscopic thermodynamics, there are two main classes of applications: (1) as an aid in analyzing, classifying, and correlating equilibrium experimental data on ``small systems'' such as (non-interacting) colloid particles, liquid droplets, crystallites, macromolecules, polymers, polyelectrolytes, nucleic acids, proteins, etc.; and (2) to verify, stimulate, and provide a framework for statistical mechanical analysis of models of finite (i.e., ``small'') systems. A well-known experimental and theoretical example (in which there are sizable effects of chain length) is the helix-coil transition in synthetic polypeptides and polynucleotides. Unlike macroscopic thermodynamics, thermodynamic functions are different for different environments (open, closed, isothermal, siobaric, etc.). Although it is possible to derive a single set of thermodynamic equations applicable to all environments, it proves useful to give a separate analysis for each environment. Several cases are discussed, and a few simple statistical mechanical models are used for purposes of illustration. The partition function for a ``completely open'' small system can be used without any special technique such as is required when this partition function is applied to a macroscopic system. Solvent effects are discussed and details are given in one case. The present method provides an invariant treatment of the spherical interface of a drop or bubble, independent of any choice of dividing surface. Usually, only mean values of fluctuating extensive variables appear in thermodynamic equations. This is justified in macroscopic thermodynamics because fluctuations are generally unimportant. The situation is different for small systems and we derive, in one case, a hierarchy of thermodynamic equations involving higher moments of the probability distribution of fluctuating extensive properties.

Cortisone in ulcerative colitis; final report on a therapeutic trial.

We describe a general statistical framework for the genetic analysis of quantitative trait data in inbred line crosses. Our main result is based on the observation that, by conditioning on the unobserved QTL genotypes, the problem can be split into two statistically independent and manageable parts. The first part involves only the relationship between the QTL and the phenotype. The second part involves only the location of the QTL in the genome. We developed a simple Monte Carlo algorithm to implement Bayesian QTL analysis. This algorithm simulates multiple versions of complete genotype information on a genomewide grid of locations using information in the marker genotype data. Weights are assigned to the simulated genotypes to capture information in the phenotype data. The weighted complete genotypes are used to approximate quantities needed for statistical inference of QTL locations and effect sizes. One advantage of this approach is that only the weights are recomputed as the analyst considers different candidate models. This device allows the analyst to focus on modeling and model comparisons. The proposed framework can accommodate multiple interacting QTL, nonnormal and multivariate phenotypes, covariates, missing genotype data, and genotyping errors in any type of inbred line cross. A software tool implementing this procedure is available. We demonstrate our approach to QTL analysis using data from a mouse backcross population that is segregating multiple interacting QTL associated with salt-induced hypertension.