Search

Summary: "Analysis of Variance, Design, and Regression: Linear Modeling for Unbalanced Data, Second Edition presents linear structures for modeling data with an emphasis on how to incorporate specific ideas (hypotheses) about the structure of the data into a linear model for the data. The book carefully analyzes small data sets by using tools that are easily scaled to big data. The tools also apply to small relevant data sets that are extracted from big data. New to the Second Edition: Reorganized to focus on unbalanced data ; Reworked balanced analyses using methods for unbalanced data Introductions to nonparametric and lasso regression ; Introductions to general additive and generalized additive models ; Examination of homologous factors ; Unbalanced split plot analyses ; Extensions to generalized linear models R, Minitab®, and SAS code on the author's website. The text can be used in a variety of courses, including a yearlong graduate course on regression and ANOVA or a data analysis course for upper-division statistics students and graduate students from other fields. It places a strong emphasis on interpreting the range of computer output encountered when dealing with unbalanced data"--Provided by publisher.

Recent trends in the pharmaceutical sector are changing the way protein purification processes are designed and executed, moving from operating the process in a fixed point to allowing a permissible region in the operating space known as design space. This trend is driving product development to design quality into the manufacturing process (Quality by Design) and not to rely exclusively on testing quality in the product. A typical purification step has numerous operating parameters that can impact its performance. Therefore, optimization and robustness analysis in purification processes can be time-consuming since they are mainly grounded on experimental work. A valuable approach consists in the combination of an adequate risk analysis technique for selecting the relevant factors influencing process performance and the design of experiment methodology. The latter allows for many process variables which can be studied at the same time; thus, the number of tests will be reduced in comparison with the conventional approach based on trial and error. These multivariate studies permit a detailed exploration in the experimental range and lay the foundation of Quality by Design principles application. This article outlines a recommended sequence of activities toward the establishment of an expanded design space for a purification process. [ABSTRACT FROM AUTHOR]

A new method for estimation of the mean and variance models of the responses, taking into account the correlation between the multiple responses, together with the heteroscedasticity of the observations due to errors in the factor levels is proposed. The application of this method gives us the possibility to use raw industrial data for mean and variance model estimation and leads to reduction of the predicted variance of the responses in production conditions. Recommendations for the experimental design sequential generation, based on the proposed method are made. The introduced new combined method is applied to an electron beam welding experiment. [ABSTRACT FROM AUTHOR]

One of the main principles in a design of experiment is to use blocking factors whenever it is possible. On the other hand, if the blocking information is not precisely defined and subjective in nature, it is usually discarded. This paper introduces a new design that utilizes available subjective information on experimental units to create artificial blocking factors and develops a rank-sum test to test the difference between two population medians. [ABSTRACT FROM AUTHOR]

In this study, response surface models (RSMs) based on limited data are developed. Experimental cutting force data for the flat end milling process are employed to build these models. Four RSM models are developed in terms of process variables. The first model is used to build the mean cutting force. Similarly, the second, third and fourth models are used to build models for the variance, skewness and curtosis coefficients of the cutting forces. A bi-objective optimization procedure is developed and solved to generate a set of optimal process settings. An approximation scheme resulted in 6 possible objective combinations. The Pareto set of solutions (or part of) are generated for the six possible combinations. The merit of this study lies in the fact that response surfaces are built from limited data, often experienced in reality. [ABSTRACT FROM AUTHOR]

Chapter 3 of the book "The Six Sigma Manual for Small and Medium Businesses: What You Need to Know Explained Simply" is presented. It discusses the use of scientific methods or experiments in investigating and determining the root cause of problems and developing the best course of action. Scientific methods start with hypotheses and the author warns against assumptions that may affect the design of the experiment.

Chapter 2 of the book "The Six Sigma Manual for Small and Medium Businesses: What You Need to Know Explained Simply" is presented. It explains terminologies used in Six Sigma methodologies including Critical to X (CTX), Critical to Quality (CTQ) and Design of Experiments (DOE). Gage Repeatability and Reproducibility (R&R) measures tolerance and process variation while Voice of the Customer refers to a process for gathering feedback from customers for use as input to the design of the best product or service.

The double sampling (DS) control chart can efficaciously reduce sample sizes and increase performance of process monitoring. Hsu [Int. j. prod. res., vol. 42, no. 5, pp. 1043-1047] mentioned that the statistical design model of DS x̅ control chart could merely minimize sample sizes during in-control process monitoring but fail to decrease sample sizes during detection of process shifts. In this study, with minimization of sample sizes for both in-control process and out-of-control process, a multi-objective programming method and genetic algorithm are proposed for statistical designs of DS x̅ control chart. In comparison with both statistical design models, it is quite obvious that our model can effectively lower sample sizes of two process situations. [ABSTRACT FROM AUTHOR]

Base of the Pyramid (BoP) is a 4 billion design target group, which is composed of people living on an income less than US$3 per day. Design for BoP is a new research topic and design practice started from 2003 in the faculty of Industrial Design Engineering, Delft University of Technology (IDE/TU). To explore some unknown research questions about it, a research project on "Product design for BoP" had been set up. This paper is an analysis research based on 24 case studies from IDE/TU, and the aim is to conclude design decision factors for BoP. [ABSTRACT FROM AUTHOR]

The main aim of research is to introduce a new data-driven user simulation approach for the quality and usability evaluation for spoken dialogue systems. [ABSTRACT FROM AUTHOR]

Chapter 61 of the book "HANDBOOK OF DEVELOPMENT ECONOMICS," VOLUME 4, edited by T. Paul Schultz and John Strauss is presented. It highlights the use of randomization as a solution to selection bias and a partial solution to publication biases. Discussions of various ways in introducing randomization in field settings, design issues, analyzing data and drawing general conclusions from randomized evaluations are presented.

Chapter 59 of the book "HANDBOOK OF DEVELOPMENT ECONOMICS," VOLUME 4, edited by T. Paul Schultz and John Strauss is presented. It highlights Martin Ravillion's review of non-experimental methods relying on social experiments. It presents studies used in the methods to evaluate anti-poverty programs including the context dependent. Details on propensity-score methods, exploiting program design, higher-order differences and instrumental variables are also offered.

There is not a method for analysing unreplicated factorial designs that performs well for various configurations of number and size of active effects. Moreover, the most popular tool that has been applied is informal and subjective. To overcome these drawbacks, this paper suggests a multiple testing to help practitioners making objective decisions and identifying active effects with a minimum number of additional runs for a wide range of active effects configurations. The selected methods are efficient and easy to implement by practitioners, including those who do not have a profound knowledge of statistics. Two examples from the literature exemplify and justify the proposed approach. [ABSTRACT FROM AUTHOR]

Although experimentation is a frequent activity of engineers, they usually use primitive strategies to carry on their experiments. A survey identifies the needs of using an efficient and practical technique for the experimentation. Although Six Sigma improvement initiative emphasized the use of Design of Experiments (DoE) for experimentation, engineers still consider it a difficult technique to apply and interpret. For this reason, this paper presents a methodology that tries to make easier the implementation of DoE, as an approach to bridge the existing gap between the technique and industries. [ABSTRACT FROM AUTHOR]

Automatic Process Control (APC) based on Design of Experiment (DOE) is an efficient approach to further improve the manufacturing quality. The process changes both in mean and variance due to online parameter adjustment make it hard to apply traditional SPC chart in DOE-based APC applied process. This paper develops an alternative design of adaptive control charts to give a solution for this problem. Adaptive control chart parameters are adjusted online to better track the process changes and to gain the possible SPC run cost reduction. An example is provided to illustrate the proposed adaptive control chart design approach. [ABSTRACT FROM AUTHOR]

Cyclodextrin glucanotransferase production from Bacillus clausii E16, a new bacteria isolated from Brazilian soil samples was optimized in shake-flask cultures. A 24 full-factorial central composite design was performed to optimize the culture conditions, using a response surface methodology. The combined effect among the soluble starch concentration, the peptone concentration, the yeast extract concentration, and the initial pH value of the culture medium was investigated. The optimum concentrations of the components, determined by a 24 full-factorial central composite design, were 13.4 g/L soluble starch, 4.9 g/L peptone, 5.9 g/L yeast extract, and initial pH 10.1. Under these optimized conditions, the maximum cyclodextrin glucanotransferase activity was. 5.9 U/mL after a 48-h fermentation. This yield was 68% higher than that obtained when the microorganism was cultivated in basal culture medium. [ABSTRACT FROM AUTHOR]

In the present work, the production of rhamnolipid from residual soybean oil (RSO) from food frying facilities was studied using a strain of Pseudomonas aeruginosa of contaminated lagoon, isolated from a hydrocarbon contaminated soil. The optimization of RSO, amonium nitrate, and brewery residual yeast concentrations was accomplished by a central composite experimental design and surface response analysis. The experiments were performed in 500-mL Erlenmeyer flasks containing 50mL of mineral medium, at 170 rpm and 30±1°C, for a 48-h fermentation period. Rhamnolipid production has been monitored by measurements of surface tension, rhamnose concentration, and emulsifying activity. The best-planned results, located on the central point, have corresponded to 22g/L of RSO, 5.625 g/ L of NH4NO3' and 11.5 g/L of brewery yeast. At the maximum point the values for rhamnose and emulsifying index were 2.2g/L and 100%, respectively. [ABSTRACT FROM AUTHOR]

Chapter 27 of the book "Start With a Story: The Case Study Method of Teaching College Science: An NSTA Press Journals Collection" is presented. It presents an interrupted case study, based on a 1994 journal article about the parenting behavior of American coots, which can be used to examine the hypothesis of parental favoritism. It says that the case allows students to develop hypotheses and design experiments, resembling the way that scientists conduct research.

IMPACT ON RECIDIVISM In this book I examined correctional interventions, management strategies, and treatment and rehabilitation programs to identify which were effective in reducing the recidivism of offenders and delinquents. Hereafter, in this chapter, I refer to all of these as programs or interventions. Two hundred eighty-four evaluations of the programs were located and scored for the quality of the research methods. The significance and effectiveness were examined for each study. Table 15.1 shows the correctional programs examined, the chapter for each, the number of evaluations located, the number of the evaluations scored five (successful random assignment), the number scored two (too poorly done to include in decision making), and the conclusion drawn about the effectiveness of the program. What Works As shown in Table 15.1, at this point in time, the research evidence demonstrates the following programs were effective in reducing recidivism: academic education, vocation education, MRT, R&R, cognitive restructuring, cognitive behavior treatment for sex offenders, behavioral treatment for sex offenders, hormonal/surgical treatment of sex offenders, MST for juveniles, drug courts, drug treatment in the community, and incarceration-based drug treatment. All of the effective programs have what I refer to as human service components except the hormonal/surgical treatment for sex offenders. The effective programs are not based on a control or deterrent philosophy. Furthermore, the results are consistent with the results of the theoretical meta-analyses (see Chapter 4). As found in theoretical meta-analyses such as those conducted by Andrews et al. (1990) and Lipsey (1992; 1995), many of the programs target dynamic criminogenic factors. [ABSTRACT FROM AUTHOR]

INTRODUCTION During the 1980s, in response to the record numbers of convicted offenders and widespread prison crowding, correctional officials in the United States expanded the range of sentencing options for convicted offenders (Morris & Tonry, 1990; Cullen, Wright, & Applegate, 1996; Tonry & Lynch, 1996; Byrne, Lurigio, & Petersilia, 1992; Harland, 1996; Smykla & Selke, 1995). Prior to this, there were few sentencing options for punishing criminals. Offenders were either incarcerated or given standard probation. In some jurisdictions, ordinary probation often equated with perfunctory supervision. The problem was that the range of severity of crimes did not fall neatly into these two categories of punishments. For many offenders, prison was often too severe a punishment whereas probation was too lenient. To remedy the situation, many states began to develop alternatives. These alternatives, variously labeled intermediate sanctions, correctional options, alternative punishments, or community corrections, were designed to be punishments somewhere between prison and routine probation with respect to harshness and restrictiveness. Rather than being limited to a choice of either prison or ordinary probation, offenders could be given one of a variety of sanctions that ranged between prison and probation in severity. Numerous different types of intermediate sanctions were developed including house arrest, electronic monitoring, intensive supervision, boot camps, split sentences, day reporting centers, fines, and community service. As a result of the disillusionment with the effectiveness of rehabilitation and the focus on justice and incapacitation, intermediate sanctions were proposed as an ideal way to provide a range of sanctions between probation and parole (Morris & Tonry, 1990; Tonry, 1996a). [ABSTRACT FROM AUTHOR]

INTRODUCTION Correctional boot camps are short-term incarceration programs modeled after military basic training (MacKenzie & Parent, 1992; MacKenzie & Hebert, 1996). Since their inception in the adult correctional systems in the United States in 1983 in Georgia and Oklahoma, the programs have rapidly grown, first within adult correctional systems and later in local jails and in juvenile institutions. Boot camps for juveniles were developed after the adult camps; however, they developed rapidly during the 1990s and by 2000 there were seventy juvenile camps in the United States (Koch Crime Institute, 2001). Today, correctional boot camps exist in the United States in federal, state, and local adult jurisdictions, and in state and local juvenile facilities. Correctional boot camps have also been opened in Canada and England. There are some similarities among most of the correctional boot camps. As in military basic training, inmates often enter the camps in squads or platoons. There may be an elaborate intake ceremony where inmates are immediately required to follow the rules, respond to staff in appropriate ways, stand at attention, and have their heads shaved. While in the program, participants are required to follow a rigorous daily schedule of activities including physical exercise, drill, and ceremony. They arise early each morning and are kept active for most of the day. Staff are addressed by military titles. Punishment for misbehavior is immediate and swift and frequently involves some type of physical activity such as push-ups. A graduation ceremony marks successful completion of the program. [ABSTRACT FROM AUTHOR]

INTRODUCTION The United States declared a “war on drugs” in the early 1980s. This “war” was based on a control philosophy of corrections and focused on the use of deterrence and incapacitation to try to reduce drug use and drug-related crimes. Severe penalties for drug-involved offenders were expected to reduce drug use and its related criminal activities. These policies resulted in a tremendous increase in the number of drug-involved offenders in prisons. For example, according to a 1997 survey of incarcerated offenders, 57 percent of the state inmates and 45 percent of federal inmates reported drug use in the month prior to their offense. Surveys of state inmates in Texas and Ohio reported that 56 percent and 51 percent, respectively, were diagnosed to need drug treatment. Many of these offenders were convicted of drug crimes (Lo & Stephens, 2000; Peters, Greenbaum, Edens, Carter, & Ortiz, 1998). During the 1990s, it became obvious that the war on drugs was not having the desired effect. Federal, state, and local law enforcement agencies had responded to the drug crisis by targeting drug suppliers and users. Severe penalties for such drug-involved offenders were expected to keep the offenders in prison for longer periods of time and deter others from dealing or using drugs. These tactics proved to be ineffective in reducing the supply of drugs in most communities. Drug-involved offenders continued to be a problem for criminal justice systems. [ABSTRACT FROM AUTHOR]

INTRODUCTION Drug courts are a management strategy designed to use the authority of the courts to reduce crime by changing defendants' drug-using behavior. Drug-involved defendants are diverted to drug courts in exchange for the possibility of dismissed charges or reduced sentences. Judges preside over drug court proceedings and monitor the progress of the defendants. Working in collaboration with prosecutors, defense attorneys, treatment providers, probation agents, and others, judges proscribe sanctions for noncompliance with program requirements and rewards for compliance. In comparison to traditional courtrooms, the drug court model is less punitive and more healing and restorative in nature. Courtroom procedures are more informal. Direct exchanges between the participant and the judge are common, and interactions between the state and defense are nonadversarial. A new working relationship occurs between the criminal court and health and treatment systems (Goldkamp, 1999). The role of the judge galvanizes the treatment process into a more powerful and accountable form of rehabilitation than previously available in the criminal justice system. The focus is not on the disposition of criminal cases but instead on rehabilitating drug-involved offenders. The major goals of most drug courts are to reduce drug use and criminal activity. These goals are accomplished by engaging and retaining drug-involved offenders in supervision and treatment programs. Expertise about drug cases is concentrated in a single courtroom. Advocates of the drug courts argue that use of one courtroom increases the effectiveness of both clinical assessment of defendants' needs and case management. [ABSTRACT FROM AUTHOR]

THE JUVENILE JUSTICE SYSTEM The first juvenile court in the United States was established more than one hundred years ago in Chicago. Prior to this, children as young as seven were processed the same as adults. They could stand trial in criminal court and be sentenced to prison or death. The juvenile court reflected a change in the philosophy toward juveniles. The new courts focused on the welfare of the child. The delinquent child was viewed as in need of benevolent intervention from the court. Juvenile courts sought to rehabilitate delinquents so they would become productive members of society. By 1925, all but two states had established either juvenile courts or probation services for juveniles. During the next fifty years, juvenile courts in most states had exclusive jurisdiction over all youth under age eighteen who were charged with violating criminal laws. As occurred with adult offenders in the United States, attitudes toward the rehabilitation of juveniles changed during the 1970s. There was less support for rehabilitation. The public perceived that serious juvenile crime was increasing and that the system was too lenient with juvenile delinquents. In response, many states passed more punitive laws. Some of these laws removed young lawbreakers from the juvenile system and required that they be tried in adult criminal courts. During the 1990s, all but three states passed laws designed to crack down on juvenile crime. These laws expanded eligibility for criminal court processing and adult correctional sanctioning and reduced confidentiality protections for juveniles. [ABSTRACT FROM AUTHOR]

INTRODUCTION Estimates of the occurrence of domestic violence differ depending upon the survey used to collect the data; however, it is clear that this phenomenon affects a large part of the population. According to the 1992 NCVS, more than 1 million women were victimized by intimate partners, a category that includes boyfriends, girlfriends, spouses, and former spouses, in that year (Andrews & Bonta, 2003; Healey, Smith, & O'Sullivan, 1998). According to the Bureau of Justice Statistics (BJS), approximately 840,000 women were assaulted, raped, or robbed by intimate partners in 1996 (Greenfeld et al., 1998). Estimates from the National Violence Against Women Survey indicate approximately 1.8 million women and 1 million men were raped, physically assaulted, or stalked by an intimate partner in the past twelve months (Tjaden & Thoennes, 2000). Straus, Gelles, and Steinmetz (1980) estimated that 2 million women were victimized each year by their husbands and further speculated that if ex-husbands and boyfriends were included in the calculations, the estimate could be as high as 4 million. Whichever figure is used, it is clear that domestic violence affects a large number of women sometime in their life. Physical violence can have far reaching effects on women and can continue long after surface injuries have healed. Research has shown that women who suffer domestic assault at the hands of an intimate partner are more likely to show increased rates of depression, suicide, anxiety, post-traumatic stress syndrome, low self-esteem, and substance abuse problems (Tjaden & Thoennes, 2000). [ABSTRACT FROM AUTHOR]

INTRODUCTION Educational programming in prisons originally focused on religious instruction. Such instruction was expected to help offenders achieve spiritual enlightenment (Gerber & Fritsch, 1995; Teeters, 1955). Consistent with the beliefs of those responsible for the early penitentiary movement, offenders were thought to need time to reflect on their crimes and repent. Religious instruction was expected to help in this process of penitence. However, during the reformatory era, the focus of education programs changed from religious instruction to basic literacy and communication skills. The reformatory age began in 1876 when Zebulon R. Brockway first proposed his theory of rehabilitation at the first conference of the American Prison Association (the forerunner of the current American Correctional Association). Since that time academic education has been a cornerstone of correctional programming. As superintendent of the first reformatory at Elmira, New York, Brockway brought his revolutionary ideas into practice. The goal was to reform youth. The reformatory was designed to provide a physically and mentally healthy environment where youth would have access to academic education and extensive vocational training. Brockway believed that law-abiding behavior was attainable through legitimate industry and education (Reagan & Stoughton, 1976). Elmira was used as a model for both adult prisons and juvenile reformatories throughout the United States. During the age of rehabilitation, from the turn of the century until the mid-1970s, educational programs were considered a mainstay of correctional rehabilitation. [ABSTRACT FROM AUTHOR]

SCIENTIFIC EVIDENCE FOR WHAT WORKS Strategies for reducing crime should be based on scientific evidence. By the term “evidence-based corrections,” I refer to the need to use scientific evidence to make informed decisions about correctional policy (MacKenzie, 2002; MacKenzie, 2001; MacKenzie, 2000; MacKenzie 2005). That is the subject of this book. In this book, I use four different approaches to evaluate the research and assess the effectiveness of crime reduction strategies: (1) reviews of the research literature, (2) meta-analyses used to examine groups of studies, (3) simulation studies, and (4) a scientific methods scoring system that uses the direction and significance of studies. There are several reasons for using different methodologies to review and assess the research and to draw conclusions about effectiveness in crime prevention. First, some strategies of crime prevention do not lend themselves to program evaluation that can be easily categorized using the scientific methods score designed to assess the quality of the science. Second, by using different methods to assess effectiveness, it is possible to discern agreement among the different methods. LITERATURE REVIEWS Reviews of the research literature have traditionally been used to determine whether there is evidence that some particular program, strategy, or intervention is effective in producing the intended outcome (e.g., reduce recidivism). The reviewer, usually someone who is knowledgeable about the topic area, examines the studies and draws conclusions about the findings based on the direction and significance of the results. [ABSTRACT FROM AUTHOR]

This chapter describes the goals of robust engineering and two methods to achieve these goals. The objective is to select controllable factor settings so that the effects of uncontrollable factors are not harmful. The first method presented is an extention of standard response surface methods (RSM) called RDPM. This method was originally developed in Allen et al. (2001). The second approach is the so-called static Taguchi Method. The benefits of the first method include it derives the profit optimal balance between quality and revenues and can easily handle situations involving multiple quality characteristics. Benefits of Taguchi Methods include simplicity and cost advantages in cases when all noise factors are easy-to-change. Taguchi Methods also have the obvious problem that decision-making is ambiguous if more than a single response or quality characteristic is of interest. For this reason the extension of Taguchi Methods in Song et al. (1995) is described. [ABSTRACT FROM AUTHOR]

This chapter describes the application of so-called response surface methods (RSM). These methods generally result in a relatively accurate prediction of all response variable averages related to quantities measured during experimentation. An important reason why the predictions are relatively accurate is that so-called “interactions” which relate to the combined effects of factors are included explicity in the predicted models. Three types of methods were presented. Box Behnken designs (BBDs) were argued to generate relatively accurate predictions because they minimize so-called “bias” errors under certain reasonable assumptions. Central composite designs (CCDs) were presented and explained to offer the advantage that they permit certain level adjustments and can be used in two-step sequential response surface methods. In these methods, there is a chance that the experimental will stop with relatively few runs and decide his or her prediction model is satisfactory. The third class of experimental designs presented is the expected integrated mean squared error (EIMSE) designs which are available for a variety of numbers of runs and offer predictive advantages of Box Benken designs. The EIMSE criteria is also used at the end to clarify the relative prediction errors and to help method users decide whether a given experimental design is appropriate for their own prediction accuracy goals. [ABSTRACT FROM AUTHOR]