Methods For Categorical Data Analysis Research Articles

Categorical linkage-data analysis.

Analysis of integrated data often requires record linkage in order to join together the data residing in separate sources. In case linkage errors cannot be avoided, due to the lack a unique identity key that can be used to link the records unequivocally, standard statistical techniques may produce misleading inference if the linked data are treated as if they were true observations. In this paper, we propose methods for categorical data analysis based on linked data that are not prepared by the analyst, such that neither the match-key variables nor the unlinked records are available. The adjustment is based on the proportion of false links in the linked file and our approach allows the probabilities of correct linkage to vary across the records without requiring that one is able to estimate this probability for each individual record. It accommodates also the general situation where unmatched records that cannot possibly be correctly linked exist in all the sources. The proposed methods are studied by simulation and applied to real data.

Survey of Unreported Psychiatric Morbidity Induced by Job Dissatisfaction among Teachers: Inferential Point of View

Elegant statistical methods for categorical data analysis are rapidly evolving and being adopted, particularly for biomedical and social sciences data analysis. This study presents a case study for the application of the discrete Johnson systems of distribution approach for the analysis of secondary school teachers’ job satisfaction (JS). This new approach accommodates relative frequency behavioural patterns in the analysis of categorical data using the entropy measure of discrete Johnson systems of distribution (DJSD). The approach offers a better alternative to the existing chi-square and likelihood ratio tests because it captures more shared information compared to known measures of association. A focus on the JS of about 393 teachers, showed that above 60% of the teacher’s eventually developed job dissatisfaction induced psychiatric disorders before the end of their career. Further examples were used to illustrate the applicability of the approach and enhance its reproducibility.

Analytic Hierarchy Process Analysis of Patient Preferences for Contrast-Enhanced Mammography Versus MRI as Supplemental Screening Options for Breast Cancer

ObjectiveTo guide implementation of supplemental breast screening by assessing patient preferences for contrast-enhanced mammography (CEM) versus MRI using analytic hierarchy process (AHP) methodology. MethodsIn an institutional review board–approved, HIPAA-compliant protocol, from March 23 to June 3, 2022, we contacted 579 women who had both CEM screening and MRI. Women were e-mailed an invitation to complete an online survey developed using an AHP-based model to elicit preferences for CEM or MRI. Methods for categorical data analysis were used to evaluate factors affecting preferences, under the Bonferroni correction for multiplicity. ResultsComplete responses were received from 222 (38.3%) women; the 189 women with a personal history of breast cancer had a mean age 61.8 years, and the 34 women without a personal history of breast cancer had a mean age of 53.6 years. Of 222 respondents, 157 (70.7%, confidence interval [CI]: 64.7-76.7) were determined to prefer CEM to MRI. Breast positioning was the most important criterion for 74 of 222 (33.3%) respondents, with claustrophobia, intravenous line placement, and overall stress most important for 38, 37, and 39 women (17.1%, 16.7%, and 17.6%), respectively, and noise level, contrast injection, and indifference being emphasized least frequently (by 10 [4.5%], 11 [5.0%], and 13 [5.9%] women, respectively). CEM preference was most prevalent (MRI least prevalent) for respondents emphasizing claustrophobia (37 of 38 [97%], CI: 86.2-99.9); CEM preference was least prevalent (MRI most prevalent) for respondents emphasizing breast positioning (40 of 74 [54%], CI: 42.1-65.7). ConclusionsAHP-based modeling reveals strong patient preferences for CEM over MRI, with claustrophobia favoring preference for CEM and breast positioning relatively favoring preference for MRI. Our results should help guide implementation of screening CEM and MRI.

Consumers’ Behavior in the Field of Organic Agriculture and Food Products during the COVID-19 Pandemic in the Czech Republic: Focus on a Comparison of Hyper-, Super- and Farmers’ Markets and Direct Purchases from Producers

The aim of this paper is to assess Czech food consumers’ behavior when buying organic products during the COVID-19 pandemic, with an emphasis on the place of purchase of organic agriculture and food products—especially those purchases with the shortest logistics value chain, i.e., purchase at farmers’ markets, or directly from the producer—and a comparison with the current most common places of purchase of organic products in the Czech Republic, supermarkets and hypermarkets. Categorical data analysis methods were used to create a profile of the consumer according to the most frequent purchase locations. To create mathematical–statistical models and interpretations, the methods of logistic regression, correspondence analysis and contingency table analysis were chosen. According to the results of the survey, respondents under 25 years of age are the least likely to make purchases at farmers’ markets or directly from the producer. Consumers aged 26–35 and with a university degree are the most likely to buy organic agriculture and food products at this location, followed closely by older respondents in the categories 36–45 and 46+ and with a secondary education. It is important for manufacturers to have an overview of where, in what quantities, and for what reasons consumers buy their products, especially for reasons of production optimization and planning, ecological concerns, rural development, and the impact on local areas and the value chain.

Magnetic Resonance Imaging Findings in Childhood Epilepsy at a Tertiary Hospital in Kenya.

Background: Neuroimaging is important for determining etiology and guiding care in early childhood epilepsy. However, access to appropriate imaging in sub-Saharan Africa is modest, and as a consequence, etiological descriptions of childhood epilepsy in the region have been limited. We sought to describe MRI findings in children with epilepsy presenting to a tertiary hospital in Nairobi, Kenya, over a 6-year period of routine care.Materials and Methods: We undertook a retrospective review of MRI findings of children aged between 0 and 18 years with a diagnosis of epilepsy presenting to the pediatric neurology department of Aga Khan University Hospital in Nairobi, Kenya, between January 2014 and July 2020. Over this period, the hospital had 1.5T MRI machines (GE1.5T Signa Excite and GE 1.5T Signa Explorer) and a 3T MRI machine (Philips 3T Ingenia). MRI images were independently reviewed by two study radiologists, and the findings were summarized and categorized into a study database. Related clinical and electroencephalographic (EEG) details were extracted from patient records. Categorical data analysis methods were applied to investigate for relationships between clinically relevant neuroimaging findings and key clinical and EEG observations.Results: Over the study period, 288 children with a confirmed diagnosis of epilepsy had an MRI. They were of median age of 6 [interquartile range (IQR) 2–11] years. Ninety-five (33%) children had abnormal findings on imaging. The most common findings were encephalomalacia related to chronic infarcts (n = 18: 6.3%), cerebral atrophy (n = 11: 3.8%), disorders of neuronal migration (n = 11: 3.8%), periventricular leukomalacia (n = 9: 3.1%), and hippocampal sclerosis (n = 8: 2.8%). Findings related to infectious etiology were only observed in four children. Clinical comorbidity and inter-ictal epileptiform activity on EEG were independently associated with abnormal findings on imaging.Conclusion: Up to a third of the children who underwent an MRI had a positive yield for abnormal findings. Imaging findings related to infectious etiologies were little observed in our cohort, in contradistinction to etiology studies in similar settings. At the time of the study, comorbidity and inter-ictal epileptiform activity on EEG were associated with abnormal findings on imaging and should be considered in informing prioritization for imaging in childhood epilepsy in this setting.

Differentiating between fatal and non-fatal mining accidents using artificial intelligence techniques

ABSTRACT Using statistical methods for categorical data analysis, namely multiple correspondence analysis and Artificial Intelligence through Bayesian networks, we analysed a database of occupational mining accidents for Spain for the period 2004–2017 to identify the factors most associated with the occurrence of fatal and non-fatal accidents. The results obtained allow to shed light on the hidden patterns present in different work situations where accidents can have fatal consequences. In addition, this study exemplifies the application of statistical techniques suitable for Big Data and data-driven decision making in the mining sector.

Multinomial Logistic Modelling of Socio-Economic Factors Influencing Spending Behavior of University Students

This study aims at determining the use of Multinomial Logistic Regression (MLR) model which is one of the important methods for categorical data analysis. This model particularly deals with one nominal or ordinal response variable that has more than two categories. Despite the fact that many researchers have applied this model in data analysis in many areas, for instance behavioral, social, health, and educational, a study on spending habits of University students have never been done. To identify the model by practical way, we conducted a survey research among students from University of Embu. Segment of the population of students in undergraduate level, a sample of 376 was selected. We employed the use stratified random sampling and simple random sampling without replacement in each stratum. The response variable consisted of five categories. Four of explanatory variables were used for building the primary (MLR) model. The model was tested through a set of statistical tests to ensure its appropriateness for the data. From the results, the study reveals that year of study, family financial level, gender and school are significant factors in explaining spending habits of students. Despite the fact that gender is one of the deterministic factors of financial behavior of student, this model identified family level of income as a major deterministic factor. Conclusively, using MLR model accurately defines the relationship between the group of explanatory variables and the response variable. It also identifies the effect of each of the variables, and we can predict the classification of any individual case. The researchers recommend that, the Universities peer counselling department, should hold trainings on the basis of major determinant of financial spending behavior i.e. family financial level.

Zaufanie do lekarzy oraz przekonania osób starszych na temat efektywności medycyny w walce z chorobami nowotworowymi

Wstęp: Polska należy do krajów o najwyższych wskaźnikach umieralności na raka. Ponad 60% zachorowań na raka w Europie i USA odnotowuje się u osób powyżej 65. r.ż. W walce z chorobami nowotworowymi niezbędne są nie tylko działania makrostrukturalne, lecz również poprawa samej relacji lekarz–pacjent. Ważnym elementem tej relacji jest zaufanie pacjenta.W niniejszym artykule przedstawione zostały badania dotyczące zaufania osób starszych do lekarzy oraz ich przekonania na temat efektywności medycyny w walce z chorobami nowotworowymi, atakże wpływ takich zmiennych jak płeć, status ekonomiczny i wykształcenie na poziom zaufania do lekarza i medycyny.Materiały i metody: Badania socjologiczne zrealizowano w 2012 r. we Wrocławiu wśród 329 starszych osób. Zastosowano próbę kwotową. Jako metodę zastosowano wywiad kwestionariuszowy owysokim poziomie standaryzacji. Zastosowano dostępne metody analizy danych jakościowych.Wyniki: Większość badanych deklaruje zaufanie do lekarza, ale w różnym stopniu. Poziom zaufania do lekarza zależny jest od dochodów i wykształcenia; 44,98% zgadza się ze stwierdzeniem, że„współczesna medycyna (…) skutecznie radzi sobie z leczeniem nowotworów”, natomiast 55,02% respondentów nie zgadza się z tym stwierdzeniem. Istnieje zależność pomiędzy tym przekonaniem awysokością dochodów badanych.Wnioski: Poziom zaufania do lekarzy i medycyny jest zróżnicowany ze względu na zmienne społeczno-demograficzne wyznaczające nierówności w sferze zdrowia, tj. wiek, sytuację ekonomiczną ipoziom wykształcenia. Uzyskane wyniki posłużą do wypracowania skuteczniejszych działań z zakresu promocji zdrowia i profilaktyki onkologicznej ukierunkowanych na osoby starsze.

Repulsed by the “Other”

We offer an integration of theory and method in the study of intergroup social associations. Specifically, we show that models for intergroup association tables developed using generic log-linear methods for categorical data analysis embody a general theoretical point of view on the driving force behind intergroup association, namely, as the outcome of a probabilistic process of repulsion from dissimilar others. We develop this argument and illustrate it with intermarriage data. We conclude by identifying the advantages that accrue to both theory and method when the theoretical assumptions underlying the application of a generic statistical methodology are clearly understood.

Effects of Personal Selling on Sales: A Case of Women Groups in Imenti North District, Meru County, Kenya

As firms increase in number, modern marketing calls for more than just producing the products, pricing them and making them accessible to the target market, in addition, they must also communicate to present and potential consumers and the general public. The objective of this study was to evaluate the effects of personal selling on sales of agricultural products produced by women groups. The population of the study comprised of 100 women groups registered with the Ministry of Sports, Culture and Social Services by December 2012 in Imenti North District in Meru County Kenya engaging in agricultural activities to generate income. Simple random sampling technique was used to determine the sample size of 79. One official from each of these groups was picked as the respondent. Primary data was collected using structured questionnaires which were administered personally. Data was presented by use of frequency tables then analyzed using descriptive statistical techniques such as frequencies and percentages. Categorical data analysis methods were used to aggregate the influence of promotion elements on sales. Chi Square test was used to test the hypotheses. Personal selling was found to be widely used promotion method by women groups where they largely target women. Personal selling was found to have influence on sales of agricultural products produced by women groups. Background of the Study Globally, there has been a call to uplift the standard of women in the society. Hence, women do come together to combine their efforts through the formation of women groups. According to Chun (2005) in china women come together to uplift their economic standards through producing handcrafts and interior decorations for sale, they use price and promotion as a marketing tool. As Aidoo (1995) points out in Ghana rural women have been empowered to form self help groups where they engage in various income generating activities. Ghana women form groups to facilitate succession, wealthy inheritance, affiliation, residence and economic potential (Aidoo, 1995). This has enabled them to uplift their living standards. In Kenya maendeleo ya wanawake was formed at independence to uplift and empower women as a means of alleviating poverty and creating a better environment and quality of life. Other

Multivariate data in the estimation of consumer risk

The risk of consumer behaviour, as a part of the widely understood studies on risk, is still an uncharted and undiscovered area of human activity. The main goal of this paper is to draw attention to the issue of the measurement of risk perceived by the consumers` unsuccessful purchase, as well as presenting a multidimensional analysis of data on risk research perceived by consumers in the decision making process. Some of the well-known multivariate methods are presented: analysis of variance, correspondence analysis and some graphical methods for categorical data analysis, such as mosaic, sieve, association and doubledecker plot. In the paper, the qualitative analysis aimed at risk identification and interpretation in the decisions process of consumers will be conducted. The exploration of different types of risks and the influence on consumer behaviour will be identified. The perception of risk was examined based on the examples of food, home appliances and travel services (trips, holidays).

A Categorical Data Analysis on Financial Failures in Vietnam, 2007-2013

In this paper, we examined 256 cases of financial failure and fraud occuring during the recent Vietnam’s chaotic years from 2007 to 2013, employing methods of categorical data analysis. Reported results suggest that the rent-seeking approach, or resource-based orientation, alone does not help explain the outcome of a business intention while the association between Orientation and Approach is the best-fit predictor. Rampant financial collapse not only increases the cost of funds but also erodes trust in the economy. Entrepreneurship development and creativity capacity building are necessary to improve socio-economic conditions and the environment. This work also introduces intuitive and cognitive factors to predict ex-ante outcome of a financing scheme.

Section Editor's Notes

This is my first issue as associate editor of software reviews for The American Statistician. In this column, I will introduce myself, comment on the types of software reviews that can be published in this section of The American Statistician, and encourage others in the profession to consider taking on the task of reviewing statistical software packages. I first began reading statistical software reviews in this journal when enrolled in my doctoral program in biostatistics. I read a number of excellent reviews and found the information they contained helpful in my own research, in class projects, and in collaborative work that I was doing at the time. I became interested in writing software reviews while completing my dissertation and contacted Dr. Richard Goldstein, at that time the journal’s software reviews editor. Dr. Goldstein explained the process of submitting software reviews, gave me valuable pointers on writing these reviews as well as obtaining the statistical software that I was interested in evaluating, and helped me focus my ideas. I completed my first software review after finishing my dissertation. During this period, Dr. Joseph Hilbe assumed the role of software reviews editor. Both Drs. Goldstein and Hilbe were instrumental in helping me get that first review published. That review focused on statistical software designed specifically for, or with strong components in, analyses of data obtained from epidemiology studies (Oster 1998). I subsequently developed an interest in statistical research in categorical data analysis methods in general, and in exact statistical methods in particular. Since then, I have authored (and coauthored with Dr. Hilbe) numerous comparative reviews regarding categorical data analysis and exact statistical methods (Oster 2002, 2003; Oster and Hilbe 2003, 2005, 2008a, 2008b). In these reviews, there is an emphasis on evaluating the implementation and performance of statistical methods in the various software packages. Writing these reviews has been a rewarding experience for me personally and also for my career. Various types of software reviews have appeared in The American Statistician. These reviews have traditionally been of the comparative type. A comparative review involves the evaluation of two or more software packages with a focus on a specific statistical or scientific area (e.g., Oster and Hilbe 2008a, 2008b). This type of review tends to be thorough and involves an evaluation of all capabilities of each software package with

An Application on Multinomial Logistic Regression Model

Normal 0 false false false EN-US X-NONE X-NONE This study aims to identify an application of Multinomial Logistic Regression model which is one of the important methods for categorical data analysis. This model deals with one nominal/ordinal response variable that has more than two categories, whether nominal or ordinal variable. This model has been applied in data analysis in many areas, for example health, social, behavioral, and educational.To identify the model by practical way, we used real data on physical violence against children, from a survey of Youth 2003 which was conducted by Palestinian Central Bureau of Statistics (PCBS). Segment of the population of children in the age group (10-14 years) for residents in Gaza governorate, size of 66,935 had been selected, and the response variable consisted of four categories. Eighteen of explanatory variables were used for building the primary multinomial logistic regression model. Model had been tested through a set of statistical tests to ensure its appropriateness for the data. Also the model had been tested by selecting randomly of two observations of the data used to predict the position of each observation in any classified group it can be, by knowing the values of the explanatory variables used. We concluded by using the multinomial logistic regression model that we can able to define accurately the relationship between the group of explanatory variables and the response variable, identify the effect of each of the variables, and we can predict the classification of any individual case.

On Nonparametric Statistical Process Control of Univariate Processes

This article considers statistical process control (SPC) of univariate processes when the parametric form of the process distribution is unavailable. Most existing SPC procedures are based on the assumption that a parametric form (e.g., normal) of the process distribution can be specified beforehand. In the literature, it has been demonstrated that their performance is unreliable in cases when the prespecified process distribution is invalid. To overcome this limitation, some nonparametric (or distribution-free) SPC charts have been proposed, most of which are based on the ordering information of the observed data. This article tries to make two contributions to the nonparametric SPC literature. First, we propose an alternative framework for constructing nonparametric control charts, by first categorizing observed data and then applying categorical data analysis methods to SPC. Under this framework, some new nonparametric control charts are proposed. Second, we compare our proposed control charts with several representative existing control charts in various cases. Some empirical guidelines are provided for users to choose a proper nonparametric control chart for a specific application. This article has supplementary materials online.

Statistical Methods for Categorical Data Analysis, Second Edition by Daniel A. Powers, Yu Xie

Statistical Methods for Categorical Data Analysis, Second Edition by Daniel A. Powers, Yu Xie

Bayesian inference for categorical data analysis

This article surveys Bayesian methods for categorical data analysis, with primary emphasis on contingency table analysis. Early innovations were proposed by Good (1953, 1956, 1965) for smoothing proportions in contingency tables and by Lindley (1964) for inference about odds ratios. These approaches primarily used conjugate beta and Dirichlet priors. Altham (1969, 1971) presented Bayesian analogs of small-sample frequentist tests for 2 x 2 tables using such priors. An alternative approach using normal priors for logits received considerable attention in the 1970s by Leonard and others (e.g., Leonard 1972). Adopted usually in a hierarchical form, the logit-normal approach allows greater flexibility and scope for generalization. The 1970s also saw considerable interest in loglinear modeling. The advent of modern computational methods since the mid-1980s has led to a growing literature on fully Bayesian analyses with models for categorical data, with main emphasis on generalized linear models such as logistic regression for binary and multi-category response variables.

OCCUPATIONAL RISKY BUSINESS: INJURY PREVENTION BEHAVIORS OF FARM WOMEN AND CHILDREN

On farms in the United States, there are approximately 100 fatal and 32,800 nonfatal injuries annually in children 19 years and younger (). Up to 40% of nonfatally injured children are left with permanent disabilities. The impact of injury and death on children associated with farming in the United States is substantial.Research suggests that modeling of health behaviors may be an effective technique for the socialization of children's health behaviors. The purpose of this study is twofold. First, the study describes the participation and use of protective farm equipment or practices by the caregiver and the child. The second purpose is to compare the practices of the caregivers and the children in relation to the use of protective farm equipment and practices.Descriptive and categorical data analysis methods were used to examine the associations of 177 pairs of caregivers and children and their use of protective equipment and preventative behaviors. Greater than 50% of the children under the age of 7 had handled or touched livestock, as well as rode as passengers on tractors. Many children in all age categories had ridden in the back of pick-up trucks. Modeling of the prevention behavior by the caregivers did parallel the use of the equipment by the children. However, the opposite also was true; if the caregiver did not use the equipment, neither did the child.

CHILDREN'S INJURIES IN AGRICULTURE RELATED EVENTS: THE EFFECT OF SUPERVISION ON THE INJURY EXPERIENCE

An international health problem and the leading cause of death and disability among children in the United States are unintentional injuries. Children in rural areas in the United States have the highest death rate related to unintentional injuries regardless of age (Crawley, 1996). Using Haddon's Injury Model as the theoretical framework, the purpose of this study is threefold. First the study describes actual injuries that were sustained by farm children. Second, the research identifies the type of supervision the farm children and adolescents were receiving at the time of the injury, and finally the study examines injury risk in relation to supervision. Descriptive and categorical data analysis methods were used to examine the associations between farm-related injury and supervision type. Out of 177 children living in the home under 18 years of age, 32 children sustained at least one injury and eight sustained two injuries within one year from the time of the survey. The majority of children needed medical attention because of their injuries ( n = 37). Children were more likely to sustain farm-related injury when they were supervised by a caregiver engaged in farm work versus supervised at home ( p = .007). The findings of this study support Haddon's Injury Model, which suggests injuries occur because of an uncontrolled interaction between a host, an agent, and the environment. Examining the children's role within the framework of Haddon's Injury Model, will assist researchers in designing evidenced-based research that addresses the interaction between the host, agent, and environmental factors. Results from these studies will be useful in identifying effective interventions in the pre-event phase, as well as maximizing quality of life in the postevent phase.

Applications of bootstrap methods for categorical data analysis

Simultaneous confidence regions for proportions of a single multinomial population, and for a finite number of contrasts from several multinomial populations are proposed. In this paper bootstrap methods are used to construct confidence regions. We compare the performance of bootstrap methods with other methods in terms of average coverage probability by Monte Carlo simulation. Advantages of the bootstrap methods are discussed.