Model selection of GLMMs in the analysis of count data in single-case studies: A Monte Carlo simulation.

Generalized linear mixed models (GLMMs) have great potential to deal with count data in single-case experimental designs (SCEDs). However, applied researchers have faced challenges in making various statistical decisions when using such advanced statistical techniques in their own research. This study focused on a critical issue by investigating the selection of an appropriate distribution to handle different types of count data in SCEDs due to overdispersion and/or zero-inflation. To achieve this, I proposed two model selection frameworks, one based on calculating information criteria (AIC and BIC) and another based on utilizing a multistage-model selection procedure. Four data scenarios were simulated including Poisson, negative binominal (NB), zero-inflated Poisson (ZIP), and zero-inflated negative binomial (ZINB). The same set of models (i.e., Poisson, NB, ZIP, and ZINB) were fitted for each scenario. In the simulation, I evaluated 10 model selection strategies within the two frameworks by assessing the model selection bias and its consequences on the accuracy of the treatment effect estimates and inferential statistics. Based on the simulation results and previous work, I provide recommendations regarding which model selection methods should be adopted in different scenarios. The implications, limitations, and future research directions are also discussed.

An interpretable Bayesian clustering approach with feature selection for analyzing spatially resolved transcriptomics data.

Recent breakthroughs in spatially resolved transcriptomics (SRT) technologies have enabled comprehensive molecular characterization at the spot or cellular level while preserving spatial information. Cells are the fundamental building blocks of tissues, organized into distinct yet connected components. Although many non-spatial and spatial clustering approaches have been used to partition the entire region into mutually exclusive spatial domains based on the SRT high-dimensional molecular profile, most require an ad hoc selection of less interpretable dimensional-reduction techniques. To overcome this challenge, we propose a zero-inflated negative binomial mixture model to cluster spots or cells based on their molecular profiles. To increase interpretability, we employ a feature selection mechanism to provide a low-dimensional summary of the SRT molecular profile in terms of discriminating genes that shed light on the clustering result. We further incorporate the SRT geospatial profile via a Markov random field prior. We demonstrate how this joint modeling strategy improves clustering accuracy, compared with alternative state-of-the-art approaches, through simulation studies and 3 real data applications.

A robust regression model for bounded count health data.

Bounded count response data arise naturally in health applications. In general, the well-known beta-binomial regression model form the basis for analyzing this data, specially when we have overdispersed data. Little attention, however, has been given to the literature on the possibility of having extreme observations and overdispersed data. We propose in this work an extension of the beta-binomial regression model, named the beta-2-binomial regression model, which provides a rather flexible approach for fitting a regression model with a wide spectrum of bounded count response data sets under the presence of overdispersion, outliers, or excess of extreme observations. This distribution possesses more skewness and kurtosis than the beta-binomial model but preserves the same mean and variance form of the beta-binomial model. Additional properties of the beta-2-binomial distribution are derived including its behavior on the limits of its parametric space. A penalized maximum likelihood approach is considered to estimate parameters of this model and a residual analysis is included to assess departures from model assumptions as well as to detect outlier observations. Simulation studies, considering the robustness to outliers, are presented confirming that the beta-2-binomial regression model is a better robust alternative, in comparison with the binomial and beta-binomial regression models. We also found that the beta-2-binomial regression model outperformed the binomial and beta-binomial regression models in our applications of predicting liver cancer development in mice and the number of inappropriate days a patient spent in a hospital.

Counting the days: Exploring the post-mortem interval factors in sexual homicides.

This study explores the offender, victim, and environmental characteristics that significantly influence the number of days a sexual homicide victim remains undiscovered. Utilizing a sample of 269 cases from the Homicide Investigation Tracking System database an in-depth analysis was conducted to unveil the factors contributing to the delay in the discovery of victims' bodies. The methodological approach involves applying a negative binomial regression analysis, which allows for the examination of count data, specifically addressing the over-dispersion and excess zeros in the dependent variable - the number of days until the victim is found. The findings reveal that certain offender characteristics, victim traits, and spatio-temporal factors play a pivotal role in the time lag experienced in locating the bodies of homicide victims. These findings have crucial implications for investigative efforts in homicide cases, offering valuable insights that can inform and enhance the efficacy and efficiency of future investigative procedures and strategies.

A flexible time-varying coefficient rate model for panel count data.

Panel count regression is often required in recurrent event studies, where the interest is to model the event rate. Existing rate models are unable to handle time-varying covariate effects due to theoretical and computational difficulties. Mean models provide a viable alternative but are subject to the constraints of the monotonicity assumption, which tends to be violated when covariates fluctuate over time. In this paper, we present a new semiparametric rate model for panel count data along with related theoretical results. For model fitting, we present an efficient EM algorithm with three different methods for variance estimation. The algorithm allows us to sidestep the challenges of numerical integration and difficulties with the iterative convex minorant algorithm. We showed that the estimators are consistent and asymptotically normally distributed. Simulation studies confirmed an excellent finite sample performance. To illustrate, we analyzed data from a real clinical study of behavioral risk factors for sexually transmitted infections.

Incorporating calibration errors in oral reading fluency scoring.

Oral reading fluency (ORF) assessments are commonly used to screen at-risk readers and evaluate interventions' effectiveness as curriculum-based measurements. Similar to the standard practice in item response theory (IRT), calibrated passage parameter estimates are currently used as if they were population values in model-based ORF scoring. However, calibration errors that are unaccounted for may bias ORF score estimates and, in particular, lead to underestimated standard errors (SEs) of ORF scores. Therefore, we consider an approach that incorporates the calibration errors in latent variable scores. We further derive the SEs of ORF scores based on the delta method to incorporate the calibration uncertainty. We conduct a simulation study to evaluate the recovery of point estimates and SEs of latent variable scores and ORF scores in various simulated conditions. Results suggest that ignoring calibration errors leads to underestimated latent variable score SEs and ORF score SEs, especially when the calibration sample is small.

Analyzing health outcomes measured as bounded counts.

This paper assesses analytical strategies that respect the bounded-count nature of health outcomes encountered often in empirical applications. Absent in the literature is a comprehensive discussion and critique of strategies for analyzing and understanding such data. The paper's goal is to provide an in-depth consideration of prominent issues arising in and strategies for undertaking such analyses, emphasizing the merits and limitations of various analytical tools empirical researchers may contemplate. Three main topics are covered. First, bounded-count health outcomes' measurement properties are reviewed and their implications assessed. Second, issues arising when bounded-count outcomes are the objects of concern in evaluations are described. Third, the (conditional) probability and moment structures of bounded-count outcomes are derived and corresponding specification and estimation strategies presented with particular attention to partial effects. Many questions may be asked of such data in health research and a researcher's choice of analytical method is often consequential.

High-dimensional covariate-augmented overdispersed poisson factor model.

The current Poisson factor models often assume that the factors are unknown, which overlooks the explanatory potential of certain observable covariates. This study focuses on high dimensional settings, where the number of the count response variables and/or covariates can diverge as the sample size increases. A covariate-augmented overdispersed Poisson factor model is proposed to jointly perform a high-dimensional Poisson factor analysis and estimate a large coefficient matrix for overdispersed count data. A group of identifiability conditions is provided to theoretically guarantee computational identifiability. We incorporate the interdependence of both response variables and covariates by imposing a low-rank constraint on the large coefficient matrix. To address the computation challenges posed by nonlinearity, two high-dimensional latent matrices, and the low-rank constraint, we propose a novel variational estimation scheme that combines Laplace and Taylor approximations. We also develop a criterion based on a singular value ratio to determine the number of factors and the rank of the coefficient matrix. Comprehensive simulation studies demonstrate that the proposed method outperforms the state-of-the-art methods in estimation accuracy and computational efficiency. The practical merit of our method is demonstrated by an application to the CITE-seq dataset. A flexible implementation of our proposed method is available in the R package COAP.

Mapping QTL controlling count traits with excess zeros and ones using a zero-and-one-inflated generalized Poisson regression model.

The research on the quantitative trait locus (QTL) mapping of count data has aroused the wide attention of researchers. There are frequent problems in applied research that limit the application of the conventional Poisson model in the analysis of count phenotypes, which include the overdispersion and excess zeros and ones. In this article, a novel model, that is, the zero-and-one-inflated generalized Poisson (ZOIGP) model, is proposed to deal with these problems. Based on the proposed model, a score test is performed for the inflation parameter, in which the ZOIGP model with a constant proportion of excess zeros and ones is compared with a standard generalized Poisson model. To illustrate the practicability of the ZOIGP model, we extend it to the QTL interval mapping application that underpins count phenotype with excess zeros and excess ones. The genetic effects are estimated utilizing the expectation-maximization algorithm embedded with the Newton-Raphson algorithm, and the genome-wide scan and likelihood ratio test is performed to map and test the potential QTLs. The statistical properties exhibited by the proposed method are investigated through simulation. Finally, a real data analysis example is used to illustrate the utility of the proposed method for QTL mapping.

A novel correction method for modelling parameter-driven autocorrelated time series with count outcome.

BACKGROUND: Count time series (e.g., daily deaths) are a very common type of data in environmental health research. The series is generally autocorrelated, while the widely used generalized linear model is based on the assumption of independent outcomes. None of the existing methods for modelling parameter-driven count time series can obtain consistent and reliable standard error of parameter estimates, causing potential inflation of type I error rate. METHODS: We proposed a new maximum significant ρ correction (MSRC) method that utilizes information of significant autocorrelation coefficient ρ estimate within 5 orders by moment estimation. A Monte Carlo simulation was conducted to evaluate and compare the finite sample performance of the MSRC and classical unbiased correction (UB-corrected) method. We demonstrated a real-data analysis for assessing the effect of drunk driving regulations on the incidence of road traffic injuries (RTIs) using MSRC in Shenzhen, China. Moreover, there is no previous paper assessing the time-varying intervention effect and considering autocorrelation based on daily data of RTIs. RESULTS: Both methods had a small bias in the regression coefficients. The autocorrelation coefficient estimated by UB-corrected is slightly underestimated at high autocorrelation (≥ 0.6), leading to the inflation of the type I error rate. The new method well controlled the type I error rate when the sample size reached 340. Moreover, the power of MSRC increased with increasing sample size and effect size and decreasing nuisance parameters, and it approached UB-corrected when ρ was small (≤ 0.4), but became more reliable as autocorrelation increased further. The daily data of RTIs exhibited significant autocorrelation after controlling for potential confounding, and therefore the MSRC was preferable to the UB-corrected. The intervention contributed to a decrease in the incidence of RTIs by 8.34% (95% CI, -5.69-20.51%), 45.07% (95% CI, 25.86-59.30%) and 42.94% (95% CI, 9.56-64.00%) at 1, 3 and 5 years after the implementation of the intervention, respectively. CONCLUSIONS: The proposed MSRC method provides a reliable and consistent approach for modelling parameter-driven time series with autocorrelated count data. It offers improved estimation compared to existing methods. The strict drunk driving regulations can reduce the risk of RTIs.

Socio-environmental predictors of diabetes incidence disparities in Tanzania mainland: a comparison of regression models for count data.

BACKGROUND: Diabetes is one of the top four non-communicable diseases that cause death and illness to many people around the world. This study aims to use an efficient count data model to estimate socio-environmental factors associated with diabetes incidences in Tanzania mainland, addressing lack of evidence on the efficient count data model for estimating factors associated with disease incidences disparities. METHODS: This study analyzed diabetes counts in 184 Tanzania mainland councils collected in 2020. The study applied generalized Poisson, negative binomial, and Poisson count data models and evaluated their adequacy using information criteria and Pearson chi-square values. RESULTS: The data were over-dispersed, as evidenced by the mean and variance values and the positively skewed histograms. The results revealed uneven distribution of diabetes incidence across geographical locations, with northern and urban councils having more cases. Factors like population, GDP, and hospital numbers were associated with diabetes counts. The GP model performed better than NB and Poisson models. CONCLUSION: The occurrence of diabetes can be attributed to geographical locations. To address this public health issue, environmental interventions can be implemented. Additionally, the generalized Poisson model is an effective tool for analyzing health information system count data across different population subgroups.

Conditional modeling of panel count data with partly interval-censored failure event.

In longitudinal follow-up studies, panel count data arise from discrete observations on recurrent events. We investigate a more general situation where a partly interval-censored failure event is informative to recurrent events. The existing methods for the informative failure event are based on the latent variable model, which provides indirect interpretation for the effect of failure event. To solve this problem, we propose a failure-time-dependent proportional mean model with panel count data through an unspecified link function. For estimation of model parameters, we consider a conditional expectation of least squares function to overcome the challenges from partly interval-censoring, and develop a two-stage estimation procedure by treating the distribution function of the failure time as a functional nuisance parameter and using the B-spline functions to approximate unknown baseline mean and link functions. Furthermore, we derive the overall convergence rate of the proposed estimators and establish the asymptotic normality of finite-dimensional estimator and functionals of infinite-dimensional estimator. The proposed estimation procedure is evaluated by extensive simulation studies, in which the finite-sample performances coincide with the theoretical results. We further illustrate our method with a longitudinal healthy longevity study and draw some insightful conclusions.

Simultaneous variable selection and estimation in semiparametric regression of mixed panel count data.

Mixed panel count data represent a common complex data structure in longitudinal survey studies. A major challenge in analyzing such data is variable selection and estimation while efficiently incorporating both the panel count and panel binary data components. Analyses in the medical literature have often ignored the panel binary component and treated it as missing with the unknown panel counts, while obviously such a simplification does not effectively utilize the original data information. In this research, we put forward a penalized likelihood variable selection and estimation procedure under the proportional mean model. A computationally efficient EM algorithm is developed that ensures sparse estimation for variable selection, and the resulting estimator is shown to have the desirable oracle property. Simulation studies assessed and confirmed the good finite-sample properties of the proposed method, and the method is applied to analyze a motivating dataset from the Health and Retirement Study.

Large-scale data decipher children's scale errors: A meta-analytic approach using the zero-inflated Poisson models.

Scale errors are intriguing phenomena in which a child tries to perform an object-specific action on a tiny object. Several viewpoints explaining the developmental mechanisms underlying scale errors exist; however, there is no unified account of how different factors interact and affect scale errors, and the statistical approaches used in the previous research do not adequately capture the structure of the data. By conducting a secondary analysis of aggregated datasets across nine different studies (n = 528) and using more appropriate statistical methods, this study provides a more accurate description of the development of scale errors. We implemented the zero-inflated Poisson (ZIP) regression that could directly handle the count data with a stack of zero observations and regarded developmental indices as continuous variables. The results suggested that the developmental trend of scale errors was well documented by an inverted U-shaped curve rather than a simple linear function, although nonlinearity captured different aspects of the scale errors between the laboratory and classroom data. We also found that repeated experiences with scale error tasks reduced the number of scale errors, whereas girls made more scale errors than boys. Furthermore, a model comparison approach revealed that predicate vocabulary size (e.g., adjectives or verbs), predicted developmental changes in scale errors better than noun vocabulary size, particularly in terms of the presence or absence of scale errors. The application of the ZIP model enables researchers to discern how different factors affect scale error production, thereby providing new insights into demystifying the mechanisms underlying these phenomena. A video abstract of this article can be viewed at https://youtu.be/1v1U6CjDZ1Q RESEARCH HIGHLIGHTS: We fit a large dataset by aggregating the existing scale error data to the zero-inflated Poisson (ZIP) model. Scale errors peaked along the different developmental indices, but the underlying statistical structure differed between the in-lab and classroom datasets. Repeated experiences with scale error tasks and the children's gender affected the number of scale errors produced per session. Predicate vocabulary size (e.g., adjectives or verbs) better predicts developmental changes in scale errors than noun vocabulary size.

Integrated distance sampling models for simple point counts.

Point counts (PCs) are widely used in biodiversity surveys but, despite numerous advantages, simple PCs suffer from several problems: detectability, and therefore abundance, is unknown; systematic spatiotemporal variation in detectability yields biased inferences, and unknown survey area prevents formal density estimation and scaling-up to the landscape level. We introduce integrated distance sampling (IDS) models that combine distance sampling (DS) with simple PC or detection/nondetection (DND) data to capitalize on the strengths and mitigate the weaknesses of each data type. Key to IDS models is the view of simple PC and DND data as aggregations of latent DS surveys that observe the same underlying density process. This enables the estimation of separate detection functions, along with distinct covariate effects, for all data types. Additional information from repeat or time-removal surveys, or variable survey duration, enables the separate estimation of the availability and perceptibility components of detectability with DS and PC data. IDS models reconcile spatial and temporal mismatches among data sets and solve the above-mentioned problems of simple PC and DND data. To fit IDS models, we provide JAGS code and the new "IDS()" function in the R package unmarked. Extant citizen-science data generally lack the information necessary to adjust for detection biases, but IDS models address this shortcoming, thus greatly extending the utility and reach of these data. In addition, they enable formal density estimation in hybrid designs, which efficiently combine DS with distance-free, point-based PC or DND surveys. We believe that IDS models have considerable scope in ecology, management, and monitoring.

A new modified biased estimator for Zero inflated Poisson regression model.

Zero-inflated Poisson (ZIP) model is widely used for counting data with excessive zeroes. The multicollinearity is the common factor in the explanatory variables of the count data. In this context, typically, maximum likelihood estimation (MLE) generates unsatisfactory results due to inflation of mean square error (MSE). In the solution of this problem usually, ridge parameters are used. In this study, we proposed a new modified zero-inflated Poisson ridge regression model to reduce the problem of multicollinearity. We experimented within the context of a specified simulation strategy and recorded the behavior of proposed estimators. We also apply our proposed estimator to the real-life data set and explore how our proposed estimators perform well in the presence of multicollinearity with the help of ZIP model for count data.

Determination of correlations in multivariate count data with informative observation times.

We consider there are various types of recurrent events and the total number of occurrences are collected at the random observation times. It has concerned that the observation process may not be independent to the multivariate event processes, hence the total counts and observation times may be correlated and the dependence may exist among different types of the event processes as well. Many methods have developed nonparametric models to accommodate such unknown structures; however, it is difficult to assess and directly quantify their correlation relationships. A multivariate frailty model is proposed to this study, in which the event and observation processes are linked by frailty variables whose joint distribution can be implicitly specified through the multivariate normal distribution with some unknown covariance matrix. The Bayesian inference method is conducted to obtain the estimates of the regression coefficients and correlation parameters. We use a form of trigonometric functions to represent the covariance matrix, so that it meets the positive-definiteness condition efficiently during the estimation schemes. The simulation studies demonstrate the utility of the proposed models. We apply the model to a skin cancer prevention study, and aim to determine the covariate and association effects. We found treatment is significant in determining the duration of examination times; prior-counts, age and gender are significant variables on the occurrence rates of tumor counts. Using the covariance matrix to access the underlying dependent structure, the mutual correlations among them are all positive, and the basal cell counts are more related to the examination times.

Multilevel modeling in single-case studies with zero-inflated and overdispersed count data.

Count outcomes are frequently encountered in single-case experimental designs (SCEDs). Generalized linear mixed models (GLMMs) have shown promise in handling overdispersed count data. However, the presence of excessive zeros in the baseline phase of SCEDs introduces a more complex issue known as zero-inflation, often overlooked by researchers. This study aimed to deal with zero-inflated and overdispersed count data within a multiple-baseline design (MBD) in single-case studies. It examined the performance of various GLMMs (Poisson, negative binomial [NB], zero-inflated Poisson [ZIP], and zero-inflated negative binomial [ZINB] models) in estimating treatment effects and generating inferential statistics. Additionally, a real example was used to demonstrate the analysis of zero-inflated and overdispersed count data. The simulation results indicated that the ZINB model provided accurate estimates for treatment effects, while the other three models yielded biased estimates. The inferential statistics obtained from the ZINB model were reliable when the baseline rate was low. However, when the data were overdispersed but not zero-inflated, both the ZINB and ZIP models exhibited poor performance in accurately estimating treatment effects. These findings contribute to our understanding of using GLMMs to handle zero-inflated and overdispersed count data in SCEDs. The implications, limitations, and future research directions are also discussed.

Understanding Ability and Reliability Differences Measured with Count Items: The Distributional Regression Test Model and the Count Latent Regression Model.

In psychology and education, tests (e.g., reading tests) and self-reports (e.g., clinical questionnaires) generate counts, but corresponding Item Response Theory (IRT) methods are underdeveloped compared to binary data. Recent advances include the Two-Parameter Conway-Maxwell-Poisson model (2PCMPM), generalizing Rasch's Poisson Counts Model, with item-specific difficulty, discrimination, and dispersion parameters. Explaining differences in model parameters informs item construction and selection but has received little attention. We introduce two 2PCMPM-based explanatory count IRT models: The Distributional Regression Test Model for item covariates, and the Count Latent Regression Model for (categorical) person covariates. Estimation methods are provided and satisfactory statistical properties are observed in simulations. Two examples illustrate how the models help understand tests and underlying constructs.

Importance of temporary and permanent snow for new second homes.

This study investigates empirically how natural snow depth and permanent snow affect the number of new second homes in Norway. One out of four Norwegian municipalities is partly covered by glaciers and permanent snow. In the winter seasons of 1983-2020, there is a decline in snow depth from 50 to 35 cm on average (based on 41 popular second-home areas in the mountains). Results of the fixed effects Poisson estimator with spatial elements show that there is a significant and positive relationship between natural snow depth in the municipality and the number of second homes started. There is also a significant and negative relationship between the number of new second homes in the municipality and a scarcity of snow in the surrounding municipalities. However, the magnitude of both effects is small. Estimates also show a strong positive relationship between the proportion of surface covered by permanent snow or glaciers in the municipality and new second homes. This implies that a decline in permanent snow and glaciers may make these areas less attractive for the location of second homes.