A Bayesian hierarchical model for the analysis of visual analogue scaling tasks.

In psychophysics and psychometrics, an integral method to the discipline involves charting how a person's response pattern changes according to a continuum of stimuli. For instance, in hearing science, Visual Analog Scaling tasks are experiments in which listeners hear sounds across a speech continuum and give a numeric rating between 0 and 100 conveying whether the sound they heard was more like word "a" or more like word "b" (i.e. each participant is giving a continuous categorization response). By taking all the continuous categorization responses across the speech continuum, a parametric curve model can be fit to the data and used to analyze any individual's response pattern by speech continuum. Standard statistical modeling techniques are not able to accommodate all of the specific requirements needed to analyze these data. Thus, Bayesian hierarchical modeling techniques are employed to accommodate group-level non-linear curves, individual-specific non-linear curves, continuum-level random effects, and a subject-specific variance that is predicted by other model parameters. In this paper, a Bayesian hierarchical model is constructed to model the data from a Visual Analog Scaling task study of mono-lingual and bi-lingual participants. Any nonlinear curve function could be used and we demonstrate the technique using the 4-parameter logistic function. Overall, the model was found to fit particularly well to the data from the study and results suggested that the magnitude of the slope was what most defined the differences in response patterns between continua.

The Dental Researcher's Guide to Data Analysis - Part 2.

It is important for dental researchers to have a general idea of the different types of data that can be collected from a study and the available statistical tools that can be used with such data. By knowing about what is available, researchers will have an informed idea of the types of studies that should be conducted, the data that should be collected, and the proper statistical methods for analyzing collected data. In the first part of this general overview of statistical methods, we walked through the various types of data that traditional statistical techniques, such as t tests and linear regression, can handle. In this second part, we explore more complex types of data that traditional statistical techniques are unable to handle. Specifically, we discuss longitudinal and time-to-event data because both occur frequently in dental studies and require special modeling techniques in order to analyze correctly. Using two different simulated dental datasets, the proper application of techniques such as repeated measures ANOVA, linear mixed modeling, generalized estimating equations, log-rank test, and Cox proportional hazards models are discussed and illustrated in depth.

The Dental Researcher's Guide to Data Analysis: Part 1.

One of the most important parts of any research project is the statistical analysis through which results are described and deemed significant or otherwise. Given this role of statistical analyses, it is imperative to understand what statistical tests are available and, most importantly, when they are appropriate. This is because choosing an improper statistical test can lead to inaccurate results and meaningless conclusions. We have written this data analysis guide to assist dental researchers with carrying out a proper analysis plan in consultation with a statistician. In this first part, we detail common types of outcome and predictor variables and include some common tests that can be used for each scenario. Using different simulated datasets, we also provide examples of common data visualization techniques and how a proper statistical test can be applied. We hope this guide provides some general background knowledge on common statistical techniques and allows for easier and more effective conversations with your statistician.

ComBat Harmonization of Myocardial Radiomic Features Sensitive to Cardiac MRI Acquisition Parameters.

Purpose: To investigate the effect of ComBat harmonization methods on the robustness of cardiac MRI-derived radiomic features to variations in imaging parameters. Materials and Methods: This Health Insurance Portability and Accountability Act-compliant retrospective study used a publicly available data set of 11 healthy controls (mean age, 33 years ± 16 [SD]; six men) and five patients (mean age, 52 years ± 16; four men). A single midventricular short-axis section was acquired with 3-T MRI using cine balanced steady-state free precision, T1-weighted, T2-weighted, T1 mapping, and T2 mapping imaging sequences. Each sequence was acquired using baseline parameters and after variations in flip angle, spatial resolution, section thickness, and parallel imaging. Image registration was performed for all sequences at a per-individual level. Manual myocardial contouring was performed, and 1652 radiomic features per sequence were extracted using baseline and variations in imaging parameters. Radiomic feature stability to change in imaging parameters was assessed using Cohen d sensitivity. The stability of radiomic features was assessed both without and after ComBat harmonization of radiomic features. Three ComBat methods were studied: parametric, nonparametric, and Gaussian mixture model (GMM). Results: For all sequences combined, 51.4% of features were robust to changes in imaging parameters when no ComBat method was applied. ComBat harmonization substantially increased the number of stable features to 95.1% (95% CI: 94.9, 95.3) when parametric ComBat was used and 90.9% (95% CI: 90.6, 91.2) when nonparametric ComBat was used. GMM combat resulted in only 52.6% stable features. Conclusion: ComBat harmonization improved the stability of radiomic features to changes in imaging parameters across all cardiac MRI sequences.Keywords: Cardiac MRI, Radiomics, ComBat, Harmonization Supplemental material is available for this article. © RSNA, 2023.

Story Generation and Narrative Retells in Children Who Are Hard of Hearing and Hearing Children.

PURPOSE: The primary goal for this study was to compare story generation and narrative retell performance in children who are hard of hearing (CHH) and hearing children. A secondary goal was to determine factors that influence narrative performance. Research on this topic is important because narrative language competency is an essential communication function. METHOD: Participants included 86 CHH and 53 seven-year-old hearing children who had completed a test battery composed of vocabulary, grammar, phonological processing, story generation, and narrative retell tasks. Coders who were blind to hearing status used a scoring rubric to judge the quality of narrative macrostructure in story generation and narrative retells. Data were analyzed using analysis of variance to determine group differences and correlational models to examine individual differences. RESULTS: At 7 years of age, CHH demonstrated significant deficits in narrative macrostructure compared to hearing children, with larger delays in narrative retell than story generation. Vocabulary, grammar, and phonological memory acted as mediators in the relationship between hearing status and story generation; grammar acted as a mediator between hearing status and narrative retell. Auditory access variables accounted for a significant proportion of shared variance in story generation skills for CHH. CONCLUSIONS: School-age CHH are at risk for delays in narrative production, particularly with retelling stories. The results of this study highlight a narrative coding approach and task procedures that are sensitive to differences in language levels and may be clinically useful for professionals working with early school-age children.

Correlation in Dental Studies: Why It Matters and What to Do About It.

In dental research, it is particularly common for studies to collect data that are fundamentally correlated. Some common dental situations in which correlation arises include patients being observed across multiple teeth and/or across multiple time points, such as before and after treatment, or groups of patients being clustered (ie, familial units). For a number of traditional statistical tests and modeling techniques, the assumption of independence between observations is imperative in order to receive valid results and make accurate conclusions. This article describes how ignoring inherent correlations in data can lead to erroneous results when using traditional methods as well as the types of modeling techniques that are available to handle correlated data. Furthermore, two simulation studies are performed to further illustrate and prove the advantages of adequately handling correlated data in statistical analyses. Int J Oral Maxillofac Implants 2023;38:417-421. doi: 10.11607/jomi.10285.

Bayesian Statistics: A Walkthrough with a Simulated Dental Dataset.

When a clinician sees a patient with a complication, they often go through a Bayesian style of logic, most likely without even knowing it. They assess whether they have seen the complication before, provide an intervention based on historical knowledge of what leads to improvement, and then later assess how the intervention is performed. This process, which is routine in clinical practice, can be mathematically extended into an alternative way of performing statistical analyses to assess clinical research. However, this process is contrary to the most common statistical methods used in dental research: frequentist statistics. Though powerful, frequentist methods come with advantages and disadvantages. Bayesian statistics are an alternative method, one that mirrors how we as researchers think and process new information. In this primer, a walkthrough of Bayesian statistics is performed by constructing priors, defining the likelihood, and using the posterior result to draw conclusions on parameters of interest. The motivating example for this walkthrough was a Bayesian analog to logistic regression, fit using a simulated dental-related dataset of 50 patients who received a dental implant-classified as either within or outside normal limits-from practitioners who did or did not receive a training course in implant placement. The results of the Bayesian and traditional frequentist logistic regression models were compared, resulting in very similar conclusions regarding which parameters seemed to be strongly associated with the outcome.

Increased rate of listening difficulties in autistic children.

INTRODUCTION: Auditory challenges are both common and disruptive for autistic children and evidence suggests that listening difficulties may be linked to academic underachievement (Ashburner, Ziviani & Rodger, 2008). Such deficits may also contribute to issues with attention, behavior, and communication (Ashburner et al., 2008; Riccio, Cohen, Garrison & Smith, 2005). The present study aims to summarize the auditory challenges of autistic children with normal pure-tone hearing thresholds, and perceived listening difficulties, seen at auditory-ASD clinics in the US and Australia. METHODS: Data were compiled on a comprehensive, auditory-focused test battery in a large clinical sample of school-age autistic children with normal pure-tone hearing to date (N = 71, 6-14 years). Measures included a parent-reported auditory sensory processing questionnaire and tests of speech recognition in noise, binaural integration, attention, auditory memory and listening comprehension. Individual test performance was compared to normative data from children with no listening difficulties. RESULTS: Over 40% of patients exhibited significantly reduced speech recognition in noise and abnormal dichotic integration that were not attributed to deficits in attention. The majority of patients (86%) performed abnormally on at least one auditory measure, suggesting that functional auditory issues can exist in autistic patients despite normal pure-tone sensitivity. CONCLUSION: Including functional listening measures during audiological evaluations may improve clinicians' ability to detect and manage the auditory challenges impacting this population. Learner Outcomes: 1) Readers will be able to describe the auditory difficulties experienced by some autistic patients (ASD). 2) Readers will be able to describe clinical measures potentially useful for detecting listening difficulties in high-functioning autistic children.