scholarly journals Application of Multilevel Models to Morphometric Data. Part 2. Correlations

2003 ◽  
Vol 25 (4) ◽  
pp. 187-191 ◽  
Author(s):  
O. Tsybrovskyy ◽  
A. Berghold
Keyword(s):  
Multilevel Modeling ◽  
Multilevel Models ◽  
Correlation Coefficients ◽  
Morphometric Data ◽  
Cell Level ◽  
Units Of Analysis ◽  
Major Interest ◽  
Level Statistics ◽  
Indispensable Tool ◽  
Multivariate Multilevel Models

Multilevel organization of morphometric data (cells are “nested” within patients) requires special methods for studying correlations between karyometric features. The most distinct feature of these methods is that separate correlation (covariance) matrices are produced for every level in the hierarchy. In karyometric research, the cell‐level (i.e., within‐tumor) correlations seem to be of major interest. Beside their biological importance, these correlation coefficients (CC) are compulsory when dimensionality reduction is required. Using MLwiN, a dedicated program for multilevel modeling, we show how to use multivariate multilevel models (MMM) to obtain and interpret CC in each of the levels. A comparison with two usual, “single‐level” statistics shows that MMM represent the only way to obtain correct cell‐level correlation coefficients. The summary statistics method (take average values across each patient) produces patient‐level CC only, and the “pooling” method (merge all cells together and ignore patients as units of analysis) yields incorrect CC at all. We conclude that multilevel modeling is an indispensable tool for studying correlations between morphometric variables.

scholarly journals Application of Multilevel Models to Morphometric Data. Part 1. Linear Models and Hypothesis Testing

2003 ◽  
Vol 25 (4) ◽  
pp. 173-185 ◽  
Author(s):  
O. Tsybrovskyy ◽  
A. Berghold
Keyword(s):  
Multilevel Modeling ◽  
Multilevel Models ◽  
Linear Models ◽  
Nuclear Morphology ◽  
Hierarchical Data ◽  
Morphometric Data ◽  
Statistical Software ◽  
Level Statistics ◽  
Thyroid Adenomas ◽  
Theoretical Issues

Morphometric data usually have a hierarchical structure (i.e., cells are nested within patients), which should be taken into consideration in the analysis. In the recent years, special methods of handling hierarchical data, called multilevel models (MM), as well as corresponding software have received considerable development. However, there has been no application of these methods to morphometric data yet. In this paper we report our first experience of analyzing karyometric data by means of MLwiN – a dedicated program for multilevel modeling. Our data were obtained from 34 follicular adenomas and 44 follicular carcinomas of the thyroid. We show examples of fitting and interpreting MM of different complexity, and draw a number of interesting conclusions about the differences in nuclear morphology between follicular thyroid adenomas and carcinomas. We also demonstrate substantial advantages of multilevel models over conventional, single‐level statistics, which have been adopted previously to analyze karyometric data. In addition, some theoretical issues related to MM as well as major statistical software for MM are briefly reviewed.

scholarly journals 87. Impact of State of Residence on Adult Vaccination Uptake: A Multilevel Modeling Approach

2020 ◽  
Vol 7 (Supplement_1) ◽  
pp. S175-S175
Author(s):  
Shannon Hunter ◽  
Diana Garbinsky ◽  
Elizabeth M La ◽  
Sara Poston ◽  
Cosmina Hogea
Keyword(s):  
Multilevel Modeling ◽  
Vaccination Coverage ◽  
Multilevel Models ◽  
State Level ◽  
Childhood Vaccination ◽  
Material Support ◽  
Adult Vaccination ◽  
Direct Compensation ◽  
Cluster Variation ◽  
Coverage Rates

Abstract Background Previous studies on adult vaccination coverage found inter-state variability that persists after adjusting for individual demographic factors. Assessing the impact of state-level factors may help improve uptake strategies. This study aimed to: • Update previous estimates of state-level, model-adjusted coverage rates for influenza; pneumococcal; tetanus, diphtheria, and acellular pertussis (Tdap); and herpes zoster (HZ) vaccines (individually and in compliance with all age-appropriate recommended vaccinations) • Evaluate effects of individual and state-level factors on adult vaccination coverage using a multilevel modeling framework. Methods Behavioral Risk Factor Surveillance System (BRFSS) survey data (2015–2017) were retrospectively analyzed. Multivariable logistic regression models estimated state vaccination coverage and compliance using predicted marginal proportions. BRFSS data were then combined with external state-level data to estimate multilevel models evaluating effects of state-level factors on coverage. Weighted odds ratios and measures of cluster variation were estimated. Results Adult vaccination coverage and compliance varied by state, even after adjusting for individual characteristics, with coverage ranging as follows: • Influenza (2017): 35.1–48.1% • Pneumococcal (2017): 68.2–80.8% • Tdap (2016): 21.9–46.5% • HZ (2017): 30.5–50.9% Few state-level variables were retained in final multilevel models, and measures of cluster variation suggested substantial residual variation unexplained by individual and state-level variables. Key state-level variables positively associated with vaccination included health insurance coverage rates (influenza/HZ), pharmacists’ vaccination authority (HZ), presence of childhood vaccination exemptions (pneumococcal/Tdap), and adult immunization information system participation (Tdap/HZ). Conclusion Adult vaccination coverage and compliance continue to show substantial variation by state even after adjusting for individual and state-level characteristics associated with vaccination. Further research is needed to assess additional state or local factors impacting vaccination disparities. Funding GlaxoSmithKline Biologicals SA (study identifier: HO-18-19794) Disclosures Shannon Hunter, MS, GSK (Other Financial or Material Support, Ms. Hunter is an employee of RTI Health Solutions, who received consultancy fees from GSK for conduct of the study. Ms. Hunter received no direct compensation from the Sponsor.) Diana Garbinsky, MS, GSK (Other Financial or Material Support, The study was conducted by RTI Health Solutions, which received consultancy fees from GSK. I am a salaried employee at RTI Health Solutions and received no direct compensation from GSK for the conduct of this study..) Elizabeth M. La, PhD, RTI Health Solutions (Employee) Sara Poston, PharmD, The GlaxoSmithKline group of companies (Employee, Shareholder) Cosmina Hogea, PhD, GlaxoSmithKline (Employee, Shareholder)

From Event Reports to Protest Analysis

2019 ◽  
pp. 62-76
Author(s):  
Nils B. Weidmann ◽  
Espen Geelmuyden Rød
Keyword(s):  
Research Design ◽  
Multilevel Models ◽  
Mass Mobilization ◽  
Units Of Analysis ◽  
Computer Scientists ◽  
The Hierarchical Structure ◽  
National Unit ◽  
Internet Penetration ◽  
Mass Protest ◽  
High Level

This chapter introduces the main elements of the research design for the empirical chapters in the book. Starting with the event reports provided by the Mass Mobilization in Autocracies Database, the chapter develops a research design that studies variation in local Internet penetration and anti-regime protest. The chapter motivates the choice of the sub-national unit of observation (cities), and temporal units of analysis (years, weeks). It introduces a new measure of Internet penetration derived from network measurements, developed in collaboration with computer scientists. The high level of spatial and temporal resolution allows for one of the most detailed analyses so far in the study of mass protest. The chapter also introduces the statistical models used for the analysis. The book relies on Bayesian multilevel models, a framework that takes into account the hierarchical structure of the data and has advantages in the analysis of data with skewed dependent variables.

How Does Fidelity of Implementation Matter? Using Multilevel Models to Detect Relationships Between Participant Outcomes and the Delivery and Receipt of Treatment

2012 ◽  
Vol 33 (4) ◽  
pp. 547-565 ◽  
Author(s):  
Keith Zvoch
Keyword(s):  
Program Evaluation ◽  
Multilevel Modeling ◽  
Multilevel Models ◽  
Growth Models ◽  
Fidelity Of Implementation ◽  
Literacy Intervention ◽  
Multidimensional Construct ◽  
Modeling Techniques ◽  
Model Treatment ◽  
Participant Outcomes

Multilevel modeling techniques facilitated examination of relationships between fidelity indicators and outcomes associated with a summer literacy intervention. Three-level growth models were specified to capture the extent to which students experienced instruction and to demonstrate the ways in which dosage–response relationships manifest in program evaluation contexts. The observation that outcome-related deviations from program protocol occurred both at the provider and at the recipient levels suggests that evaluators will often need to conceptualize, measure, and model “treatment fidelity” as a multilevel, multidimensional construct.

scholarly journals Mind the step: A more insightful and robust analysis of the sentencing process in England and Wales under the new sentencing guidelines

2018 ◽  
Vol 20 (3) ◽  
pp. 268-301 ◽  
Author(s):  
Jose Pina-Sánchez ◽  
Ian Brunton-Smith ◽  
Guangquan Li
Keyword(s):  
Multilevel Models ◽  
Sentencing Guidelines ◽  
Robust Analysis ◽  
England And Wales ◽  
One Stage ◽  
Relevant Case ◽  
Case Characteristics ◽  
Multivariate Multilevel Models

The ‘England and Wales Sentencing Guidelines’ aim to promote consistency by organizing the sentencing process as a sequence of steps, with initial judicial assessments subsequently adjusted to reflect relevant case characteristics. Yet, existing evaluations of the guidelines have failed to incorporate this structure adequately, instead concentrating solely on sentence outcomes. We use multivariate multilevel models to offer new insights into the decisions made throughout the sentencing process. Focusing on cases of assault sentenced at the Crown Court we show that the level of compliance with the guidelines is high. However, we also show that some case characteristics are being unduly considered at more than one stage of the sentencing process, meaning existing studies may be underestimating their true influence.

scholarly journals Changes in General Practitioners’ consultation frequency over time for patients with hypertension or anxiety/depression symptoms: a 10-year follow-up of the Norwegian HUNT study

2019 ◽  
Author(s):  
Lena J Skarshaug ◽  
Silje L Kaspersen ◽  
Johan H Bjørngaard ◽  
Kristine Pape
Keyword(s):  
General Practitioners ◽  
Multilevel Models ◽  
Correlation Coefficients ◽  
Depression Symptoms ◽  
Complex Needs ◽  
Baseline Status ◽  
Clinical Measurements ◽  
The Individual ◽  
Anxiety Depression

Abstract Background General Practitioners’ (GPs’) workload has been suggested to increase in many countries; how does this impact patient follow-up? Objective To investigate trends in GP consultation patterns for adults according to baseline hypertension and anxiety/depression symptoms and attribution of the GP to trend differences. Methods Prospective cohort study, linking survey data and clinical measurements from the Norwegian HUNT3 study (2006–08) with national administrative data on GP list assignment and consultations with GP services. We grouped participants aged 40–59 years according to sex and their baseline status regarding hypertension and anxiety/depression symptoms. We registered GP consultations in 2007–16 and used general estimation equation models to estimate the level of GP consultations per month per year during follow-up. We used multilevel models with participants nested in their assigned regular GP to calculate GP-level intra-class correlation coefficients, reflecting to what extent patients’ consultation patterns could be attributed to the individual GP. Results In total, 47 550 HUNT3 participants were registered with 102 different GPs in Nord-Trøndelag County, Norway, in 2007. Adjusted for age, we observed an overall increase in GP consultations in 2007–16, particularly in those with a better health status at baseline. About 2% of the variance of patient consultations could be attributed to differences between GPs and 10% to the use of lengthy consultations. Out-of-hours consultations did not change much in the study period 2007–16. Conclusion Increased use of GP consultations, mainly among the healthiest participants, encourage further research into whether these patients displace patients with heavier and more complex needs.

The Role of Sex and Age on Pre-drinking: An Exploratory International Comparison of 27 Countries

2019 ◽  
Vol 54 (4) ◽  
pp. 378-385 ◽  
Author(s):  
Jason Ferris ◽  
Cheneal Puljević ◽  
Florian Labhart ◽  
Adam Winstock ◽  
Emmanuel Kuntsche
Keyword(s):  
Multilevel Models ◽  
The United States ◽  
Drinking Behaviour ◽  
Single Model ◽  
Policy Makers ◽  
Worldwide Phenomenon ◽  
Using Data ◽  
The Impact ◽  
Multivariate Multilevel Models

Abstract Aims This exploratory study aims to model the impact of sex and age on the percentage of pre-drinking in 27 countries, presenting a single model of pre-drinking behaviour for all countries and then comparing the role of sex and age on pre-drinking behaviour between countries. Methods Using data from the Global Drug Survey, the percentages of pre-drinkers were estimated for 27 countries from 64,485 respondents. Bivariate and multivariate multilevel models were used to investigate and compare the percentage of pre-drinking by sex (male and female) and age (16–35 years) between countries. Results The estimated percentage of pre-drinkers per country ranged from 17.8% (Greece) to 85.6% (Ireland). The influence of sex and age on pre-drinking showed large variation between the 27 countries. With the exception of Canada and Denmark, higher percentages of males engaged in pre-drinking compared to females, at all ages. While we noted a decline in pre-drinking probability among respondents in all countries after 21 years of age, after the age of 30 this probability remained constant in some countries, or even increased in Brazil, Canada, England, Ireland, New Zealand and the United States. Conclusions Pre-drinking is a worldwide phenomenon, but varies substantially by sex and age between countries. These variations suggest that policy-makers would benefit from increased understanding of the particularities of pre-drinking in their own country to efficiently target harmful pre-drinking behaviours.

Sample Size and Accuracy of Estimates in Multilevel Models

Methodology ◽  
2011 ◽  
Vol 7 (3) ◽  
pp. 111-120 ◽  
Author(s):  
Omar Paccagnella
Keyword(s):  
Variance Components ◽  
Multilevel Models ◽  
Intraclass Correlation ◽  
Correlation Coefficients ◽  
Estimation Procedure ◽  
Small Sample ◽  
Standard Errors ◽  
Sample Sizes ◽  
Monte Carlo Simulation Study ◽  
Small Sample Sizes

In a multilevel framework several researches have investigated the behavior of estimates in finite samples, particularly for continuous dependent variables. Some findings show poor precise estimates for the variance components. On the other hand, discrete response multilevel models have been investigated less widely. In this paper we analyze the influence of different factors on the accuracy of estimates and standard errors of estimates in a binary response 2-level model, through a Monte Carlo simulation study. We investigate the hypothesis of: (a) small sample sizes; (b) different intraclass correlation coefficients; (c) different numbers of quadrature points in the estimation procedure. Standard errors of estimates are studied through a noncoverage indicator. In all instances we have considered, the point estimates are unbiased (even with very small sample sizes), while the variance components are underestimated. The accuracy of the standard errors of variance estimates needs a very large number of groups.

Sign in / Sign up

Export Citation Format

Share Document