scholarly journals A power approximation for the Kenward and Roger Wald test in the linear mixed model

PLoS ONE ◽  
2021 ◽  
Vol 16 (7) ◽  
pp. e0254811
Author(s):  
Sarah M. Kreidler ◽  
Brandy M. Ringham ◽  
Keith E. Muller ◽  
Deborah H. Glueck
Keyword(s):  
Mixed Model ◽  
Linear Mixed Model ◽  
Wald Test ◽  
Small Sample ◽  
Power Calculation ◽  
Wald Statistic ◽  
Power Approximation ◽  
Group Randomized Trial ◽  
Approximate Moments ◽  
Image Position

We derive a noncentral F power approximation for the Kenward and Roger test. We use a method of moments approach to form an approximate distribution for the Kenward and Roger scaled Wald statistic, under the alternative. The result depends on the approximate moments of the unscaled Wald statistic. Via Monte Carlo simulation, we demonstrate that the new power approximation is accurate for cluster randomized trials and longitudinal study designs. The method retains accuracy for small sample sizes, even in the presence of missing data. We illustrate the method with a power calculation for an unbalanced group-randomized trial in oral cancer prevention.

scholarly journals Small-Area Estimation with Zero-Inflated Data – a Simulation Study

2016 ◽  
Vol 32 (4) ◽  
pp. 963-986 ◽  
Author(s):  
Sabine Krieg ◽  
Harm Jan Boonstra ◽  
Marc Smeets
Keyword(s):  
Simulation Study ◽  
Small Area ◽  
Mixed Model ◽  
Linear Mixed Model ◽  
Real Data ◽  
Small Sample ◽  
Artificial Data ◽  
Normal Distributions ◽  
Area Estimation ◽  
Small Sample Sizes

Abstract Many target variables in official statistics follow a semicontinuous distribution with a mixture of zeros and continuously distributed positive values. Such variables are called zero inflated. When reliable estimates for subpopulations with small sample sizes are required, model-based small-area estimators can be used, which improve the accuracy of the estimates by borrowing information from other subpopulations. In this article, three small-area estimators are investigated. The first estimator is the EBLUP, which can be considered the most common small-area estimator and is based on a linear mixed model that assumes normal distributions. Therefore, the EBLUP is model misspecified in the case of zero-inflated variables. The other two small-area estimators are based on a model that takes zero inflation explicitly into account. Both the Bayesian and the frequentist approach are considered. These small-area estimators are compared with each other and with design-based estimation in a simulation study with zero-inflated target variables. Both a simulation with artificial data and a simulation with real data from the Dutch Household Budget Survey are carried out. It is found that the small-area estimators improve the accuracy compared to the design-based estimator. The amount of improvement strongly depends on the properties of the population and the subpopulations of interest.

scholarly journals Solutions for Determining the Significance Region Using the Johnson-Neyman Type Procedure in Generalized Linear (Mixed) Models

2011 ◽  
Vol 36 (6) ◽  
pp. 699-719 ◽  
Author(s):  
Ann A. Lazar ◽  
Gary O. Zerbe
Keyword(s):  
Error Rate ◽  
Multiple Testing ◽  
Mixed Model ◽  
Linear Mixed Model ◽  
Type I Error ◽  
Hierarchical Linear Model ◽  
Wald Test ◽  
Analysis Of Covariance ◽  
Type I ◽  
Type I Error Rate

Researchers often compare the relationship between an outcome and covariate for two or more groups by evaluating whether the fitted regression curves differ significantly. When they do, researchers need to determine the “significance region,” or the values of the covariate where the curves significantly differ. In analysis of covariance (ANCOVA), the Johnson-Neyman procedure can be used to determine the significance region; for the hierarchical linear model (HLM), the Miyazaki and Maier (M-M) procedure has been suggested. However, neither procedure can assume nonnormally distributed data. Furthermore, the M-M procedure produces biased (downward) results because it uses the Wald test, does not control the inflated Type I error rate due to multiple testing, and requires implementing multiple software packages to determine the significance region. In this article, we address these limitations by proposing solutions for determining the significance region suitable for generalized linear (mixed) model (GLM or GLMM). These proposed solutions incorporate test statistics that resolve the biased results, control the Type I error rate using Scheffé’s method, and uses a single statistical software package to determine the significance region.

scholarly journals Power Analysis for Stepped Wedge Trials with Two Treatments

2020 ◽  
Author(s):  
Phillip T Sundin ◽  
Catherine M Crespi
Keyword(s):  
Statistical Power ◽  
Mixed Model ◽  
Linear Mixed Model ◽  
Random Effect ◽  
Power Calculation ◽  
Stepped Wedge ◽  
Cross Sectional ◽  
Closed Form Solutions ◽  
Multiple Treatments ◽  
Cluster Randomized

AbstractStepped wedge designs (SWDs) are designs for cluster randomized trials that feature staggered, unidirectional cross-over, typically from a control to a treatment condition. Existing literature on statistical power for SWDs primarily focuses on designs with a single treatment. However, SWDs with multiple treatments are being proposed and conducted. We present a linear mixed model for a SWD with two treatments, with and without an interaction between them. We derive closed form solutions for the standard errors of the treatment effect coefficients for such models along with power calculation methods. We consider repeated cross-sectional designs as well as open and closed cohort designs and different random effect structures. Design features are examined to determine their impact on power for main treatment and interaction effects.

Adolescent Patient Text Messaging Through Peer Networks to Increase Youth Access Healthcare and STI Testing (Preprint)

2021 ◽  
Author(s):  
Marguerita Lightfoot ◽  
Joi Jackson-Morgan ◽  
Lance Pollack ◽  
Ayanna Bennett
Keyword(s):  
Text Messaging ◽  
Mixed Model ◽  
Linear Mixed Model ◽  
Small Sample Size ◽  
Healthcare Services ◽  
Small Sample ◽  
Text Messages ◽  
Adolescent Medicine ◽  
Post Intervention ◽  
Sti Testing

UNSTRUCTURED Adolescents are disproportionately affected by sexually transmitted infections (STI), including HIV. Many youth with asymptomatic STI or related symptoms do not seek treatment and may not be screened if accessing the healthcare system for other reasons. We examined the feasibility and acceptability of a peer-driven, text messaging strategy to connect youth to STI and HIV services. Using an interrupted time series design consecutive patients at an adolescent medicine clinic were enlisted to send five text messages to peers they believed were sexually active and lived in the clinic's service area. Analyses conducted in 2015 used a generalized linear mixed model. Data were collected from 2013-2014. Of 153 patients approached to participate, 100 agreed to send text messages. Most (55%) reported no concerns with sending the text message. No adverse events or negative outcomes were reported. Adolescent STI testing, positive test results and reported risk behavior increased post intervention, although not statistically significant, likely because of the small sample size. Given low youth uptake of healthcare services and STI/HIV screening, in particular, new strategies are needed to address access barriers. Common approaches for reaching youth are resource-intensive and often miss youth not connected to school or community programs. The peer-based, text messaging strategy showed promise for both increasing the number of youth accessing health services and finding youth engaging in sexual risk behaviors and most in need of sexual health screening and services.

Analyzing discontinuities in longitudinal count data: A multilevel generalized linear mixed model.

2020 ◽  
Author(s):  
James L. Peugh ◽  
Sarah J. Beal ◽  
Meghan E. McGrady ◽  
Michael D. Toland ◽  
Constance Mara
Keyword(s):  
Count Data ◽  
Mixed Model ◽  
Linear Mixed Model ◽  
Generalized Linear Mixed Model ◽  
Longitudinal Count Data

Predicting bycatch hotspots based on suitable habitat derived from fishery-independent data

2020 ◽  
Vol 641 ◽  
pp. 159-175
Author(s):  
J Runnebaum ◽  
KR Tanaka ◽  
L Guan ◽  
J Cao ◽  
L O’Brien ◽  
...  
Keyword(s):  
Habitat Suitability ◽  
Mixed Model ◽  
Linear Mixed Model ◽  
Gulf Of Maine ◽  
Homarus Americanus ◽  
Management Tool ◽  
American Lobster ◽  
Suitable Habitat ◽  
Climate Conditions ◽  
Suitability Index

Bycatch remains a global problem in managing sustainable fisheries. A critical aspect of management is understanding the timing and spatial extent of bycatch. Fisheries management often relies on observed bycatch data, which are not always available due to a lack of reporting or observer coverage. Alternatively, analyzing the overlap in suitable habitat for the target and non-target species can provide a spatial management tool to understand where bycatch interactions are likely to occur. Potential bycatch hotspots based on suitable habitat were predicted for cusk Brosme brosme incidentally caught in the Gulf of Maine American lobster Homarus americanus fishery. Data from multiple fisheries-independent surveys were combined in a delta-generalized linear mixed model to generate spatially explicit density estimates for use in an independent habitat suitability index. The habitat suitability indices for American lobster and cusk were then compared to predict potential bycatch hotspot locations. Suitable habitat for American lobster has increased between 1980 and 2013 while suitable habitat for cusk decreased throughout most of the Gulf of Maine, except for Georges Basin and the Great South Channel. The proportion of overlap in suitable habitat varied interannually but decreased slightly in the spring and remained relatively stable in the fall over the time series. As Gulf of Maine temperatures continue to increase, the interactions between American lobster and cusk are predicted to decline as cusk habitat continues to constrict. This framework can contribute to fisheries managers’ understanding of changes in habitat overlap as climate conditions continue to change and alter where bycatch interactions could occur.

Long-term seizure outcomes after pediatric temporal lobectomy: does brain MRI lesion matter?

2019 ◽  
Vol 24 (2) ◽  
pp. 200-208
Author(s):  
Ravindra Arya ◽  
Francesco T. Mangano ◽  
Paul S. Horn ◽  
Sabrina K. Kaul ◽  
Serena K. Kaul ◽  
...  
Keyword(s):  
Repeated Measures ◽  
Mixed Model ◽  
Linear Mixed Model ◽  
Hippocampal Sclerosis ◽  
Brain Mri ◽  
Temporal Lobectomy ◽  
Seizure Freedom ◽  
Seizure Onset

OBJECTIVEThere is emerging data that adults with temporal lobe epilepsy (TLE) without a discrete lesion on brain MRI have surgical outcomes comparable to those with hippocampal sclerosis (HS). However, pediatric TLE is different from its adult counterpart. In this study, the authors investigated if the presence of a potentially epileptogenic lesion on presurgical brain MRI influences the long-term seizure outcomes after pediatric temporal lobectomy.METHODSChildren who underwent temporal lobectomy between 2007 and 2015 and had at least 1 year of seizure outcomes data were identified. These were classified into lesional and MRI-negative groups based on whether an epilepsy-protocol brain MRI showed a lesion sufficiently specific to guide surgical decisions. These patients were also categorized into pure TLE and temporal plus epilepsies based on the neurophysiological localization of the seizure-onset zone. Seizure outcomes at each follow-up visit were incorporated into a repeated-measures generalized linear mixed model (GLMM) with MRI status as a grouping variable. Clinical variables were incorporated into GLMM as covariates.RESULTSOne hundred nine patients (44 females) were included, aged 5 to 21 years, and were classified as lesional (73%), MRI negative (27%), pure TLE (56%), and temporal plus (44%). After a mean follow-up of 3.2 years (range 1.2–8.8 years), 66% of the patients were seizure free for ≥ 1 year at last follow-up. GLMM analysis revealed that lesional patients were more likely to be seizure free over the long term compared to MRI-negative patients for the overall cohort (OR 2.58, p < 0.0001) and for temporal plus epilepsies (OR 1.85, p = 0.0052). The effect of MRI lesion was not significant for pure TLE (OR 2.64, p = 0.0635). Concordance of ictal electroencephalography (OR 3.46, p < 0.0001), magnetoencephalography (OR 4.26, p < 0.0001), and later age of seizure onset (OR 1.05, p = 0.0091) were associated with a higher likelihood of seizure freedom. The most common histological findings included cortical dysplasia types 1B and 2A, HS (40% with dual pathology), and tuberous sclerosis.CONCLUSIONSA lesion on presurgical brain MRI is an important determinant of long-term seizure freedom after pediatric temporal lobectomy. Pediatric TLE is heterogeneous regarding etiologies and organization of seizure-onset zones with many patients qualifying for temporal plus nosology. The presence of an MRI lesion determined seizure outcomes in patients with temporal plus epilepsies. However, pure TLE had comparable surgical seizure outcomes for lesional and MRI-negative groups.

Sign in / Sign up

Export Citation Format

Share Document