Dataset Information

Consequences of Misspecifying Levels of Variance in Cross-Classified Longitudinal Data Structures.

ABSTRACT: The purpose of this study was to determine if modeling school and classroom effects was necessary in estimating passage reading growth across elementary grades. Longitudinal data from 8367 students in 2989 classrooms in 202 Reading First schools were used in this study and were obtained from the Progress Monitoring and Reporting Network maintained by the Florida Center for Reading Research. Oral reading fluency (ORF) was assessed four times per school year. Five growth models with varying levels of data (student, classroom, and school) were estimated in order to determine which structures were necessary to correctly partition variance and accurately estimate standard errors for growth parameters. Because the results illustrate that not modeling higher-level clustering inflated lower-level variance estimates and in some cases led to biased standard errors, the authors recommend the practice of including classroom cross-classification and school nesting when predicting longitudinal student outcomes.

SUBMITTER: Gilbert J

PROVIDER: S-EPMC4870234 | biostudies-literature | 2016

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Consequences of Misspecifying Levels of Variance in Cross-Classified Longitudinal Data Structures.

Gilbert Jennifer J Petscher Yaacov Y Compton Donald L DL Schatschneider Chris C

Frontiers in psychology 20160518

The purpose of this study was to determine if modeling school and classroom effects was necessary in estimating passage reading growth across elementary grades. Longitudinal data from 8367 students in 2989 classrooms in 202 Reading First schools were used in this study and were obtained from the Progress Monitoring and Reporting Network maintained by the Florida Center for Reading Research. Oral reading fluency (ORF) was assessed four times per school year. Five growth models with varying levels ...[more]

PMID: 27242608

Similar Datasets

Project description:Background:Diabetes is a known independent risk factor for stroke. However, whether higher glucose levels (126-139.9?mg/dl) can increase the risk of stroke in people without diabetes is still unknown. Moreover, as a fluctuating parameter, long-term glucose levels may also be related to the risk of stroke outcome. It is important to explore the correlation between long-term average blood glucose, as well as its variability, and stroke. Methods:We used 40,975 clinical measurements of glucose levels and 367 measurements of glycated hemoglobin A1c levels from 12,321 participants without stroke to examine the relationship between glucose levels and the risk of stroke. Participants were from the Weitang Geriatric Diseases study, including 5,707 men and 6,614 women whose mean age at baseline was 60.8 years; 1,011 participants had diabetes, and 11,310 did not. We estimated the long-term average blood glucose level based on the multilevel Bayesian model and fit in Cox regression models, stratified according to diabetes status. Results:Over a median follow-up period of 5 years, stroke developed in 279 of the 12,321 participants (244 without diabetes and 35 with). For people with an average glucose level of 126-139.9?mg per deciliter, compared with 90-99.9?mg per deciliter, the adjusted hazard ratio (HR) for total stroke was 1.78 (95% confidence interval (CI), 1.16-2.75), and the HR for levels higher than 140?mg per deciliter was 1.89 (95% CI, 1.09-3.29). Among those without diabetes whose glucose level was higher than 140?mg per deciliter, compared with 90-99.9?mg per deciliter, the adjusted HRs for total stroke and fatal stroke were 3.66 (95% CI, 1.47-9.08) and 5 (95% CI, 1.77-14.15), respectively. For a glucose standard deviation level higher than 13.83?mg per deciliter, compared with that lower than 5.91?mg per deciliter, the adjusted HR for total stroke was 2.31 (95% CI, 1.19-4.48). Conclusions:Our results suggest that higher average glucose levels (126-139.9?mg/dl) and variance may be risk factors for stroke, even among people without diabetes diagnosis.

Dataset Information

Consequences of Misspecifying Levels of Variance in Cross-Classified Longitudinal Data Structures.

Publications

Consequences of Misspecifying Levels of Variance in Cross-Classified Longitudinal Data Structures.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets