Time Series Data Analysis of Single Subject Experimental Designs Using Bayesian Estimation (open access)

Time Series Data Analysis of Single Subject Experimental Designs Using Bayesian Estimation

This study presents a set of data analysis approaches for single subject designs (SSDs). The primary purpose is to establish a series of statistical models to supplement visual analysis in single subject research using Bayesian estimation. Linear modeling approach has been used to study level and trend changes. I propose an alternate approach that treats the phase change-point between the baseline and intervention conditions as an unknown parameter. Similar to some existing approaches, the models take into account changes in slopes and intercepts in the presence of serial dependency. The Bayesian procedure used to estimate the parameters and analyze the data is described. Researchers use a variety of statistical analysis methods to analyze different single subject research designs. This dissertation presents a series of statistical models to model data from various conditions: the baseline phase, A-B design, A-B-A-B design, multiple baseline design, alternating treatments design, and changing criterion design. The change-point evaluation method can provide additional confirmation of causal effect of the treatment on target behavior. Software codes are provided as supplemental materials in the appendices. The applicability for the analyses is demonstrated using five examples from the SSD literature.
Date: August 2015
Creator: Aerts, Xing Qin
System: The UNT Digital Library
Comparing Three Effect Sizes for Latent Class Analysis (open access)

Comparing Three Effect Sizes for Latent Class Analysis

Traditional latent class analysis (LCA) considers entropy R2 as the only measure of effect size. However, entropy may not always be reliable, a low boundary is not agreed upon, and good separation is limited to values of greater than .80. As applications of LCA grow in popularity, it is imperative to use additional sources to quantify LCA classification accuracy. Greater classification accuracy helps to ensure that the profile of the latent classes reflect the profile of the true underlying subgroups. This Monte Carlo study compared the quantification of classification accuracy and confidence intervals of three effect sizes, entropy R2, I-index, and Cohen’s d. Study conditions included total sample size, number of dichotomous indicators, latent class membership probabilities (γ), conditional item-response probabilities (ρ), variance ratio, sample size ratio, and distribution types for a 2-class model. Overall, entropy R2 and I-index showed the best accuracy and standard error, along with the smallest confidence interval widths. Results showed that I-index only performed well for a few cases.
Date: December 2015
Creator: Granado, Elvalicia A.
System: The UNT Digital Library
Reliability Generalization: a Systematic Review and Evaluation of Meta-analytic Methodology and Reporting Practice (open access)

Reliability Generalization: a Systematic Review and Evaluation of Meta-analytic Methodology and Reporting Practice

Reliability generalization (RG) is a method for meta-analysis of reliability coefficients to estimate average score reliability across studies, determine variation in reliability, and identify study-level moderator variables influencing score reliability. A total of 107 peer-reviewed RG studies published from 1998 to 2013 were systematically reviewed to characterize the meta-analytic methods employed and to evaluate quality of reporting practice against standards for transparency in meta-analysis reporting. Most commonly, RG studies meta-analyzed alpha coefficients, which were synthesized using an unweighted, fixed-effects model applied to untransformed coefficients. Moderator analyses most frequently included multiple regression and bivariate correlations employing a fixed-effects model on untransformed, unweighted coefficients. Based on a unit-weighted scoring system, mean reporting quality for RG studies was statistically less than that for a comparison study of 198 meta-analyses in the organizational sciences across 42 indicators; however, means were not statistically significantly different between the two studies when evaluating reporting quality on 18 indicators deemed essential to ethical reporting practice in meta-analyses. Since its inception a wide variety of statistical methods have been applied to RG, and meta-analysis of reliability coefficients has extended to fields outside of psychological measurement, such as medicine and business. A set of guidelines for conducting and reporting RG …
Date: December 2015
Creator: Holland, David F.
System: The UNT Digital Library