A holistic view on a targeted intervention on exclusionary discipline using generalized additive models for location, scale, and shape
No Thumbnail Available
Authors
Meeting name
Sponsors
Date
Journal Title
Format
Thesis
Subject
Abstract
The use of Generalized Linear Models (GLMs) to analyze count data has become a common practice in social work research. However, most applications of GLMs to count data fail to test for various distributions, which can result in inaccurate estimates. An alternative approach that provides more flexibility is the Generalized Additive Models for Location, Scale, and Shape (GAMLSS). GAMLSS is an innovative statistical approach that tests over 30 different count distributions by comparing each model and selects the best fitted models for the data. Despite its advantages, a preliminary search of the social work literature yielded no published papers utilizing this approach. This raises questions about the accuracy of current analyses of models using count data. The purpose of this dissertation is to demonstrate the use of GAMLSS in analyzing the effectiveness of the Self-Monitoring and Regulation Training Strategy (SMARTS), a school-based behavioral intervention, on in-school suspension (ISS), out-of-school suspension (OSS), and exclusionary discipline (ED) in elementary schools. Overall, this study provides evidence that GAMLSS is a powerful and flexible tool that can be used to analyze social work data. GAMLSS can be applied to important social work areas, such as addiction studies, health disparities, crime, aging-related outcomes, and homelessness, making it a tremendous utility for social work research.
Table of Contents
DOI
PubMed ID
Degree
Ph. D.