In this work, we introduce a novel oversampling technique, the theory of inheritance and Gower distance-based oversampling (TIGO) method, designed to address class imbalance issues in mixed categorical and continuous variables data set. Drawing inspiration from genetic inheritance principles, TIGO synthesizes new minority class data,…

Predicting student performance using educational data is a significant area of machine learning research. However, class imbalance in datasets and the challenge of developing interpretable models can hinder accuracy. This study compares different variations of the Synthetic Minority Oversampling Technique (SMOTE) combined with classification…

The purpose of this dissertation is to develop principles and strategies for and identify limitations of multisite cluster randomized trials in the context of partially and fully nested designs. In the first study, I develop principles of estimation, sampling variability, and inference for studies that leverage multisite designs within the context…

The samples of evidence we use to make inferences in everyday and formal settings are often subject to selection biases. Two property induction experiments examined group and individual sensitivity to one type of selection bias: sampling frames - causal constraints that only allow certain types of instances to be sampled. Group data from both…

The simple average of student growth scores is often used in accountability systems, but it can be problematic for decision making. When computed using a small/moderate number of students, it can be sensitive to the sample, resulting in inaccurate representations of growth of the students, low year-to-year stability, and inequities for…

Learning processes often occur at a situational level. Changes in learning context have implications on how students are motivated or are able to cognitively process information. To study such situational phenomena, Experience Sampling Method (ESM) can help assess psychological variables in the moment and in context. However, data collected via…

International survey researchers are increasingly turning to a geospatial sampling approach known as gridded population sampling for finer resolution and a more updated sampling frame than conventional census-based sampling. To date, there have been no direct comparisons of accuracy between survey results derived from the two methods. This…

Place-based initiatives often use resident surveys to inform and evaluate interventions. Sampling based on well-defined sampling frames is important but challenging for initiatives that target subpopulations. Databases that enumerate total population counts can produce overinclusive sampling frames, resulting in costly outreach to ineligible…

This paper explores the effect of using multitask learning for abstractive summarization in the context of small training corpora. In particular, we incorporate four different tasks (extractive summarization, language modeling, concept detection, and paraphrase detection) both individually and in combination, with the goal of enhancing the target…

In order to promote the use of increasingly available large-scale assessment data in education and expand the scope of analytic capabilities among applied researchers, this study provides step-by-step guidance, and practical examples of syntax and data analysis using Maples. Concise overview and key unique aspects of large-scale assessment data…

Probabilistic machine learning increasingly informs critical decisions in medicine, economics, politics, and beyond. To aid the development of trust in these decisions, we develop a taxonomy delineating where trust in an analysis can break down: (1) in the translation of real-world goals to goals on a particular set of training data, (2) in the…

In recent years, harmonic regression models have been applied to implement quality control for educational assessment data consisting of multiple administrations and displaying seasonality. As with other types of regression models, it is imperative that model adequacy checking and model fit be appropriately conducted. However, there has been no…

Background: Several statistical applications including Mplus, STATA, and R are available to conduct analyses such as structural equation modeling and multi-level modeling using large-scale assessment data that employ complex sampling and assessment designs and that provide associated information such as sampling weights, replicate weights, and…

Socially anxious adolescents commonly experience impaired interpersonal functioning with unfamiliar, same-age peers. Yet, we lack short screening tools for assessing peer-related impairments. Recent work revealed that a parent-reported, three-item screening tool produced scores that uniquely related to social anxiety concerns. However, this tool…

This dissertation developed a new learner corpus of Japanese and introduced an error and linguistic annotation scheme specifically designed for Japanese particles. The corpus contains texts written by learners who are in the first year to fourth year university level Japanese courses. The texts in the corpus were tagged with part-of-speech and…