Predicting students' success is crucial in educational settings to improve academic performance and prevent dropouts. This study aimed to improve student performance prediction by combining advanced machine learning (ML) approaches. Convolutional Neural Networks (CNNs) and attention mechanisms were used for extracting relevant features from…

School attendance/absenteeism (SA/A) is a crucial indicator of health and development in youth but educational policies and health-based practices in this area rely heavily on a simple metric of physical presence or absence in a school setting. SA/A data suffer from problems of quality (reliability, construct validity, data integrity) and utility…

The preschool landscape is complex, consisting of several publicly funded programs supported by federal, state, and local funds. Included in this landscape is Head Start, a critical early childhood education (ECE) program that serves--in every state and territory--young children in families with incomes at or below the federal poverty line,…

Many educators in California are unaware of just how harmful out of school suspensions can be. When suspended students are barred from attending school, more often than not, the rule broken was some form of minor misconduct. This update of "Lost Instruction Time in California Schools" demonstrates that despite the important efforts by…

"Fix Injustice, Not Kids and Other Principles for Transformative Equity Leadership" offers a deep dive into the leadership values, commitments, and practices that help educational leaders create and sustain equitable schools and districts. Drawing from their extensive equity and inclusion work with schools, Paul Gorski and Katy Swalwell…

Institutional staff such as enrollment managers, business officers, and institutional researchers are often asked to predict enrollments. Developing any predictive model can be intimidating, particularly when there is no textbook to follow. This paper provides a practical framework for generating enrollment projection options and for evaluating…

This paper discusses the application value of the writing teaching mode combined with the mixed teaching mode in college English writing teaching against the background of big data. Focusing on production-oriented approach (POA) theory, this paper proposes a mixed learning writing model for English teaching and applies the POA mixed learning…

This paper combines a theoretical-based model with a data-driven approach to develop an Early Warning System that detects students who are more likely to dropout. The model uses innovative multilevel statistical and machine learning methods. The paper demonstrates the validity of the approach by applying it to administrative data from a leading…

The increased use of quantitative education data is often regarded by scholars as evidence of the emergence of 'governing by numbers'. These scholars ascribe major stakeholders such as the OECD and nation states agency as they produce, distribute and consume data, and respond to these with policy and management initiatives. This paper argues that…

Higher education leaders have repeatedly called for improved diversity, equity, and inclusion efforts, but many institutions continue to fall short. Data can play an integral role in this work; key among them are data on student demographics, including race/ethnicity. Meeting diversity, equity, and inclusion goals requires a thorough and nuanced…

The "Every Student Succeeds Act" requires that student's test scores be disaggregated by racial characteristics. Nevertheless, the author's recent study suggests that K-12 school principals may not intentionally think about race when they collect, interpret, analyze, and make decisions about data. By not disaggregating data by race,…

To successfully reduce student attrition, it is imperative to understand what the underlying determinants of attrition are and which students are at risk of dropping out. We develop an early detection system (EDS) using administrative student data from a state and private university to predict student dropout as a basis for a targeted…

Prospective longitudinal data collection is an important way for researchers and evaluators to assess change. In school-based settings, for low-risk and/or likely-beneficial interventions or surveys, data quality and ethical standards are both arguably stronger when using a waiver of parental consent--but doing so often requires the use of…

Analysing the behaviour of student performance in classroom education is an active area in educational research. Early prediction of student performance may be helpful for both teacher and the student. However, the influencing factors of the student performance need to be identified first to build up such early prediction model. The existing data…

While the statewide trends and disparities suggest that the rate of lost instruction in California due to out-of-school suspension (OSS) is about where it was before the COVID-19 school closures, this is the first report to highlight how post-COVID suspensions in 2021-2022 have added to the pandemic's harmful impact of instructional loss,…