Research Spotlight: Dr. Ann O’Connell

Dr. Ann A. O’Connell is Professor of Educational Statistics, Measurement, and Evaluation in the Department of Educational Psychology at Rutgers University. She was formerly a Professor in the Educational Studies Department at The Ohio State University. Dr. O’Connell’s research interests are in multilevel and generalized linear models, the use of research evidence and evidence-based interventions, and evaluation of interventions in health and education.

Can you tell us about a new research project you are working on?

As a methodologist, I have worked on many different types of projects and data sources with teams in education, health, and internationally. One of my most recent research projects is a pilot study of early education classrooms, teachers, and kindergarten student outcomes in Ethiopia. A small grant enabled my Ethiopian colleagues and me to train 10 graduate students from Addis Ababa University (students in STEM and early childhood education (ECE) programs) to conduct classroom observations, teacher surveys, head-of-school interviews, and an early grade assessment. Our observational tool was the TIPPS (Teacher Instructional Practices and Processes System), designed at NYU, which is a comprehensive tool for evaluating teacher practices and student engagement. For the student assessment, we used the IDELA (International Development and Early Learning Assessment), designed by Save the Children, which is a globally recognized tool for measuring early childhood development. We have data from 60 children, 10 schools/principals, and 25 teachers, and are currently analyzing the data and preparing a publication. Our results should help us secure larger external funding to pursue some specific research questions around teacher training, quality, and student outcomes. Sadly, there has been a lot of strife in Ethiopia in the past three years, on top of its being one of the poorest countries in the world. While our pilot findings can contribute to some recommendations for teachers and schools, one of the most rewarding aspects of our project was the collaborative training of the graduate students. Their hands-on engagement in this project helps build capacity in the education research sector and directly benefits the Ph.D. students in their research training, opening doors to other research endeavors. Several of my US students are collaborating with AAU colleagues and students on papers and projects beyond the ECE project – and my hope is to include interested Rutgers students in developing other opportunities for deeper collaboration.

What kind of methodological and theoretical approaches do you use? 

I use a lot of different methodological approaches in my work, although I likely am best known as a quantitative methodologist. I have always been intrigued by complex data and methods for analyzing data from multilevel structures and with non-normal outcomes – like counts or ordinal data. Even in the project I described above, one can see the multilevel structure at play: data collected from children nested within teachers/classrooms and schools. These complexities add some challenges to how we learn from and analyze data.      

And why are these important to your work? 

These approaches – using multilevel designs, often with complex samples and limited or discrete outcomes (again, like counts or ordinal data) can help us with research questions that can’t be directly answered through the use of single-level samples that often rely on the more typical strategies like ANOVA or regression techniques, although they are related. For example, suppose we want to know how teacher behaviors, attitudes, or school/classroom ecologies affect student learning. In that case, we need to be sampling and measuring variables at these multiple levels. Multilevel designs allow us to examine how higher-level variables affect outcomes at lower levels in a hierarchy. Ultimately, these involve interactions across levels, which can help us clarify what works best for what kinds of students.

 What’s next for you in terms of research? 

What’s next for me? Well in addition to settling in at Rutgers and getting to know my colleagues and students better, I’m very interested in learning from the community here about what their needs and interests are, particularly in quantitative science, evaluation and measurement. I have a lot of varied interests and while I love working with data and figuring things out, my hope in general is to help increase capacity in education research and methods so that we can all make a better difference in the world, in big and small ways.