Low cost Digital Tool to Improve Early Learning Outcomes

At SSEL, we believe we can use big data and tap into rising technology around machine learning and data analytics with an “early warning” tool to help improve developmental outcomes of early learning students, 6 weeks to 5 years, improve family well-being by identifying services a family needs, and also provide information used to prove the impact of high-quality educators.

The SSEL “early warning” tool identifies early learning students not making sufficient academic and developmental progress and can help identify the needs of the family unit. Child care centers and preschools collect upwards of 2,475 data points per child. For SSEL, that is nearly 284,625 data points that could be used to make more informed decisions related to maximizing a child’s growth and development. The information received as a result of the tool would increase our ability to respond more effectively, thus improving the developmental outcomes of those students.

Since 1922, SSEL has been providing holistic, high-quality care and early childhood education to the children and families living in poverty in Columbus, Ohio’s South Side community. We aspire to live in a world where all children can grow up to access the same opportunities and find success. 

SSEL’s teaching and Family Service Center staff provide amazing outcomes for our students. 92% of our most recent graduating class left SSEL ready for kindergarten—a staggering 49% higher than the country average (Nationwide Children’s). To reach our goal of having 100% of typically developing graduates leaving our program kindergarten ready by 2025, SSEL is building an “early warning” model to identify students who are not making sufficient academic and developmental progress using data already collected on academic outcomes, attendance, family services provided, and other measures. 

Our program generates much data and outputs. With help from the Data Solutions for Change program, SSEL is building the tools and systems to improve kindergarten readiness for children on Columbus’ South Side. This tool will help staff respond more effectively to a child’s physical, social, emotional, language, literacy and cognitive developmental needs, thus improving early childhood developmental outcomes.

The South Side community of Columbus, Ohio has infant mortality rates of 18.3 deaths per 1,000 live births compared to the County rates of 8.4 infant deaths per 1,000 live births. On the South Side, 19.8% of adults 25+ years are without a high school diploma or GED and 12.4% of the labor force is unemployed. In our school, 84% of families live in poverty – 43% have a household income under $10,000.

We consistently measure the progress of each child in our program on six developmental markers, Social Emotional, Physical, Language, Cognitive, Literacy, and Mathematics. Each child at SSEL is paired with a Family Specialist whose job is to make sure the entire family’s basic needs are being met, along with support goals aimed at economic stability and self-sufficiency because parental or guardian success is just as important in ensuring the success of our students. 

Using this tool to provide real time information to staff so they are able to respond to developmental deficiencies will improve the academic and developmental progress of students, but also alert staff to family needs, as well.

The SSEL tool to improve early learning developmental outcomes is powered by big data and machine learning technology to determine how to identify students not making sufficient academic and developmental progress. Child care centers and preschools collect upwards of 2,475 data points per child. For South Side Early Learning (SSEL), that is nearly 284,625 data points that could be used to make more informed decisions related to maximizing a child’s growth and development. Using archived data from our past classes of students, and data that we regularly collect for each student and their families, we built an algorithm to help determine trends in child and family behaviors that result in delayed development, as well as critical family issues, that will be used by staff to intervene sooner and improve early education developmental outcomes, but also to indicate necessary services that could benefit the family unit. 

By using big data for little learners, SSEL will be able to understand how to better serve our children and families by identifying a need for intervention sooner. This solution will also enable us to look at our work longitudinally, tell a comprehensive story about our programming and highlight the impact of our work, while using the output to inform decision making. This system could be used by anyone having a computer or tablet, the digital tool, and a desire to improve the early learning outcomes for their students.

We hope to use this tool to become a model of support for our children and families, but also for early learning programs everywhere.

In our nearly 100 years, it has been SSEL’s experience that early education schools do not typically track data on students. Nor do they utilize data analytics to inform decisions related to the developmental impact and outcomes of students. IF data relating to the progress of students is collected, it is generally kept in paper form in binders or submitted to various funders in bits and pieces. If computer systems are used to store data, they do not communicate with one another, so trends are not recognized and opportunities to respond to developmental needs are lost. 

By using the machine learning technology and creating a tool that can constantly review and compare data, information needed to better serve students will be created. As data is entered into the system, this tool will track it, determine any trends or “red flag areas” and produce information that can be used to improve the response to developmental needs of students in real time, as well as to the needs of the entire family unit. Teachers enter daily observations and data, including any information related to the household, then this tool will be able to review that information and offer outputs in real time enabling teachers to respond immediately to academic or developmental deficiencies, as well as household needs.

Our "early warning" solution uses machine learning and data analytics along with a data warehouse that SSEL can use to make informed decisions and more immediately and effectively intervene when alerted to academic or developmental deficiencies of students during early learning.

The achievement gaps that exist for children living in poverty who begin kindergarten without high-quality education put them at an extreme disadvantage over their peers. SSEL conducts quarterly developmental assessments of our students that prove after just 90 days in one of our programs a majority of our children are at or exceed developmental proficiency in six of these domains - Social, Emotional, Physical, Language, Cognitive, Literacy, and Mathematics - and they graduate from SSEL kindergarten ready. 

In 2018, SSEL served 165 children from 20 Franklin County zip codes. Of these, 84% were living in poverty and 46% had special needs. Despite these challenges, our students still achieved 94% success on the quarterly developmental assessments.

Currently, using a holistic approach to child development that leverages the expertise of our classroom teachers, administrative staff, and social work team in the Family Service Center, when problems are identified our comprehensive approach allows for support to the child and family at all levels—increasing the likelihood for success. However, we know from research in early childhood that many developmental challenges need to be addressed quickly in order to minimize later impact. By using big data for little learners, and constantly evolving and refining our algorithm though machine learning, we will be able to identify early warning signs of risk for developmental delay and family crisis and intervene with supports before problems arise. Being able to intervene as quickly as possible can be especially critical during the first 2,000 days of a child’s brain development.

In 2018, SSEL served 165 children and is currently serving 110 students. After developing our tool, SSEL plans to develop a training on using the tool and making it available to other schools and centers as a professional development opportunity. In the next five years, in addition to the nearly 150 students served by SSEL annually, we plan to train over 480 early childhood educators and administrators so that our tool can help impact the lives of thousands of children. Our goal is to impact 5,000 children over the next 5 years with this tool.

By creating a big data solution powered by machine learning to improve the early childhood development for all children in early education schools, we are laying the foundation for a better life for these children, which could result in increased graduation rates, sustained employment, and better overall health. Because more children would have improved developmental outcomes, our hope is that they would become contributing members of society, resulting in an improved economy. But the best news is, according to exciting new research findings, the impact of high-quality early education is experiencing generational (Heckman, 2019). So, long term – children of preschoolers who attend high-quality early learning are reaping the benefits as adults (Heckman).

We plan to offer a professional development on using big data for little learners, helping early childhood educators and administrators set up systems for good data collection, data informed decision making, and use on this tool. Access to the tool will be included in the professional development. 

Another benefit of our solution that would make a difference for millions of early educators is the fact that the results of our solution will enable us, and anyone else using it, to prove the impact of their early educators. By using the outcome data, we are confident, the value of educators will be apparent and we hope to shift thinking from early educators being babysitters to thinking of them as contributors to community development. 

The current barriers to accomplishing our five-year goals are financial and technical in nature. SSEL is fortunate to have been accepted into two social enterprise accelerators to help develop a sustainable business model to support the longevity of this project. While we have volunteers through Columbus Digital Service, a group of IT and data professionals from Columbus’ corporate community, there are still costs associated with our tool to develop a minimally viable product to test our assumptions. 

We are planning to overcome these barriers by pitching at two events through the social enterprise accelerators to try and secure the funding needed. Additionally we will continue to work to foster community support and leverage volunteered goods and services to bring this project to fruition.

Three full-time employees and a 16 person advisory committee.

Serving the children and families in this community for nearly a century has positioned us to get to know our neighbors and learn about their needs. As an early education school, the community has come to depend on our programs and staff.  We use our programs, partnerships, and resources to promote the positive development of the students in our care, but also help our families holistically by providing resources for critical needs, because the children can’t thrive if the family is not thriving. SSEL serves a fragile population that has many issues with trust, but we have developed the difficult in-road to be able to support the care and services needed. SSEL’s nearly 100 years of experience, knowledge and successful collaborations places us in a prime position to make a positive collective impact on early learning for all children.

Columbus Digital Service - Helping us with data cleaning and analysis SEA Change - Social enterprise accelerator helping with business development Impact Catalyst by Cause Impact - Social enterprise accelerator helping with business development.