Talk Data to Me: Looking into the Future with Absence Prediction
Recently, the Frontline Research & Learning Institute analyzed over five years of absence data from over 7,000 school districts nationwide to uncover trends with the goal of making future absences more predictable. In this blog post, we’ll dive into this up-to-date data and how the insights and takeaways can help your district understand the depth of your substitute pool.
5+ years of absence data
> 7,000 school districts
The Key Takeaways
Mondays and Fridays tend to have more absences than Tuesdays, Wednesdays, and Thursdays in almost every week of the school year.
Absence totals tend to rise through the fall dip in the early spring and peak in the late spring.
Human Capital Analytics and Machine Learning
Frontline’s Human Capital Analytics team decided to take this analysis a step further by applying machine learning models to daily absence data to see if they could uncover more trends, with the goal of predicting future daily absence totals.
The team figured that if district and school administrators could anticipate which days in the future were to require more substitutes than others, they could have the foresight to allocate their human capital resources most effectively to ensure high absence fill rates.
This foresight may also help:
Substitute teacher hiring
Placement of current substitutes
Implications for PD planning to minimize the effects of related absences.
After testing a few different machine learning models, it turns out that the district level, daily absence totals are actually pretty predictable!
Below is a chart showing daily absence totals from the 22-23 school year in light blue, versus the predicted totals generated by Human Capital Analytics’ machine learning model in dark blue. You can see that the machine learning model accurately predicted peaks and valleys and that the two lines generally overlap with a few exceptions.
The average district in the testing sample had an average daily absence total of 63 absences and the model predicted daily totals, within plus or minus three absences, 91% of the time.
Looking Into the Future
The good news? You don’t need to wave a wand or click your heels in shiny red slippers to predict absence trends in your district. With Human Capital Analytics, you can access machine learning predictions for the entire 23-24 school year. This tool will empower district leaders to analyze which days they can reasonably expect more or less sub-required absences and see how their absence trends rise and fall throughout the year. Look Into Your District’s Future with Human Capital Analytics
Roll Call: Best Practices for Ensuring Substitute Coverage: Consistency and continuity are key to ensuring that students receive the best possible learning experience. Yet, teacher absences are an inevitable reality. Whether it’s due to illness, professional development, or personal reasons, teacher absences can impact student learning.
Kevin is a Product Manager of Human Capital Analytics for Frontline Education. He is a former high school mathematics teacher and holds a Master's Degree in Educational Curriculum and Instruction, a Master's Degree in Educational Psychology, and is working on a dissertation toward a Ph.D. in Educational Psychology.