September 24, 2018
Katie Brohawn
In late August, this article in Chalkbeat caught our attention, touting the benefits of a longer school day in Florida on students’ reading scores. We were fortunate to get direct access to the authors of the study, Dr. David Figlio from Northwestern University and NBER and Drs. Kristian Holden and Umut Ozek from The American Institutes for Research who kindly answered some of our most pressing questions. But first, a brief overview of the findings:
In 2012, a mandate was put into place in Florida requiring elementary schools with the lowest reading scores to extend the school day by an hour, focusing that extra hour specifically on literacy instruction. Further, the schools were required to implement a number of practices common in our own expanded learning schools including, implementing research-based practices, adapting instruction to student ability and providing curricular bridging (i.e. reading material from social studies, science, and math classes). The researchers noted that schools varied somewhere in their approach to using the extra time:
- More than three-quarters chose to either extend the end of the school day as well as rearrange the instructional day to provide additional instruction at a different time or to extend the end of the school day without rearranging
- Over two-thirds (69%) of ESD schools used a combination of small and large group instruction while 28% used small group instruction exclusively
- Nearly a third (30%) of schools used regular classroom teachers to provide the additional hour of instruction while 58% used a combination of classroom teachers and other staff such as reading coaches, though nearly all schools new to implementing an ESD reported hiring additional staff (e.g., reading coaches, teachers, paraprofessionals, or volunteers) for the additional hour
The analyses in the new article entitled Do Students Benefit from Longer School Days? Regression Discontinuity Evidence from Florida’s Additional Hour of Literacy Instruction was incredibly sophisticated. Further, they helped address limitations highlighted by researchers in two previous studies of this same initiative. These prior studies had compared all of the schools required to implement the longer school day (the lowest performing in the state) to the rest of the ~1,700 schools in the state. However, as the authors of those studies noted, these schools differed substantially in underlying characteristics with respect to poverty, demographics, etc. and thus interpretation of the findings was limited.
In the current study, the researchers used the most natural comparison group available – focusing on comparing students enrolled in schools whose reading accountability scores fell immediately below the ESD cutoff to those who immediately missed the cutoff. When doing so, results revealed that students served by the longer school day scored .05 standard deviations above those in the comparison group – a gain equivalent to approximately a month of school (which is more or less the amount of extra time provided by the ESD when added together). However, important to note is that the findings were most robust, not for students with the lowest reading skills, but rather for those with basic, yet limited, skills in reading. This is very much in line with our MS ExTRA reading initiative which specifically targets students with foundational reading skills that struggle with comprehension. Further, results were stronger among students living in poverty (eligible for free/reduced price lunch) relative to those not eligible.
Moving forward, the researchers highlight the need to examine the long-term impacts of implementing an ESD, as the current implementation practices in Florida (the schools required to implement ESD each year change based on their performance the year prior, thus moving successful ESD-implementers into the non-implementation group) have yet to allow robust research to draw firm conclusions.
Q&A:
1. Did you see any effects on school attendance, positive or negative, over time? Students that we talk to in New York are pretty clear they don’t want a longer school day—until you ask what they’d like to do and they talk about learning to read, being in safe spaces, exploring passions and working on projects that are important to them. Can you say anything about the receptivity of teachers, parents, and young people to a longer day? How might these two ideas (extra time and engagement) be working against each other and how we can prevent this from happening?
A: This is a great point! Reduced attendance has the potential to moderate, or even entirely mitigate, the effects of extended school days. While our student administrative data didn’t include this as an outcome, we’ve revisited this question using school level data, and find very little difference in attendance between students near the ESD cutoff. This suggests that the policy may not be associated with positive or negative attendance effects.
2. Did you compare the cost of extending the day to other interventions, or compare the size of the impact to what the decision-makers might have achieved with another intervention at scale? How do the results compare to other efforts Florida might have implemented in hundreds of schools?
A: Cost is an important part of ESD policies, though unfortunately, we’ve had to rely on self-reported estimates of costs from district superintendents, which roughly suggest a cost of $800 per student each year. We have thought about comparisons to other interventions, especially class size reduction. One reason that the ESD policy may be attractive is because it has a relatively small cost. For example, Krueger (2003) estimates that reducing elementary class size from 22 to 15 students was about $3,500 per student. So, even though Krueger finds much larger estimates (around 0.2 standard deviations), ESD could be more effective per dollar.
It would be important to keep in mind that class size reduction has also demonstrated important lessons about implementing class size at scale. For example, Jepsen and Rivkin (2009) have found that class size reduction in California had unanticipated effects on the pool of candidate teachers due to the large increase in hiring. Similar issues may be present if Florida attempted to expand the program to the majority of schools.
3. What are the lessons for school leaders around the country who also oversee large numbers of schools serving under-resourced families and possibly struggling with literacy gains?
A. ESD policies could be an effective tool for improving student outcomes, though as mentioned above, there are a few issues to keep in mind. First, the benefits are likely to be concentrated in low-performing readers, though not the lowest performers. Second, these policies are potentially unpopular with students, teachers, and parents who may not see the additional time as useful.
4. What’s next for your research? Will you be exploring the long-term impacts of the extended day more comprehensively?
A. We feel that there is much more to be learned about extended day policies, in terms of long-term impacts of the program in Florida, but also for the effects at higher grade levels, and the degree to which differences in implementation across ESD programs are likely to affect student outcomes.
