The spread of Coronavirus in the U.S. has not only threatened the economy, but also the American educational system. Over the course of 2020, U.S. schools have radically adapted to meet student needs. To offset the disruption caused by March school closures, most districts switched to pass-fail grading. According to research from the University of Washington Bothell, only 23% of districts maintained the traditional A-F grading scale.
After the summer break, all but four states implemented either hybrid or remote learning models with some mix of online classes, digital assignments, and reinstated standard grading. Of course, varied factors at home can create uneven learning experiences for students such as access to laptops, reliable internet access, and parental support to name a few. With some parents working or taking care of siblings, among other factors, would students engage academically? In this report, Flurry measures academic engagement by looking at one of the most common activities that directly competes with classroom and study time —playing mobile games on smartphones.
Flurry Analytics, owned by Verizon Media, is used in over 1 million mobile applications, providing insights from 2 billion mobile devices per month. For this analysis, Flurry hand-curated a sample of gaming apps. We excluded the four states that required students to physically attend school: Texas, Florida, Iowa and Arkansas. For household income levels, we used U.S. Census Bureau data and then adjusted each state by its cost of living index using data from the Council for Community and Economic Research. Finally, we used the Pew Research Center disposable income definitions for income tiers. And please note that our study looks at Gen Z users between the ages of 13 to 24, since we do not collect data for users under the age of 13.
Let’s first look at how Gen Z engagement with school has changed over the course of 2020 using mobile game usage as a signal.
In the chart above, we show the number of daily Gen Z mobile game app sessions from January through October. That’s represented by the entire span of the blue area. Each rise and fall across that topography shows how game usage cycles between weekdays and weekends, with weekend usage spiking. Within the blue area, there are four time periods. The first section is “normal learning” during which teens and young adults physically attended classes before COVID-19. The second section entitled “school closures” captures the end of the 2019-2020 school year after schools began closing due to the new pandemic. The third section shows the time period during which most schools were on summer break. Finally, the “New Normal Learning” shows the return to school for the 2020-2021 academic year during which the significant majority of all schools are teaching by video conference.
Above the blue area chart are three categories where we combined the middle two sections, as play behavior was very similar across these two middle periods. They go as follows 1) “In-person Learning” — the normal way school is attended — and which serves as our baseline, 2) “Interrupted Learning” where students had highly varied demands for attending class and doing schoolwork, or were simply on summer break and 3) “Remote Learning” where the standards for school have returned to normal, except that all but four U.S. states started including remote learning.
The key takeaway from this chart is that there is an inverse relationship between students learning in-class and how much they play mobile games. As classes shifted from in-person to a remote format, the level of supervision the instructor provided to students by being present during study time decreased. For instance, some of the exercises and material previously taught in-class by the instructor is now done online asynchronously without the instructor’s supervision. And as students take classes in a way that is less supervised by the instructor, they’re also playing a lot more games on their phone. We see this by comparing the number of gaming sessions seen during the week, Monday through Friday, versus over the weekend. For example, take a look at how often students played games when they attended school supervised by the instructor all day long before COVID-19, as shown in the “Normal Learning” part of our chart. You’ll see very pronounced spikes on the weekends. This time period has the most distinct cyclicality, with lulls during the week and peaks on the weekends. Comparing weekday to weekend usage during that period showed that students played games 43% less during the week than on the weekend.
Starting mid-march, state officials ordered schools to close resulting in most students suddenly at home on school days. Although many schools began to facilitate distance learning, the transition to a different instruction format in times of economic and health crises required a period of adjustment. Survey data shows that by May 7, only 37% of instructors had interactions with the majority of their students at least once per day when teaching remotely. And 71% of instructors shared that they were spending less time on student instruction than before the pandemic. In another survey, 72% of teachers said that pausing formal evaluations and grading made the most sense during this time. We therefore consider this period as interrupted teaching. Without at least daily class sessions supervised by the instructor, many students had additional time to fill during school days. Our data indicates that their usage of game apps during school days surged by 46%, reaching similar levels as during the summer break, when game app usage is only 1% lower on weekdays than weekends. In other words, when students did not have at least daily classes supervised by the instructor, either due to school being interrupted or students being on summer break, students switched to playing games significantly more on weekdays.
Most recently, with the return to school in a remote capacity, daily teaching time supervised by the instructor picked up again compared to the summer break and school closures. For instance, survey data shows that live-instruction, supervised by the instructor, increased from 21% during school closures to 92% during remote learning. With more class time guided by instructors, students’ game app usage during school days has gradually decreased.
Don’t You Forget About Me
Comparing in-person learning in early Spring to remote learning in early Fall may introduce some seasonal variations in gaming usage between Spring and Fall semesters that are not due to the shift in instruction type. To better isolate this change, we next compare this year’s Fall 2020, when learning is remote, to the same time period last year in Fall 2019, when learning was conducted in-person. Additionally, in order to factor out the change in users over time, we look at gaming usage per user instead of usage across all users.
In the chart above, the light blue area shows the number of daily game app sessions per user during Fall 2019, when classes were still held in-person. The bold blue line shows the same metric this Fall 2020, when learning has shifted online. Compared to last year, students have more time to fill, and our data shows that they’re playing games an average of 15% more during school days than last year.
Let’s now combine the findings from the two above charts. Considering gaming usage during in-person learning as the baseline, usage grew 15% higher during remote learning and 46% higher during interrupted learning. This shows that as class instruction time became less supervised by the instructor, students increased their usage of games during school. In other words, as supervised instruction time increases, school engagement too increases. This suggests that by providing supervised class instruction to students continuously throughout the school day, in-person teaching best curbs smartphone game usage during school days, and therefore better engages students.
Dazed and Confused
Let’s next find out how this inverse relation between class time supervised by the instructor and gaming usage varies by student income level.
In this last chart, we show the change in game usage during school days (Monday-Friday) throughout 2020 by student income level. We represent the upper income level in orange, middle income in grey, and lower income in blue. For this chart, we set usage for each income segment against its respective January baseline.
During the in-person learning time period, all income levels exhibit similar usage. After schools closed in mid-March, all students —regardless of their income level— increased their usage of game apps, with the largest surge in usage coming from the upper income student segment. During the summer break, only the lower income segment continued to play games at this elevated level, while upper and middle income students decreased their usage. With many parents working from home during weekdays this summer, the upper and middle income students may have benefited from more at-home parental supervision or restrictions on gaming compared to lower income students, whose parents may have been more concerned and affected by the economic downturn.
Most recently, with the return to school remotely, while the lower income segment plays games 87% more than during in-person learning, the middle and upper income segments play games only 2% more and 22% less respectively. This suggests that during the pandemic 1) remote learning leaves lower income students behind, who in turn play games during school days more than their peers compared to in-person learning 2) with the change from in-person to remote learning, the lower the student’s household income, the greater the increase in gaming usage during school becomes, indicating a greater decrease in engagement with school.
Note that during in-person learning in February and March, when class instruction was the most continuously supervised throughout school days, there was very little disparity in gaming usage during school across the three income curves. This again suggests that more guided and supervised class instruction effectively curbs smartphone game usage across income levels, thereby equalizing students’ engagement with school.
The primary takeaway is that as teaching time supervised by the instructor decreases, mobile game usage increases as school engagement decreases. Our study also shows that with the recent introduction of remote learning, the lower a student’s household income, the more likely they are to play mobile games during school. This suggests that during remote learning, increasing teaching time supervised by the instructor may help mitigate the disparities in school engagement across students based on their income level. For more reports covering important trends during the pandemic, subscribe to the Flurry Analytics blog and follow us on Twitter and LinkedIn.