Student Engagement in High School: Why It Dips and What Teachers Can Do
Student engagement is lower in high school than in elementary school, and a meta-analysis of longitudinal studies on the motivation behind it points to a gradual, modest decline. The research ties the slide to falling intrinsic motivation and subject self-concept, and to how much students value the work and how they rate their classroom. A 2019 meta-analysis of 107 longitudinal studies found motivational variables and self-esteem fell by about a tenth of a standard deviation over about 1.65 years. Two classroom moves with research behind them are relevance writing and early autonomy support.
The research is clearer about the size of the dip than about its causes or its cures, and it measures motivation more often than engagement. That affects how much weight each fix deserves.
Designing for how much students value the work
In a study of 1,601 students in grades 5 to 8, falling math engagement went with how much students valued the subject, and the 2019 meta-analysis below found intrinsic motivation among the largest declines. In Rapunzl, students choose what goes into a simulated $10,000 stock and crypto portfolio priced on live Nasdaq data, so there is a current reason to look at a position and explain a decision. Project-based learning runs through the 48 modules, and the free national scholarship competition runs January to early May, which gives the second half of the school year a shared goal.
How big is the high school engagement slump?
For motivation, the slide is small on average, though real. Scherrer and Preckel's 2019 meta-analysis pooled 107 longitudinal studies and found an overall decrease of Glass's Δ = −.108 in motivational variables and self-esteem over an average of 1.654 years, which on a normal curve would move a student at the 50th percentile of motivation to about the 46th.
Where the levels end up is a separate question. Martin's 2009 study compared motivation and engagement across elementary school, high school and university. Its abstract reports elementary students as generally the most motivated and engaged, then university students, then high school students, but it gives no gap sizes and does not say whether the same students were followed over time.
Survey numbers you may have seen look much larger. In the 2016 Gallup Student Poll, 74 percent of fifth graders counted as engaged, against about one-third of students in grades 10 through 12. Those figures come from 909,617 students in schools that chose to take part, and Gallup says they may not generalize to all U.S. students. Its definition of engaged is "involvement and enthusiasm with school."
| Study | Design | Students | Finding on the dip or the fix |
|---|---|---|---|
| Scherrer & Preckel (2019) | Meta-analysis of 107 longitudinal studies | Students across the school career | Motivational variables and self-esteem fell by Δ = −.108 over about 1.65 years |
| Martin (2009) | Comparison of three school stages | Elementary, high school and university students | High school students were the least motivated and engaged |
| Martin et al. (2015) | Two measurements, one year apart; multilevel regression | 1,601 students in grades 5 to 8, 200 classrooms, 44 schools | Math engagement fell in grades 7 and 8 compared with grade 6 |
| Gallup Student Poll (2016) | Voluntary school survey | 909,617 students in grades 5 to 12 | 74% engaged in grade 5; about one-third in grades 10 to 12 |
| Hafen et al. (2012) | Start-and-end-of-course questionnaires; cross-lagged model | 578 students, 34 high school classrooms | Early perceived autonomy went with rising engagement instead of the usual decline |
| Hulleman & Harackiewicz (2009) | Randomized field experiment | High school science students | Relevance writing raised interest and grades for students with low success expectations |
Does the slump start in high school?
Not at the transition itself. The meta-analysis found that school stage and the move to middle school or high school were not significantly associated with how much motivation changed. The drift shows up across the school years.
The earlier grades matter, though. In a study of 1,601 students in grades 5 to 8, Martin, Way, Bobis and Anderson measured math engagement twice, one year apart. Compared with sixth graders, students in grades 7 and 8 had declined significantly from the previous year. By the time students reach your room, some of the slide has already happened.
What drives the decline, and which parts can a teacher move?
The same grades 5 to 8 study tied the declines to four layers: the student (particularly math self-efficacy and valuing), home (parents' valuing of math and having a computer for math), the classroom (class-average achievement and perceived climate) and the school (socioeconomic status and ethnic composition). A teacher cannot change most of that, but the student and classroom layers are where daily work happens.
The meta-analysis shows what tends to erode. Scherrer and Preckel found the largest decreases in intrinsic motivation, in math and language academic self-concepts, and in mastery and performance-approach achievement goals. Self-esteem, general academic self-concept and academic self-efficacy showed no significant mean-level change. Enjoyment of the work and the sense of being good at a particular subject slid, while general confidence in oneself held steady on average.
On the fixes, Hulleman and Harackiewicz (2009) randomly assigned high school science students to write about how what they were learning connected to their lives. The relevance writing increased interest in science and course grades for students with low expectations of success.
Hafen and colleagues (2012) followed students through a single course. They gave 578 students in 34 high school classrooms questionnaires at the beginning and the end of the course. Engagement showed the typical decline in classrooms where students did not perceive autonomy early. Where they did, it rose, and the pattern held for both students' own reports and observations of engagement.
How firm is the evidence on the dip and the fixes?
Motivation and engagement are related but not the same thing, and the evidence here does not treat them as one. The meta-analysis tracked motivational variables and self-esteem, not classroom behavior, and Martin's scale covered both motivation and engagement.
The decline is also modest and uneven. Scherrer and Preckel report that it was larger in Europe than in North America or Asia.
The fixes are not proven to reverse the slump. Hulleman and Harackiewicz tested a relevance writing assignment in science classes, and the benefit was for students who started with low expectations. Hafen's result is correlational: it ties students' perceptions in the first weeks to later engagement, but does not test a specific teacher behavior. No study described here shows that either move closes the gap between elementary and high school levels.
What can a high school teacher do this term?
- Make the first weeks count for autonomy. In Hafen's study, engagement rose in classrooms that students experienced as supporting their autonomy early, and showed the usual decline in the others. Our back-to-school playbook covers how to open a personal finance class with that in mind.
- Ask students to write the connection themselves. Writing about how a unit touches their own lives is the approach the relevance experiment tested, and the benefit showed up for students with low expectations of success.
- Keep a running record of what students can now do. The meta-analysis found subject self-concept slid. No trial here tested a fix, but a skills checklist or a portfolio costs little to try.
- Look at the classroom climate. Perceived climate was one of the classroom factors tied to math engagement declines in grades 7 and 8.
Frequently Asked
Questions
Research ties it to several things at once, including students' valuing of the subject, their self-efficacy, the perceived classroom climate and the home and school context. A 2019 meta-analysis found the steepest declines in intrinsic motivation and subject-specific self-concept. It did not find the school transition to be significantly associated with the change.
Modestly, on average. The 2019 meta-analysis found a decrease of about one-tenth of a standard deviation in motivational variables and self-esteem over an average of 1.65 years. Survey figures such as Gallup's look larger, but they count the share of students rated engaged in a voluntary sample of schools, so they are not directly comparable.
Two studies point to relevance and autonomy. A randomized experiment in science classes found that writing about personal relevance increased interest and grades for students with low success expectations. A study of 578 students found that those who perceived autonomy early in a course became more engaged rather than less.
They overlap but are not identical. The meta-analysis measured motivational variables such as intrinsic motivation and self-concepts, while engagement refers to what students do and feel in class. Martin's 2009 study measured both and found high schoolers lowest overall.
Teachers weighing what a live portfolio could change about their spring semester can request a free Rapunzl demo and ask the team about their classroom.
By Maria Rodriguez, Curriculum Designer at Rapunzl.
Get the free Rapunzl Starter Unit
Module 1, ready to teach: a slide deck, a quiz, and the answer key. Free.