Educational research has spent decades attempting to identify which factors actually predict student outcomes. The research findings are reasonably consistent but often differ from popular assumptions about what matters in education.
This article surveys the published research on what actually drives educational outcomes.
The factors that consistently matter most
Research across multiple national contexts and decades consistently identifies a small set of factors as substantially predicting student outcomes:
1. Family socioeconomic status and parental education. The single largest predictor across most studies. Family background substantially affects academic preparation, available resources, and social environment.
2. Teacher quality. The largest school-level factor. Effective teachers produce substantially better outcomes than ineffective teachers, year over year.
3. Time on task. The amount of focused academic engagement substantially affects learning. This includes both class time and out-of-class study time.
4. Curriculum quality and rigor. What is taught and how it is sequenced substantially affects what students learn.
5. Peer environment. The students surrounding a child affect academic motivation and behavioral norms.
6. Student characteristics. Innate ability, motivation, self-regulation, and conscientiousness all affect outcomes substantially.
These factors together account for most observed variation in student outcomes.
The factors that matter less than commonly assumed
Several factors that receive substantial attention in popular education discussion turn out to matter less than is generally assumed:
School facilities and physical resources: Beyond a basic adequacy threshold, school facilities have surprisingly small effects on outcomes. The fancy gymnasium and updated science lab matter less than research-grade classroom instruction.
Class size: Class size effects exist but are smaller than commonly assumed. Reductions from very large classes (35+) to moderate classes (20-25) produce some benefit. Reductions from moderate to small (15-20) produce smaller benefit. The effect is often dominated by teacher quality differences.
Technology in classrooms: Despite enormous investment in classroom technology, the research suggests minimal impact on outcomes. Some specific technology applications produce benefits; broad technology investment generally does not.
Specific teaching credentials: Whether teachers have master's degrees or specific certifications has surprisingly small effects on student outcomes. Teaching effectiveness varies more within credentialing categories than across them.
School branding and reputation: Whether a school is "prestigious" affects enrollment and pricing more than student outcomes. Schools with similar student populations produce similar outcomes regardless of brand status.
The teacher quality finding
The teacher quality finding deserves elaboration because it is consequential and somewhat counterintuitive. Research using value-added methodologies (which estimate individual teacher effects on student outcomes) consistently shows:
- Teachers vary substantially in effectiveness, even within the same school
- The most effective teachers produce roughly 1.5x the learning of average teachers
- The least effective teachers produce roughly 0.5x the learning of average teachers
- The cumulative effect of teacher quality across years is substantial
The cumulative effect implication: a student with effective teachers across multiple years can be a full grade level or more ahead of an otherwise-comparable student with ineffective teachers across multiple years.
Despite the magnitude of teacher effects, identifying effective teachers in advance remains difficult. Credentials, experience, and observable characteristics correlate with effectiveness only weakly.
The family background finding
Family socioeconomic status and parental education substantially affect student outcomes through multiple mechanisms:
- Pre-school cognitive and language development
- Available resources for academic support
- Family expectations and cultural capital
- Peer networks and social environment
- Health, nutrition, and stability
- Access to higher-quality schools through residential choice
The family background effect is so substantial that schools serving similar student populations tend to produce similar outcomes regardless of school characteristics. Schools serving different populations produce different outcomes regardless of school quality.
This finding is sometimes discomforting because it suggests substantial limits on what schools can independently accomplish for students with weak family backgrounds. The most effective interventions typically combine school improvements with broader family support.
The implications for school choice
The research on what drives outcomes has implications for school choice decisions:
1. Within similar student populations, school differences are smaller than commonly assumed. Choosing among schools with similar student bodies often makes less difference to outcomes than is assumed.
2. Across substantially different student populations, school differences are largely peer effects. Schools with substantially better outcomes often have substantially different students.
3. Teacher quality matters more than school brand. Schools with better teachers produce better outcomes regardless of broader brand or facilities.
4. Family background matters substantially. The school can amplify or mitigate but rarely overcomes substantial family background effects.
These implications suggest that school selection should focus on teacher quality and student peer characteristics rather than facilities, brand, or marketing.
The implications for educational improvement
The research has implications for how schools should attempt to improve outcomes:
1. Improve teacher quality through hiring and development. The single highest-leverage school-level intervention.
2. Increase time on task. More focused academic engagement, both in class and out, produces better outcomes.
3. Use research-based curricula. Curriculum quality affects outcomes substantially.
4. Provide adequate resources but resist resource-focused improvement. Beyond basic adequacy, resource increases produce diminishing returns.
5. Address peer effects through composition. Schools that can influence student composition (through admissions, residential patterns, or programmatic choices) can substantially affect outcomes.
Schools that systematically pursue these interventions tend to produce better outcomes than schools that focus on more visible but less consequential changes.
The implications for families
The research suggests several practical implications for families:
- Family environment, expectations, and engagement substantially affect outcomes
- School selection should prioritize teacher quality and peer environment over brand or facilities
- Time on task matters; supporting consistent academic engagement helps
- The most consequential family interventions begin early and persist
- School cannot substitute for family role; it complements it
Families that systematically support academic engagement tend to produce better outcomes for their children than families that rely on schools to do the work alone.
The takeaway
Educational outcomes depend on a small set of factors that have been well-established by research. Many factors that receive substantial popular attention matter less than is generally assumed.
For families, schools, and policymakers, focusing on the factors that actually drive outcomes — teacher quality, time on task, curriculum quality, family engagement — produces better results than focusing on more visible but less consequential factors.
Source notes
Synthesis of published research on educational outcomes from major educational research journals, particularly meta-analyses by Hattie, value-added research from Hanushek, and educational economics research from Brookings and the National Bureau of Economic Research, 2010-2024.
