Subjects and lessons to explore.

Law

Education · Teaching and Instruction

Differentiated Instruction

11 min read 5 practice questions
10 key termsSourcesReport an error
Want it in plain words first? Jump to Eli explains — the same idea, no jargon.
On this page 9 sections
  1. In 30 seconds
  2. Why this matters
  3. The college version
  4. Eli explains
  5. Worked example
  6. Key takeaway
  7. Quick check
  8. Study tools
  9. Sources & references

In 30 seconds

Differentiated instruction adjusts teaching in response to what students actually know. Carol Ann Tomlinson's framework names four things a teacher can vary - content, process, product, and learning environment - and three student traits to vary them for: , interest, and . is the engine: you adapt to measured readiness, not to a label, a hunch, or a learning style. The framework is popular, and the evidence for it as a package is thinner than that popularity suggests.

Why this matters

Every class contains students who are already fluent with today's objective and students who are missing its prerequisite. Differentiation is the vocabulary that teacher-preparation programs, evaluation rubrics, and school improvement plans use for that problem, so you will be asked to do it and asked to defend it. Knowing the framework lets you plan adaptations that keep the same learning goal for everyone rather than quietly lowering it for some. Knowing the evidence lets you argue about it honestly, which matters more than it sounds: the underlying moves are better supported than the branded package, and a differentiation claim in an interview or a classroom observation is far stronger when you can name which assessment drove which grouping decision.

The college version

Responsive teaching, defined

Carol Ann Tomlinson and her colleagues describe differentiation as a pedagogical change rather than an organizational one. It is not a timetable, a program, or a way of sorting students into different rooms. It is a set of decisions a teacher makes about curriculum and instruction inside one mixed classroom: proactively modifying curricula, teaching methods, resources, learning activities and the products students make, because the learners in front of you vary. Practitioners usually compress the framework into four adjustable elements and three student traits. The elements are content (what students work with, and how they get access to it), process (the activities through which they make sense of it), product (how they show what they now know), and learning environment (how the room, the groupings, and the routines are arranged). The traits are readiness, interest, and learning profile. The review's first hallmark is that effective differentiation is proactive rather than reactive, and that distinction carries more weight than it sounds. A teacher who plans one lesson for everyone and improvises when it visibly fails is not differentiating; the review argues that this kind of on-the-spot tinkering with one-size-fits-all instruction is too small a modification to meet the range in a real class. Differentiation is planned before the lesson, on evidence gathered before the lesson.

Readiness is measured, not assumed

Readiness is the trait most often misread, and getting it right is what separates differentiation from streaming with friendlier vocabulary. Readiness is not ability, not intelligence, and not a judgment about how far a student can eventually go. It is where one student stands right now with respect to one objective, and it moves from unit to unit and skill to skill: the student who needs heavy support with proportional reasoning may need none at all with data displays three weeks later. Tomlinson's review grounds readiness in Vygotsky's zone of proximal development and in the broader claim that learning happens at moderate challenge - beyond what a student can already do alone, but reachable with support. Because readiness is that specific and that unstable, it has to be measured rather than inferred. Pre-assessment is the engine of the whole approach: a short quiz, a sorting task, a writing sample, or simply the end-of-unit problems handed out at the start. It converts 'I think some of them already know this' into a decision you can defend to a student, a parent, or an observer. The federal evidence here is instructive in two directions at once. The What Works Clearinghouse practice guide on response to intervention in primary reading recommends exactly this - provide differentiated reading instruction for all students based on assessments of their current reading levels - and then rates the level of evidence for that recommendation as low, resting largely on correlational work. The companion guide on using student achievement data to support instructional decision making puts every one of its five recommendations, including the ongoing data cycle itself, at the lowest evidence rating the Clearinghouse assigns. The panels recommend the practice anyway. A well-taught version of this topic says both halves out loud.

Flexible grouping, and the tracking question

means groups that form, do a specific job, and dissolve. Membership is set by current evidence about one skill, and it is expected to change when the next check says it should. Across a single unit a student might sit in a readiness group on Monday, an interest group on Wednesday, and a deliberately mixed group on Friday. Fixed ability grouping - , streaming, setting - is the opposite: a placement made once and carried for a term or a year, often across subjects. Collapsing the two is the most common way differentiation goes wrong in practice, and the research record here is genuinely mixed rather than settled. A second-order meta-analysis synthesizing roughly a century of grouping research found positive achievement effects for within-class grouping (g of about 0.19 to 0.30), for cross-grade subject grouping (g of about 0.26), and for special grouping of gifted students (g of about 0.37), but effectively nothing for between-class grouping (g of about 0.04 to 0.06) - and the effects did not differ for high-, medium-, or low-ability students. A meta-analysis of differentiation practices in primary education points the other way on one important sub-question: homogeneous ability grouping, whether between or within classes, showed a small negative effect for low-ability students and no effect for anyone else, and the authors conclude that grouping alone does not amount to differentiated instruction. Both findings can stand. The syntheses cover different populations, outcomes, and comparison conditions. What they agree on is the part that matters for practice: the act of sorting students is not what produces gains. What changes inside the group - the materials, the pace, the complexity of the task, the amount of teacher attention - is what produces them. Tomlinson's review reaches the same conclusion from the other direction, reporting meta-analytic evidence for small within-class groups and then insisting that gains are largest when materials actually differ between those groups rather than being handed out identically.

The four moves that do the work

Most of differentiation in practice reduces to four moves. Tiering builds parallel versions of one assignment: the same essential understanding and the same standard of quality, with differences in complexity, abstraction, open-endedness, or built-in support. Scaffolding supplies temporary help - a worked example, a sentence frame, a partly completed organizer, a teacher sitting with a small group. The research literature defines scaffolding by three features, and only the first is obvious: contingency, meaning the support is adjusted to what this learner actually does; , meaning it is withdrawn over time; and transfer of responsibility, meaning the student ends up doing the work. Support that is never faded fails the definition. It is not a but a permanent accommodation, and it quietly lowers the goal for the students who receive it. Choice - choice boards, task menus, product options - differentiates by interest and by working preference while holding the objective fixed; a student who can explain photosynthesis in a diagram, a paragraph, or a two-minute recording is being assessed on the same understanding either way. Varied pacing treats time as a resource rather than a constant, which the review flags as a chronic failure: instruction is typically pitched at the middle or the top and paced for one group, leaving the rest either frustrated or idle. Every one of these moves holds the learning goal still and moves the route.

What the evidence actually supports

Differentiation is far more popular than its evidence base, and a lesson that hides this does students no favors. The strongest positive synthesis is in elementary literacy: across 18 studies of teacher-led differentiation, the overall weighted mean effect was g = +0.13, with 88 percent of individual estimates positive and considerable variation by domain - letter-word reading around +0.20, writing around +0.96. That is a real effect, and it is small. The authors also report an alarming lack of information about how teachers decided what to differentiate and for whom, which is a striking gap in studies of a practice whose whole logic is decision-making. At secondary level the picture is thinner still: a systematic review covering a decade found only 12 usable empirical studies, mostly showing small to moderate positive effects, and concluded that severe knowledge gaps remain and that convincing conclusions about effectiveness cannot yet be drawn. That review also documents why the effect sizes are hard to read. Studies count wildly different things as differentiation - ability grouping, tiering, individualization, mastery learning, heterogeneous grouping, flipped classrooms - and they typically lean on research teams to build the materials and coach the teachers, so what was tested is not what a teacher does unaided on a Wednesday. is the recurring weak point, and it cuts both ways: in the correlational study behind the federal reading recommendation, students of teachers who implemented the differentiation program with greater fidelity showed greater reading growth. The honest summary is this. Differentiation as a branded package has thin support. Its ingredients - pre-assessment and formative feedback, scaffolding with fading, small-group instruction targeted by data, tasks pitched at moderate challenge - carry better-established evidence on their own. Teach the ingredients, and treat the package as a useful organizing vocabulary rather than a proven intervention.

The learning-profile problem

Tomlinson's review defines learning profile as a student's preferred mode of learning, shaped by factors it lists as learning style, intelligence preference, gender, and culture. One part of that definition has aged badly, and pretending otherwise is not defensible. The claim that instruction produces more learning when its modality is matched to a student's stated preference - the meshing hypothesis - has been tested with the design capable of showing it, and the crossover interaction it predicts essentially does not appear; a later direct experiment with adults found no relationship between auditory or visual-word preference and comprehension when instruction was matched to preference. So do not administer a VARK-style inventory, label students visual or auditory, and route instruction accordingly. That is not differentiation, and it is not supported. What survives from learning profile is a narrower and more defensible set of things: student interest, meaningful choice, the cultural and linguistic knowledge students bring to a task, and preferences about working conditions - alone or with others, in discussion or in silence, in one long block or several short ones. Those are claims about the conditions under which a person works well, not claims that information enters certain brains better through certain sensory channels. Tomlinson's own review, to its credit, cautions that no particular approach to learning is superior to another. The cognitivism topic in this subject handles the full learning-styles debunk; the point to carry away here is only that differentiation and style-matching are different things, and the second one is not evidence.

Where Universal Design for Learning fits

Universal Design for Learning, developed by CAST, is a related but distinct framework and it is worth being able to tell them apart in a seminar. UDL organizes instructional design around multiple means of engagement, representation, and action and expression, built into goals, methods, materials, assessments, and the learning environment from the outset so that fewer individual adaptations are needed afterwards. It designs options up front, for anyone who walks through the door. Differentiation adapts for the particular learners in this room, on evidence gathered about them. A common shorthand says UDL is proactive and differentiation is reactive; avoid it, because Tomlinson's first hallmark is that differentiation is proactive too. The cleaner distinction is who the design is for - all learners in general, or these learners in particular - and when the information arrives. The two are routinely used together, and the inclusive-teaching topic covers UDL in depth.

Eli, the EliExplains learning guide

Eli explains

The same idea, in plain words

Explain it like I’m 10

Imagine a teacher has to get thirty people across the same river. Reaching the far bank is not negotiable - that is the learning goal, and it is the same for everyone. But some people can wade straight across, some need stepping stones, and a few could swim it twice and be bored waiting. Differentiated instruction means finding out who is who before anyone gets wet, handing stepping stones to the people who need them, and then taking the stones away as soon as a person can wade without them. Teachers get this wrong in two ways: guessing who needs stones instead of checking, and letting some people stop halfway across because the bank was quietly moved closer for them.

Picture it like this

Think of a doctor's office. Everyone in the waiting room has the same goal - get healthy - but nobody expects the same prescription, and nobody gets a prescription before an examination. The examination is pre-assessment, the prescription is the tiered task, and the follow-up appointment is the next check that decides whether the treatment changes or stops.

Where the picture stops working

The analogy breaks in two places. A doctor sees patients one at a time; a teacher works with thirty at once and can realistically run only a handful of versions, which is exactly why differentiation is hard to implement well and why studies of it so often find weak fidelity. And medicine has trial evidence for specific prescriptions, while differentiation as a named package does not - its individual ingredients are better supported than the whole.

Worked example

A seventh-grade teacher opens a unit on proportional reasoning with a six-item pre-assessment on Monday. Of 28 students, 6 solve multi-step proportion problems correctly, 15 handle simple unit rates but stall on scaling, and 7 cannot yet write a ratio in simplest form. The objective stays identical for all three groups: use proportional relationships to solve multi-step problems. The task is tiered. The 7 work with the teacher on ratio notation using a partly completed table. The 15 get the same word problems with a worked example on the reverse. The 6 get problems with an extra unknown and no worked example. On Wednesday an exit ticket regroups everyone: 4 of the 7 have the notation and move up, so 3 remain. The pre-assessment set the groups; the exit ticket dissolved them.

Key takeaway

Differentiation adapts content, process, product, and learning environment to assessed readiness and to student interest while holding the learning goal fixed; its component moves, especially pre-assessment and scaffolds that are faded, are better supported by evidence than the packaged approach that names them.

Quick check

3 questions here, of 5 in this lesson’s practice set. Answers stay hidden until you check.

Question 1 of 3foundational

In Tomlinson's framework, which four classroom elements can a teacher differentiate?

Choose an answer, then check it.
Question 2 of 3intermediate

A teacher gives students a style inventory, labels them visual, auditory, or kinesthetic, and delivers each group's instruction in its matching modality. What is wrong with this?

Choose an answer, then check it.
Question 3 of 3intermediate

In September a teacher sorts a class into three reading groups using last year's end-of-year scores and keeps the same membership all year. Why does this fail as differentiation?

Choose an answer, then check it.
Practice all 5

Keep learning

Ready to build on this? Continue to the next lesson.

Practice this lesson
Study tools & related lessonsYou’ll learn to · Common mistakes · Easily confused · Key vocabulary · Related

You’ll learn to

  • Define differentiated instruction as responsive teaching and distinguish it from an organizational or scheduling change.
  • Explain how content, process, product, and learning environment are adjusted in response to readiness, interest, and learning profile.
  • Distinguish assessed readiness from ability, labels, and prior placement, and explain why pre-assessment is the engine of the approach.
  • Distinguish flexible grouping from fixed ability grouping and summarize the mixed research record on grouping.
  • Apply tiering, fading scaffolds, and choice to a scenario while holding the learning goal constant.
  • Evaluate the evidence base for differentiation honestly, separating the packaged approach from its component practices.

Common mistakes

  • Treating differentiation as learning styles - giving a VARK-style inventory, labeling students visual, auditory, or kinesthetic, and matching instruction to the label.

    This is the meshing hypothesis, and adequately designed studies do not find the crossover interaction it predicts; a direct test found no relationship between modality preference and comprehension under matched instruction. Tomlinson's own definition of learning profile does name learning style as a component, which is the weakest part of the framework and should not be taught as practice. Differentiate on assessed readiness, on interest, and on working conditions - not on a sensory label.

  • Treating differentiation as workload adjustment - fewer problems for the students who are struggling, extra problems for the students who are ahead.

    Tiering varies complexity, abstraction, and support while holding the essential understanding and the standard of quality constant. Cutting the number of items lowers exposure without changing difficulty; adding items rewards speed with more of the same. Both leave the actual learning goal untouched, and the first quietly removes it from the students who most need practice.

  • Grouping by a label or a prior placement - IEP status, gifted identification, last year's reading level - and calling the result differentiation.

    Readiness is specific to the objective and changes across a term, so groups have to be re-formed on current evidence about the skill being taught. A placement fixed in September has stopped responding to anything. This is also the line between flexible grouping and tracking: between-class ability grouping shows effectively no achievement benefit in the meta-analytic record, and one primary-grade meta-analysis found a small negative effect of homogeneous grouping for low-ability students.

  • Skipping pre-assessment and differentiating from impressions of who seems strong or weak.

    Without a pre-assessment there is no readiness data, only a ranking of students by confidence, compliance, or last term's grade. Pre-assessment is what makes the adaptation defensible and what makes regrouping possible; the federal practice guide recommendation is specifically to differentiate on assessments of students' current levels, not on professional impression.

  • Presenting differentiated instruction as a proven, evidence-based practice.

    The best positive synthesis, in elementary literacy, reports an overall weighted mean of g = +0.13 across 18 studies; the secondary-level review found only 12 usable studies and concluded that severe knowledge gaps remain. Say instead that the component moves - pre-assessment, scaffolding with fading, small-group instruction targeted by data, tasks at moderate challenge - have better individual support than the packaged approach does.

Easily confused

Flexible grouping vs. Tracking or streaming

Flexible groups are formed on current evidence about one skill and dissolved when the next check says so; tracking is a placement made once and carried for a term or year, often across subjects. The meta-analytic record is broadly positive for within-class grouping and effectively null for between-class grouping, though one primary-grade synthesis found a small negative effect of homogeneous grouping for low-attaining students.

Readiness vs. Ability

Readiness is a measured, task-specific, temporary state relative to one objective; ability is a general trait attribution. Differentiation is defensible only when it responds to the first, because the second turns adaptation into a permanent expectation.

Tiered task vs. More or less work

A tiered task keeps the essential understanding and the quality standard fixed and varies complexity, abstraction, or support. Adjusting how many items a student completes changes volume, not challenge, and does not adapt the learning.

Differentiated instruction vs. Universal Design for Learning

UDL designs multiple means of engagement, representation, and action and expression into goals, methods, and materials up front for all learners; differentiation adapts for the particular learners assessed in one class. They are complementary, and both are planned rather than improvised.

Scaffold vs. Permanent accommodation

A scaffold is contingent on the learner's response and is faded as responsibility transfers to the student. Support that is never withdrawn no longer meets the definition and functions as a lowered goal.

Key vocabulary

Readiness
Where a particular student stands right now with respect to a particular objective, established by assessment rather than by a general judgment about ability.
Pre-assessment
A check given before teaching begins, used to find out what students already know so that tasks and groups can be set on evidence rather than guesswork.
Flexible grouping
Forming and dissolving small groups repeatedly across a unit, with membership set by current evidence about one skill instead of fixed for the term.
Tracking
Sorting students into separate classes or streams by measured attainment, with placement typically fixed for a term or a year and often applied across subjects.
Tiered task
A set of parallel versions of one assignment, all aimed at the same essential understanding, differing in complexity, abstraction, or built-in support.
Scaffold
Temporary support that lets a learner do something they could not yet do alone, adjusted to that learner's responses and intended to be withdrawn.
Fading
The deliberate withdrawal of support as a learner takes over more of the task; without it, help becomes a permanent substitute for the goal.
Choice board
A menu of alternative tasks or products, all pointed at the same objective, from which students select according to interest or working preference.
Learning profile
In Tomlinson's framework, a student's preferred way of working, which the review attributes to learning style, intelligence preference, gender, and culture.
Implementation fidelity
How closely a practice is actually carried out as designed when it reaches a real classroom, and a recurring weak point in studies of this approach.

Sources & references

  1. Differentiating Instruction in Response to Student Readiness, Interest, and Learning Profile in Academically Diverse Classrooms: A Review of Literature (Journal for the Education of the Gifted, 27(2/3), 119-145) — Carol Ann Tomlinson, Catherine Brighton, Holly Hertberg, Carolyn M. Callahan, Tonya R. Moon, Kay Brimijoin, Lynda A. Conover, Timothy Reynolds; Education Resources Information Center (ERIC EJ787917), IES, U.S. Department of Education
  2. Differentiated Instruction in Secondary Education: A Systematic Review of Research Evidence (Frontiers in Psychology, 10, 2366) — Annemieke E. Smale-Jacobse, Anna Meijer, Michelle Helms-Lorenz, Ridwan Maulana; Frontiers Media / PubMed Central (PMC6883934)
  3. Differentiated Literacy Instruction: Boondoggle or Best Practice? (Review of Educational Research, 90(4), 459-498) — Kelly Puzio, Glenn T. Colby, Dana Algeo-Nichols; Education Resources Information Center (ERIC EJ1259298), IES, U.S. Department of Education
  4. Effective Differentiation Practices: A Systematic Review and Meta-Analysis of Studies on the Cognitive Effects of Differentiation Practices in Primary Education (Educational Research Review, 24, 31-54) — Marjolein I. Deunk, Annemieke E. Smale-Jacobse, Hester de Boer, Simone Doolaard, Roel J. Bosker; University of Groningen research portal
  5. What One Hundred Years of Research Says about the Effects of Ability Grouping and Acceleration on K-12 Students' Academic Achievement: Findings of Two Second-Order Meta-Analyses (Review of Educational Research, 86(4), 849-899) — Saiying Steenbergen-Hu, Matthew C. Makel, Paula Olszewski-Kubilius; Education Resources Information Center (ERIC EJ1121483), IES, U.S. Department of Education
  6. Assisting Students Struggling with Reading: Response to Intervention (RtI) and Multi-Tier Intervention in the Primary Grades (IES Practice Guide, NCEE 2009-4045, February 2009) — Institute of Education Sciences, What Works Clearinghouse, U.S. Department of Education
  7. Using Student Achievement Data to Support Instructional Decision Making (IES Practice Guide, NCEE 2009-4067, September 2009) — Institute of Education Sciences, What Works Clearinghouse, U.S. Department of Education
  8. Scaffolding in Teacher-Student Interaction: A Decade of Research — van de Pol, J., Volman, M., & Beishuizen, J.; Educational Psychology Review 22(3), 271-296 (2010); ERIC EJ924182
  9. Learning Styles: Concepts and Evidence (Psychological Science in the Public Interest, 9(3), 105-119) — Harold Pashler, Mark McDaniel, Doug Rohrer, Robert Bjork; author-posted copy, Bjork Learning and Forgetting Lab, UCLA
  10. Matching Learning Style to Instructional Method: Effects on Comprehension (Journal of Educational Psychology, 107(1), 64-78) — Beth A. Rogowsky, Barbara M. Calhoun, Paula Tallal; author copy hosted by the American Psychological Association
  11. CAST Universal Design for Learning Guidelines (Version 3.0) — CAST, Inc.

EliExplains lessons are original prose written from the open, credible references above. See Copyright & Licensing.

Researched 2026-08-18

Educational content only. It is not medical, legal or professional advice. Found an error? Tell us.