Skip to content

Project Space and Instructional Space

Pattern · All patterns
Evidence · 18 claims (12 for, 5 mixed, 1 against) · 23 studies (9 quant-synthesis, 4 causal, 4 theoretical, 3 review, 3 qualitative), q1–q4 · 9 of 23 report an effect size · 13 claims rest on one study

Description and scope

A reusable policy for a unit built around an authentic project in which each learner's component skills must be certified individually: learners work on the project → when a learner meets a gap in a component skill, the project is paused ("frozen") → the learner moves to a separate instructional space that tells, shows and gives practice with feedback on that one skill across varied situations → the learner meets an explicit criterion (with a speed criterion where the skill must become automatic) → the project is resumed and the new skill is used in it → repeat at the next gap. The intended changes are individual mastery of the component skills (a procedure or verbal-association knowledge, and the principle behind them, checked on each learner, not on the team product), their use in the project and in new tasks of the same kind (near-transfer), and, where stated, fluent performance of routines.

The design is Reigeluth's (2011) answer to four weaknesses he names in problem-based instruction: a team product hides whether every member learned; one project gives too few uses of each skill for transfer; projects build no automaticity; and unguided searching for information wastes time. It sits where three converted pages meet, and does not repeat their models. The Problem-based Learning pattern owns how much guidance a problem-centred unit gives and when; the Mastery Learning pattern owns the gate-correct-recheck cycle; and the 4C/ID pattern owns whole learning tasks with just-in-time procedural information and part-task practice for routines. This page owns the narrower configuration: a project that can be paused, a separate instructional space entered on a diagnosed gap, and a criterion that gates the return.

No claim in this wiki tests the project-space and instructional-space design as such, against an unmodified project unit or anything else. The claims that bear on it test its parts: project-based units against conventional teaching; unguided against guided discovery; mastery gating against conventional pacing; computer tutoring against other instruction; individual accountability in group work; and problem solving before instruction against instruction first. Each is carried to this pattern by extrapolation, and the page says how far. The sequence below is the earlier page's, with doses added; each step states its evidence status, and the switching rules are untested design proposals.

Inputs: establish the design brief

Field Record or elicit
Learner and starting observation Before the project starts, have each learner try the component skills the project will need (a short task for each, done alone) and one small whole-task attempt. Record which skills each learner can already do, how accurately and how fast, and where the whole-task attempt stalls. Preserve unknowns rather than labelling learners novice or advanced from a course title.
The project and its component skills The project's product or performance, who it is for, and how long it runs. List the component skills it needs and sort each by kind: facts to recall, routines to apply (and whether they must become automatic), concepts to understand, causal or theoretical understanding. Mark which gaps are predictable (most learners will hit them) and which are not.
Context and activities Setting (classroom, online-self-paced, online-instructor-led, workplace-clinical, or a mix), single-session or days-weeks, team or individual projects, group size, whether a simulation lets the project be paused with its state saved, and who delivers the instructional space (a tutoring system, a teacher, a peer, a job aid). How much facilitator time there is per learner per session.
Learner-valued goal Ask what the learner wants from the project (a working product, a qualification, a job skill, a real client served). Record where it diverges from the designer objective, for instance a learner who values finishing the product over practising a component to criterion.
Designer objective Which of these the unit is judged on: individual mastery of named component skills, quality of the team product, transfer to a new project, or automaticity of named routines. Name the one the assessment will weigh.
Outcome An individual check of each component skill on new items, separate from the team product; a measure of use in a new project or task of the same kind (near-transfer); a timed check where automaticity is the goal; immediate or delayed. Whether the check is written locally or is a standardized test, since mastery and tutoring effects were much smaller on standardized tests.

A request to "add an instructional space to our project unit" does not specify these. Ask what the component skills are, who will diagnose a gap and deliver the instruction, and what criterion gates the return before building the instructional space or setting the freeze rules.

Sequence and conditional policy

The default sequence is the earlier page's, with doses and a progression across sessions added. Steps 3, 5 and 7 have a neighbouring claim, carried by extrapolation; steps 1, 2, 4 and 6 are untested here, and every quantity is a labelled proposal.

  1. Launch the project and map its skills (first session, about 20–30 minutes; proposal). Give the authentic task, its audience, the product, and a sequence of three to five milestones that each build toward it. Run the starting check on component skills (inputs, above) and plan a freeze point for each predictable gap. Evidence status: untested; the milestone advice rests on student focus groups after project-based computing courses [+W], one qualitative evaluation coded r1, which reported preferences, not learning.
  2. Work in the project space (segments of about 30–45 minutes per session; proposal). Learners or small teams work on the current milestone, ideally in a simulation that keeps the project's state when paused. In a real-world project, keep the instructional space on a phone or laptop beside the work. Evidence status: untested; from the earlier page.
  3. Freeze at a diagnosed gap, not at every difficulty. Pause when the learner cannot proceed because a component skill is missing and a short check confirms it (step 4), or at a planned freeze point. For a principle (an idea or causal model) the project needs and learners can make a tractable attempt at, let them try first for a bounded time (about 10–15 minutes; proposal) and then teach from their attempts, rather than freezing at the first stall. For a routine or fact, freeze promptly. Evidence status: Unguided search learns less than explicit instruction, while guided discovery outperforms other instruction [+S] supports replacing unguided search with targeted instruction; problem solving before instruction [~S] qualifies when to freeze for conceptual targets (benefit on conceptual knowledge and transfer, none on procedures, reversed for second to fifth graders).
  4. Diagnose the gap (2–5 minutes; proposal). Give the learner one or two items of the suspected component skill outside the project. If they can do it in isolation, the gap is elsewhere (recognising when to use the skill, planning the next step, or the project being too hard) and the learner returns to the project with a prompt rather than entering the instructional space. Evidence status: untested proposal.
  5. Instruct and practise in the instructional space (about 10–25 minutes per gap; proposal). Tell, show and give practice with immediate feedback on that one skill, across at least three different situations rather than only the project's. Match the method to the kind of skill: drill with chunking and spaced repetition for facts; a rule with worked examples and practice items for a routine; analogies and contrasted cases for a concept; tracing causes and effects for causal understanding. For a routine that is too complex to practise whole, practise its parts first. Evidence status: tutoring systems that adjust to each learner raised achievement over large-group teaching, computer instruction and textbooks, though not over human tutors or small groups [+S]; isolating parts of a complex task reduced load for novices [+M] (one experiment); practice across varied tasks of a type for transfer [+W] (a conceptual paper).
  6. Gate the return on a criterion (proposal: at least 4 of 5 new items correct, drawn from at least two situations not used in practice; where the routine must become automatic, also a time limit set from what a competent performer takes). A learner who misses the criterion twice gets different instruction (a human explanation, a different example) before a third attempt, not the same practice again. Evidence status: mastery learning programmes raised examination scores over conventional pacing [+M]; effects were larger with a higher mastery standard and smaller on standardized tests [~M]. The criterion's numbers are a proposal; no claim here sets them.
  7. Unfreeze and use the skill in the project in the same session, then continue to the next gap. At each milestone, check every team member individually on the component skills used so far, and base part of the team's grade on members' individual results. Evidence status: in school cooperative learning, rewards based on each member's individual learning raised achievement more than a single group product [+S], carried from structured school group work to project teams.

Progression across sessions (proposal). In the first one or two sessions, the teacher or system triggers freezes at the planned points and the instructional space carries most of the teaching. In middle sessions, learners request a freeze themselves and the diagnostic check decides whether they enter. In late sessions, learners use a reference or job aid inside the project for skills already mastered, and the instructional space shrinks to short timed reviews of earlier skills. A learner moves on to the next milestone when its product meets the milestone description and every component used in it has passed its gate; a team moves to the next project when every member has. Keep milestone deadlines common to the class: in college mastery programmes, self-paced versions lowered course completion [-M] and produced smaller effects than group-paced ones [~M].

Observation Discriminating probe Proposed action, with its uncertainty
A learner stalls in the project Give one or two items of the suspected component skill on their own. If they fail, freeze and enter the instructional space. If they pass, the stall may be in recognising when the skill applies or in planning; give a prompt in the project ("which of the skills you have would get you past this?"). Neither branch is tested.
Passes the gate but does not use the skill in the project Ask the learner to say where in the project the skill applies, then watch them try. If they cannot say where, add practice items set in project-like situations. If they can say where but still do not use it, the gate items may have been too narrow; widen them. Proposed, untested.
Team product is good, one member fails the individual milestone check Ask that member to carry out one part of the product alone while explaining it. If they cannot, the team divided the labour so that member never practised; route them through the instructional space for those skills and assign them that part next time. If they can, the check may not match the project; review it.
A learner freezes again and again on the same project Count the distinct skills behind the freezes. Many different skills: the project is too hard for this learner now; give a simpler project of the same kind or more planned freeze points. The same skill again: the instruction did not hold; change the method, not the dose.
Accurate but slow on a routine that must become automatic Time the routine alone and inside the project. Slow alone: more spaced, timed practice in the instructional space. Fast alone, slow in the project: the load is in combining it with other steps; practise it inside short whole tasks. Proposed, untested.
A learner treats the instructional space as the whole course and avoids the project Ask what they want from the unit. If they value certification over the product, show how milestones count; if the project feels pointless, revisit its audience and purpose with them.

Fitting the design to a situation

The three facts that most change the decision:

  • What kind of skill each gap is. Routines and facts benefit from prompt instruction and practice to criterion; for concepts, an attempt before instruction helped conceptual knowledge and transfer in mathematics and science studies, though not for younger children. If unstated, ask: is this gap something to recall or carry out, or an idea to understand, and how old are the learners?
  • Who can diagnose a gap and deliver the instruction. The design needs a tutoring system, a teacher with time, or prepared self-study materials with a check. Tutoring systems did better than large-group teaching but not better than human tutors or small groups. If unstated, ask: when a learner is stuck, who will notice, check it and teach the skill, and how many minutes do they have per learner?
  • Whether individual mastery or the team product is the target. If only the product is graded, nothing certifies each member. If unstated, ask: what will each learner have to show alone, and what counts for the team?
If the brief says… Then change… Basis
Learners are novice in the domain Plan freeze points at every predictable gap from the starting check; keep project segments short (about 20–30 minutes); break complex routines into parts in the instructional space before whole use; give a worked example of each milestone. Unguided search learns less than explicit instruction [+S]; part-task practice for novices [+M] (one experiment)
Learners are advanced or already hold most component skills Test out of components at launch; freezes become rare and learner-requested; use the instructional space mainly for automaticity of routines; let them attempt conceptual gaps before any instruction. Guidance advantage recedes with prior knowledge [~S] (the narrative review within that claim); timing untested
Learners are child (roughly second to fifth grade) and a gap is conceptual Teach the concept first, then return to the project; do not rely on an unaided attempt first. Problem solving before instruction [~S]: effects reversed for second to fifth graders and for domain-general skills
The gap is a concept learners can attempt (conceptual-understanding), older learners Let the learner or team attempt it in the project for a bounded time (about 10–15 minutes), then freeze and teach, building on their attempts. Do not add hints during the attempt expecting a later gain. Problem solving before instruction [~S]; guidance during the attempt did not improve later learning [~M] (one article reporting two studies)
The target includes automaticity (a routine done fast, under load) Add a time limit to the gate and spaced timed reviews in later sessions; time the routine inside the project as well as alone. Untested; from the earlier page (Reigeluth's speed criterion)
One teacher for many learners, no tutoring system Run the instructional space as prepared self-study stations or short teacher-led mini-lessons for the several learners who hit the same planned freeze point at once; keep individual gates as short written checks. Untested proposal; tutoring systems did not beat small-group instruction [~S], so grouping learners with the same gap is not a known loss
Scarce facilitator time (volunteers, a busy shift, one tutor checking in briefly) Spend the facilitator's minutes only on diagnosis and on learners who missed a gate twice; let practice, feedback and the gate run on prepared material or a system. Untested proposal
online-self-paced, no teacher The system pauses the project, runs the practice and the gate, and resumes it; keep common milestone deadlines and send reminders; flag a learner who has not resumed within a week. Self-paced mastery programmes lowered completion [-M], college PSI courses; tutoring systems [+S]
Mixed modes (some learners remote, some in the room, or phone-only) Keep the instructional space and the gate on the phone for everyone, so remote and in-room learners freeze the same way; hold team milestone checks in a shared call or document. Untested proposal
Team projects Grade part of the team's result on members' individual milestone checks; rotate which member presents each part. Rewards on each member's individual learning [+S], school cooperative learning, carried to project teams
single-session only (one workshop) Run one milestone: 10 minutes launch, one planned freeze for the gap most learners will hit (15–20 minutes of instruction and a 5-item gate), 20 minutes back in the project. Name the further gaps for later. Untested proposal
A fixed short time budget per session (for example 50 minutes) Fit one cycle per session: 25 minutes project, up to 20 minutes instructional space for any frozen learner, 5 minutes back in the project or a milestone check; learners who did not freeze keep working. Untested proposal
Adults in a work setting, the project is real work Keep the project the real task; deliver the instructional space on a phone or terminal at the workplace; set gates on the routines the job needs. Untested; from the earlier page (Reigeluth's real-world variant)
The product or performance is for a real audience or client Freeze only before a public step, never during it; gate the skills the audience will see before the delivery date. Authentic projects and self-efficacy [+W] (two qualitative accounts, no comparison); timing untested
The brief wants the whole-task performance kept whole Do not split the performance into items; use the instructional space only between performances, and use the next whole performance as the gate for transfer, alongside a short check of the component. Untested proposal; whole-task practice for transfer [+W] is an argument, not a test
The brief names its own rubric, criteria or check-in routine Keep it as given and use its criteria as the gates; add only an individual check of each component beside it if the rubric scores the team product alone. Untested proposal
Learners have low confidence or low literacy Make the first gaps small and the first gates quick to pass; show rather than tell in the instructional space; let learners redo a gate privately. Untested proposal
Private or sensitive material (a learner's own business, health or family data in the project) The learner chooses what goes into the shared project and who sees their gate results; the instructional space uses invented cases. Untested proposal
High stakes (a licence, a safety-critical routine) Gate on new items written for the purpose, not on the practice items; recheck the critical routines at a delay before certifying. Effects were much smaller on standardized than local tests [~M]; delay untested

Eight of nineteen rows rest on a claim (novice and advanced learners, children with conceptual gaps, conceptual gaps for older learners, one teacher without a tutoring system, self-paced online, team projects and high stakes), and each carries its claim beyond the learners or design the claim page reports, as its Basis cell says. Two more cite a qualitative or argued claim for a nearby point. The rest are untested proposals or the earlier page's guidance.

Choosing configurations from evidence

Do not rank this design against an unmodified project unit or against direct instruction from these claims: none compares them, none measures the time pausing and resuming costs, and several outcomes in the mastery and tutoring studies were tests written for the course. A precise abstention names the missing comparison (the same project with and without a gated instructional space, on individual component checks and on a new project, at a delay) and the local observation that would stand in for it: each learner's gate results and their use of the skills in the next project.

The evidence by question

Further evidence, not yet read against this model

Claims this page cited before the 2026-10-15 rewrite, both now also used above for one point each. Neither tests the pattern.

Illustrative design instance and observation record

Invented illustration, not a tested course. Trainee pharmacy technicians, adults in a workplace-clinical programme taken partly from home, work in pairs on a six-week project: running a simulated community dispensary through a week of prescriptions, with a real pharmacist reviewing their final log. The starting check shows most can read a prescription but few can do dose calculations accurately or quickly, and the pharmacist needs calculation and label checking to be automatic. The table changed the default: calculation and label checking get planned freeze points and a timed gate (4 of 5 new calculations correct within a time a qualified technician takes), because they are routines that must be fast; drug-interaction reasoning, a concept, gets a 15-minute attempt in the simulation before instruction, since these are adults; each pair's grade includes each member's individual gate results, because the pair could otherwise divide the calculations to one person; the simulation saves the dispensary's state when a trainee freezes, and the instructional space runs on the trainees' phones so those joining from home freeze the same way. Facilitator time is one tutor for twenty trainees, so the tutor sees only trainees who miss a gate twice. One trainee says she wants to be trusted on the counter by her supervisor, which matches the automaticity gate; her partner wants to finish the log early, so the tutor shows him that the final review checks each member's calculations.

Record: project and milestone → component skill and its kind → starting check result → stall observed in the project → diagnostic item result → freeze or prompt → method used in the instructional space, minutes, and situations practised → gate attempts, accuracy and time → use of the skill after unfreezing → individual milestone check against the team product → candidate interpretations (missing skill, not recognising when to use it, project too hard, divided labour) and their basis → reobservation in the next project or at a delay. Preserve the learner's stated purpose and any change in it. An agent may summarize this record but must not fabricate missing inputs or assign numerical probabilities of mastery without a calibrated model.

Elements and limits

Simulation, just-in-time information, just-in-time learning, demonstration, worked examples, practice, part-task practice, feedback, mastery criterion with remediation, automaticity, individual accountability, scaffolding, assessment and application. Principles the pattern draws on: Problem-based Learning, Mastery Learning, Adaptive Learning, Deliberate Practice and Transfer of Learning.

This pattern is scoped to projects that can be paused and whose component skills can be named, diagnosed and checked one by one. Projects whose value lies in an unbroken performance, or whose skills cannot be separated from the whole, need the whole-task configurations in the 4C/ID pattern or cognitive apprenticeship. The policy supports observation and design reasoning; whether a project with a gated instructional space produces better individual mastery and transfer than the same project without it, for which learners, at what time cost and at what horizon, remains to be tested.

  • Epistemic Games — both embed learning within an authentic simulated task, though epistemic games focus on inhabiting a professional identity rather than alternating with a separate diagnostic instructional space
  • Problem-based Learning — the project space on its own, with the converted policy on how much guidance a problem-centred unit gives and when
  • Mastery Learning — the gate-correct-recheck cycle this pattern uses to decide when a learner returns to the project
  • 4C/ID — whole learning tasks with just-in-time procedural information and part-task practice; the closest relative, which keeps instruction inside the task rather than in a separate space

Examples

  • STAR LEGACY, a computer-based simulation cited by Reigeluth as an example project-space environment
  • A military training simulation where a trainee's project is paused for individualized skill remediation whenever a competency gap is detected, then resumed

Key Sources

  • Reigeluth, C. M. (2011). An instructional theory for the post-industrial age. Educational Technology, 51(5), 25–29. Republished in R. West (Ed.), Foundations of Learning and Instructional Design Technology. EdTech Books. https://edtechbooks.org/lidtfoundations/postindustrial_age_theory
  • Merrill, M. D. (1983). Component display theory. In C. M. Reigeluth (Ed.), Instructional-Design Theories and Models. Lawrence Erlbaum Associates.
Page metadata
Field Value
type pattern
id project-and-instructional-space
title Project Space and Instructional Space
description A reusable policy in which learners work on an authentic project, pause it at a diagnosed gap in a component skill, practise that skill in a separate instructional space to an explicit criterion and then resume the project; expected to improve individual mastery and transfer where gaps can be diagnosed and instruction delivered, but no claim in this wiki tests the design as a whole.
status review
grain_size unit
author Charles M. Reigeluth
canonical True
generated by claude/unspecified on 2026-10-15
trust unverified — no one has confirmed this page's content

Frontmatter is the page's machine-readable record — see the Schema & Guide.