Fractional square footage is often mistaken for modelled data. A room can be measured exactly and still require an allocation across uses, while a whole institution value can be missing and statistically imputed. This lesson separates the two processes.
This lesson continues The Gold Standard of Universities, the Gold Standard of Education and the How X Works hub. Every room, institution, use share and value is an original teaching construction. It explains facilities measurement and data treatment rather than prescribing a real campus allocation.
Keep room identity, measured NASF, research use, field shares, allocation driver, respondent-estimate status and statistical imputation flag separate. One physical room must reconcile once in every additive total.
The FY 2023 questionnaire instructs institutions to allocate shared research space among fields and uses. The public-use file guide explains that shared-space questions are not imputed and that their published denominators differ.
THE GOLD STANDARD · UNIVERSITIES
Allocation divides observed space; imputation fills missing data
A shared room can be allocated among research fields or functions from documented evidence. Statistical imputation is a different operation used for missing survey values. Keep both provenance fields when they coexist.
Fast route
Start with the decisive rule.
Start this route
Measurement route
Build the data map and evidence.
Start this route
Comparison route
Separate the nearest lookalikes.
Start this route
Transfer route
Change the context and retest.
Start this route
Reporting route
State limits and continue reading.
Start this route
Chapter contents
Open the full chapter index
- 1. Give the short two-process rule
- 2. Start with one physical room identity
- 3. Separate simultaneous from time-shared use
- 4. Do not call estimates imputation automatically
- 5. Keep total space from field distribution
- 6. Use a declared allocation driver
- 7. Avoid double allocation across dimensions
- 8. Respect non-imputation in published tables
- 9. Build denominators before percentages
- 10. Handle rounding without inventing gaps
- 11. Test allocation sensitivity
- 12. Build a changed-use transfer set
- 13. Separate allocation error from physical growth
- 14. Keep unavailable from zero
- 15. Repair common spreadsheet mistakes
- 16. Answer common questions with the process
- 17. Write a reproducible shared-space note
- 18. Use authoritative sources and continue the route
1. Give the short two-process rule
Shared-space allocation divides an observed room or area among uses or fields; statistical imputation predicts a missing survey value. One is a measurement allocation and the other is a missing-data procedure.
Original worked example. A 2,000-NASF core is observed and allocated 60 percent biology and 40 percent engineering. No value is missing, so the 1,200 and 800 NASF are allocated rather than imputed. The identifiers, values and observations are explicit so the reader can reconstruct the decision instead of trusting a label.
Changed-context transfer. Usage evidence disappears. The total room area remains observed, but the field split becomes unavailable instead of automatically becoming modelled. State what changed, predict the consequence and revise only the conclusion that depended on that condition.
Practical evidence check. Keep room ID, total NASF, allocation driver, shares, missingness and imputation status. If this check fails, mark the claim unavailable and repair the method before continuing.
Keep the raw record, classification, transformation and conclusion in separate fields. This exposes the first unsupported step and prevents a familiar total from rewriting the evidence.
End the chapter with three sentences: what the evidence establishes, what it does not establish and what observation or calculation would change the conclusion. This keeps the lesson useful when the institution, survey cycle or reporting method changes. A strong record also names the datum, denominator, section or coordinate system that governed the decision, so a later reader can repeat the reasoning without importing an unstated shortcut.
2. Start with one physical room identity
Allocation must conserve the room. Giving its full area to every beneficiary duplicates space even when all departments genuinely use it.
Original worked example. A 900-NASF microscopy room serves three fields in shares 50, 30 and 20 percent, producing 450, 270 and 180 NASF. The identifiers, values and observations are explicit so the reader can reconstruct the decision instead of trusting a label.
Changed-context transfer. A fourth field begins using the room. Re-estimate all shares so they still sum to one rather than adding another full area. State what changed, predict the consequence and revise only the conclusion that depended on that condition.
Practical evidence check. Use stable room IDs and a one-hundred-percent allocation check. If this check fails, mark the claim unavailable and repair the method before continuing.
Use a positive case, a clean negative case and an unavailable case. Unavailable evidence is not zero; it means a required observation or definition was not established.
End the chapter with three sentences: what the evidence establishes, what it does not establish and what observation or calculation would change the conclusion. This keeps the lesson useful when the institution, survey cycle or reporting method changes. A strong record also names the datum, denominator, section or coordinate system that governed the decision, so a later reader can repeat the reasoning without importing an unstated shortcut.
Previous chapter · Contents · Next chapter
3. Separate simultaneous from time-shared use
A room can be shared by schedule, by workstation, by project or by mixed function. The allocation driver should reflect the observed use rather than one convenient label.
Original worked example. A 1,200-NASF lab is research three quarters of operating time and instruction one quarter. Report 900 research NASF under the stated time basis. The identifiers, values and observations are explicit so the reader can reconstruct the decision instead of trusting a label.
Changed-context transfer. Research moves to half the time while floor area is unchanged. Research NASF changes without construction. State what changed, predict the consequence and revise only the conclusion that depended on that condition.
Practical evidence check. Retain schedules, observation period, function codes and allocation rule. If this check fails, mark the claim unavailable and repair the method before continuing.
Repeat the calculation with labels hidden and row order changed. Record disagreement at the exact step where it occurs so the repair addresses the real failure.
End the chapter with three sentences: what the evidence establishes, what it does not establish and what observation or calculation would change the conclusion. This keeps the lesson useful when the institution, survey cycle or reporting method changes. A strong record also names the datum, denominator, section or coordinate system that governed the decision, so a later reader can repeat the reasoning without importing an unstated shortcut.
Previous chapter · Contents · Next chapter
4. Do not call estimates imputation automatically
The questionnaire permits institutional estimates when exact figures are unavailable. An informed room-use estimate supplied by the respondent is still reported data, not NCSES statistical imputation.
Original worked example. Facilities staff estimate a 65 percent research share from bookings and interviews. The survey record is reported estimate with documented basis. The identifiers, values and observations are explicit so the reader can reconstruct the decision instead of trusting a label.
Changed-context transfer. NCSES later imputes a missing institution value. That model-produced value receives a different provenance flag. State what changed, predict the consequence and revise only the conclusion that depended on that condition.
Practical evidence check. Store respondent estimate status and statistical imputation flag separately. If this check fails, mark the claim unavailable and repair the method before continuing.
Carry a plausible alternative through the table. Choose the next check that separates the alternatives instead of repeating the same attractive evidence.
End the chapter with three sentences: what the evidence establishes, what it does not establish and what observation or calculation would change the conclusion. This keeps the lesson useful when the institution, survey cycle or reporting method changes. A strong record also names the datum, denominator, section or coordinate system that governed the decision, so a later reader can repeat the reasoning without importing an unstated shortcut.
Previous chapter · Contents · Next chapter
5. Keep total space from field distribution
An institution may know the total research use of a room but not its exact field split. Preserve the known margin and mark only the distribution unavailable.
Original worked example. A 3,000-NASF core is entirely research space, but field records are incomplete. Total research NASF includes 3,000 while field allocation remains unresolved. The identifiers, values and observations are explicit so the reader can reconstruct the decision instead of trusting a label.
Changed-context transfer. Usage logs establish 1,800 biomedical and 1,200 engineering. Fill the distribution without changing the institution total. State what changed, predict the consequence and revise only the conclusion that depended on that condition.
Practical evidence check. Reconcile field rows to the observed total and preserve unknown components. If this check fails, mark the claim unavailable and repair the method before continuing.
Write the result before consulting the expected answer. Compare it with independent evidence and preserve conflicts rather than smoothing them away.
End the chapter with three sentences: what the evidence establishes, what it does not establish and what observation or calculation would change the conclusion. This keeps the lesson useful when the institution, survey cycle or reporting method changes. A strong record also names the datum, denominator, section or coordinate system that governed the decision, so a later reader can repeat the reasoning without importing an unstated shortcut.
Previous chapter · Contents · Next chapter
6. Use a declared allocation driver
Possible drivers include scheduled hours, occupied workstations, project effort or documented room function. A driver is defensible only when it matches the reporting question.
Original worked example. Two fields use a core for 400 and 600 booked hours under comparable sessions, supporting a 40/60 allocation. The identifiers, values and observations are explicit so the reader can reconstruct the decision instead of trusting a label.
Changed-context transfer. One field begins running twice as many instruments per hour. Booking time may no longer represent use, so the driver is reviewed. State what changed, predict the consequence and revise only the conclusion that depended on that condition.
Practical evidence check. Record driver rationale, data window, exclusions and sensitivity to an alternative driver. If this check fails, mark the claim unavailable and repair the method before continuing.
Archive source records, calculations, crosswalks and controls. A polished summary helps reading, but it must not become the only surviving evidence.
End the chapter with three sentences: what the evidence establishes, what it does not establish and what observation or calculation would change the conclusion. This keeps the lesson useful when the institution, survey cycle or reporting method changes. A strong record also names the datum, denominator, section or coordinate system that governed the decision, so a later reader can repeat the reasoning without importing an unstated shortcut.
Previous chapter · Contents · Next chapter
7. Avoid double allocation across dimensions
Field, research function, medical-school location and animal use are separate dimensions. Allocate physical area once per additive margin while retaining independent subset flags.
Original worked example. A 1,000-NASF animal core inside a medical school is 70 percent biology and 30 percent veterinary. It remains one 1,000-NASF room with two field shares and two subset flags. The identifiers, values and observations are explicit so the reader can reconstruct the decision instead of trusting a label.
Changed-context transfer. Administrative location changes but use does not. Only the medical-school flag changes. State what changed, predict the consequence and revise only the conclusion that depended on that condition.
Practical evidence check. Run room uniqueness, field-sum and subset-less-than-total checks. If this check fails, mark the claim unavailable and repair the method before continuing.
Keep the raw record, classification, transformation and conclusion in separate fields. This exposes the first unsupported step and prevents a familiar total from rewriting the evidence.
End the chapter with three sentences: what the evidence establishes, what it does not establish and what observation or calculation would change the conclusion. This keeps the lesson useful when the institution, survey cycle or reporting method changes. A strong record also names the datum, denominator, section or coordinate system that governed the decision, so a later reader can repeat the reasoning without importing an unstated shortcut.
Previous chapter · Contents · Next chapter
8. Respect non-imputation in published tables
NCSES documentation notes that shared-space questions are not imputed. Missing shared-space responses therefore cannot be filled by silently borrowing the model used for other variables.
Original worked example. A table excludes nonrespondents for a shared-space calculation while the main NASF total includes imputed nonrespondents. The denominators differ by method. The identifiers, values and observations are explicit so the reader can reconstruct the decision instead of trusting a label.
Changed-context transfer. A later cycle supplies more observed shared-space responses. Comparability improves without retroactively creating modelled values. State what changed, predict the consequence and revise only the conclusion that depended on that condition.
Practical evidence check. Store item method, eligible denominator, response count and exclusion rule. If this check fails, mark the claim unavailable and repair the method before continuing.
Use a positive case, a clean negative case and an unavailable case. Unavailable evidence is not zero; it means a required observation or definition was not established.
End the chapter with three sentences: what the evidence establishes, what it does not establish and what observation or calculation would change the conclusion. This keeps the lesson useful when the institution, survey cycle or reporting method changes. A strong record also names the datum, denominator, section or coordinate system that governed the decision, so a later reader can repeat the reasoning without importing an unstated shortcut.
Previous chapter · Contents · Next chapter
9. Build denominators before percentages
A shared-space percentage can use institution NASF, field NASF, responding institutions or eligible rooms. State the base before interpreting movement.
Original worked example. Thirty percent of research NASF is shared among fields, while only twenty percent of institutions report any shared space. Both can be true. The identifiers, values and observations are explicit so the reader can reconstruct the decision instead of trusting a label.
Changed-context transfer. A large institution joins the response set. The space-weighted share changes more than the institution share. State what changed, predict the consequence and revise only the conclusion that depended on that condition.
Practical evidence check. Publish numerator, denominator, units, eligibility and weighting beside every percentage. If this check fails, mark the claim unavailable and repair the method before continuing.
Repeat the calculation with labels hidden and row order changed. Record disagreement at the exact step where it occurs so the repair addresses the real failure.
End the chapter with three sentences: what the evidence establishes, what it does not establish and what observation or calculation would change the conclusion. This keeps the lesson useful when the institution, survey cycle or reporting method changes. A strong record also names the datum, denominator, section or coordinate system that governed the decision, so a later reader can repeat the reasoning without importing an unstated shortcut.
Previous chapter · Contents · Next chapter
10. Handle rounding without inventing gaps
Rounded allocations can sum to 99 or 101 percent. Preserve unrounded calculations and use a declared display adjustment rather than altering a substantive category.
Original worked example. Shares 33.4, 33.3 and 33.3 display as 33, 33 and 33. The underlying NASF still reconciles exactly. The identifiers, values and observations are explicit so the reader can reconstruct the decision instead of trusting a label.
Changed-context transfer. A dashboard forces the final category to 34. Label that as display balancing, not a measured difference. State what changed, predict the consequence and revise only the conclusion that depended on that condition.
Practical evidence check. Keep unrounded shares, NASF amounts, display rule and reconciliation tolerance. If this check fails, mark the claim unavailable and repair the method before continuing.
Carry a plausible alternative through the table. Choose the next check that separates the alternatives instead of repeating the same attractive evidence.
End the chapter with three sentences: what the evidence establishes, what it does not establish and what observation or calculation would change the conclusion. This keeps the lesson useful when the institution, survey cycle or reporting method changes. A strong record also names the datum, denominator, section or coordinate system that governed the decision, so a later reader can repeat the reasoning without importing an unstated shortcut.
Previous chapter · Contents · Next chapter
11. Test allocation sensitivity
When several reasonable drivers exist, show whether the conclusion depends on the choice. A robust total may coexist with an uncertain field split.
Original worked example. Time gives 60/40 while project effort gives 55/45 for a 10,000-NASF core. Report the 500-NASF range in each field. The identifiers, values and observations are explicit so the reader can reconstruct the decision instead of trusting a label.
Changed-context transfer. A policy decision changes only if engineering exceeds 5,800 NASF. The sensitivity now matters and needs better evidence. State what changed, predict the consequence and revise only the conclusion that depended on that condition.
Practical evidence check. Compare at least two justified drivers and record decision thresholds. If this check fails, mark the claim unavailable and repair the method before continuing.
Write the result before consulting the expected answer. Compare it with independent evidence and preserve conflicts rather than smoothing them away.
End the chapter with three sentences: what the evidence establishes, what it does not establish and what observation or calculation would change the conclusion. This keeps the lesson useful when the institution, survey cycle or reporting method changes. A strong record also names the datum, denominator, section or coordinate system that governed the decision, so a later reader can repeat the reasoning without importing an unstated shortcut.
Previous chapter · Contents · Next chapter
12. Build a changed-use transfer set
Change floor area, research share, field share and imputation status one at a time. Only the corresponding totals and flags should move.
Original worked example. A four-room set includes full research, mixed research/instruction, two-field shared and missing-allocation cases. Learners rebuild totals. The identifiers, values and observations are explicit so the reader can reconstruct the decision instead of trusting a label.
Changed-context transfer. Change one room from 50 to 70 percent research while keeping field split constant. State what changed, predict the consequence and revise only the conclusion that depended on that condition.
Practical evidence check. Require before-and-after room matrix, changed fact and conserved margins. If this check fails, mark the claim unavailable and repair the method before continuing.
Archive source records, calculations, crosswalks and controls. A polished summary helps reading, but it must not become the only surviving evidence.
End the chapter with three sentences: what the evidence establishes, what it does not establish and what observation or calculation would change the conclusion. This keeps the lesson useful when the institution, survey cycle or reporting method changes. A strong record also names the datum, denominator, section or coordinate system that governed the decision, so a later reader can repeat the reasoning without importing an unstated shortcut.
Previous chapter · Contents · Next chapter
13. Separate allocation error from physical growth
A field total can rise because rooms were built, converted, remeasured or allocated differently. A change bridge should name the mechanism.
Original worked example. Engineering gains 2,000 NASF: 1,200 from a new lab and 800 from revised sharing evidence. Only 1,200 is new physical space. The identifiers, values and observations are explicit so the reader can reconstruct the decision instead of trusting a label.
Changed-context transfer. The next audit reverses 300 of the allocation change. Preserve the correction separately from construction. State what changed, predict the consequence and revise only the conclusion that depended on that condition.
Practical evidence check. Maintain opening total, additions, conversions, allocation changes, corrections and closing total. If this check fails, mark the claim unavailable and repair the method before continuing.
Keep the raw record, classification, transformation and conclusion in separate fields. This exposes the first unsupported step and prevents a familiar total from rewriting the evidence.
End the chapter with three sentences: what the evidence establishes, what it does not establish and what observation or calculation would change the conclusion. This keeps the lesson useful when the institution, survey cycle or reporting method changes. A strong record also names the datum, denominator, section or coordinate system that governed the decision, so a later reader can repeat the reasoning without importing an unstated shortcut.
Previous chapter · Contents · Next chapter
14. Keep unavailable from zero
No shared use is an observed condition; unavailable sharing evidence is not. This distinction matters to both room-level totals and published response rates.
Original worked example. Campus A confirms every core is single-field; Campus B cannot split two cores. A records zero shared rooms and B records unavailable. The identifiers, values and observations are explicit so the reader can reconstruct the decision instead of trusting a label.
Changed-context transfer. Campus B later documents a 25 percent shared area. Replace unavailable in the new release. State what changed, predict the consequence and revise only the conclusion that depended on that condition.
Practical evidence check. Use explicit zero, unknown, not applicable and observed-positive states. If this check fails, mark the claim unavailable and repair the method before continuing.
Use a positive case, a clean negative case and an unavailable case. Unavailable evidence is not zero; it means a required observation or definition was not established.
End the chapter with three sentences: what the evidence establishes, what it does not establish and what observation or calculation would change the conclusion. This keeps the lesson useful when the institution, survey cycle or reporting method changes. A strong record also names the datum, denominator, section or coordinate system that governed the decision, so a later reader can repeat the reasoning without importing an unstated shortcut.
Previous chapter · Contents · Next chapter
15. Repair common spreadsheet mistakes
Typical failures include duplicating a room for every field, applying shares twice, allocating rounded percentages to a different total and mixing current and prior schedules.
Original worked example. A pivot table sums one 2,000-NASF room three times because field labels were expanded before area allocation. Room-key reconciliation exposes 4,000 excess NASF. The identifiers, values and observations are explicit so the reader can reconstruct the decision instead of trusting a label.
Changed-context transfer. The duplicate is fixed, but percentages still come from the prior year. Version checks catch the remaining error. State what changed, predict the consequence and revise only the conclusion that depended on that condition.
Practical evidence check. Validate room keys, sum of shares, total basis, period and formula lineage. If this check fails, mark the claim unavailable and repair the method before continuing.
Repeat the calculation with labels hidden and row order changed. Record disagreement at the exact step where it occurs so the repair addresses the real failure.
End the chapter with three sentences: what the evidence establishes, what it does not establish and what observation or calculation would change the conclusion. This keeps the lesson useful when the institution, survey cycle or reporting method changes. A strong record also names the datum, denominator, section or coordinate system that governed the decision, so a later reader can repeat the reasoning without importing an unstated shortcut.
Previous chapter · Contents · Next chapter
16. Answer common questions with the process
An estimate can be reported, a missing value can be imputed, and a shared room can be allocated. These labels describe different operations, not levels of trustworthiness.
Original worked example. A reader asks whether every fractional NASF value is imputed. The answer is no; it may be the result of a documented use allocation. The identifiers, values and observations are explicit so the reader can reconstruct the decision instead of trusting a label.
Changed-context transfer. The underlying room total becomes imputed. The allocation and imputation provenance now coexist and both must be shown. State what changed, predict the consequence and revise only the conclusion that depended on that condition.
Practical evidence check. Name observation, allocation, missing-data method and release in each answer. If this check fails, mark the claim unavailable and repair the method before continuing.
Carry a plausible alternative through the table. Choose the next check that separates the alternatives instead of repeating the same attractive evidence.
End the chapter with three sentences: what the evidence establishes, what it does not establish and what observation or calculation would change the conclusion. This keeps the lesson useful when the institution, survey cycle or reporting method changes. A strong record also names the datum, denominator, section or coordinate system that governed the decision, so a later reader can repeat the reasoning without importing an unstated shortcut.
Previous chapter · Contents · Next chapter
17. Write a reproducible shared-space note
A complete note states room identities, measured areas, functions, fields, allocation driver, period, respondent-estimate status, imputation treatment and reconciliations.
Original worked example. The note rebuilds 18,000 shared NASF from twelve rooms and shows which margins remain uncertain. The identifiers, values and observations are explicit so the reader can reconstruct the decision instead of trusting a label.
Changed-context transfer. A new booking system changes two drivers. Version the allocation without rewriting prior reports. State what changed, predict the consequence and revise only the conclusion that depended on that condition.
Practical evidence check. Archive inventory, schedules, interviews, formulas, crosswalks and validation results. If this check fails, mark the claim unavailable and repair the method before continuing.
Write the result before consulting the expected answer. Compare it with independent evidence and preserve conflicts rather than smoothing them away.
End the chapter with three sentences: what the evidence establishes, what it does not establish and what observation or calculation would change the conclusion. This keeps the lesson useful when the institution, survey cycle or reporting method changes. A strong record also names the datum, denominator, section or coordinate system that governed the decision, so a later reader can repeat the reasoning without importing an unstated shortcut.
Previous chapter · Contents · Next chapter
18. Use authoritative sources and continue the route
NCSES documentation distinguishes respondent allocation and estimation from the statistical treatment of nonresponse. The rooms and values here are original.
Original worked example. Continue through research-space and animal-space lessons to keep additive field allocations separate from overlapping subset flags. The identifiers, values and observations are explicit so the reader can reconstruct the decision instead of trusting a label.
Changed-context transfer. A future survey can change its method, but one physical room still cannot be added in full to several field totals. State what changed, predict the consequence and revise only the conclusion that depended on that condition.
Practical evidence check. Save official sources, cycle, access date, driver evidence and calculation version. If this check fails, mark the claim unavailable and repair the method before continuing.
Archive source records, calculations, crosswalks and controls. A polished summary helps reading, but it must not become the only surviving evidence.
End the chapter with three sentences: what the evidence establishes, what it does not establish and what observation or calculation would change the conclusion. This keeps the lesson useful when the institution, survey cycle or reporting method changes. A strong record also names the datum, denominator, section or coordinate system that governed the decision, so a later reader can repeat the reasoning without importing an unstated shortcut.
Useful FAQ. Is a respondent allocation an imputation? No. Can one room be counted in full for every field? No. Can an unknown split coexist with a known total? Yes.
Grouped reading routes
Core route
This batch
- Separating Reported Values from Imputed Values in Research-Facility Data
- Separating S&E Field Reclassification from Physical Research-Space Change
- Separating Cross-Cycle Data Revisions from Real Research-Facility Change
Related specialist lessons
Sources and further study
FY 2023 questionnaire. Used for the standard definitions, material mechanisms or method limits developed in this lesson.
FY 2023 public-use file guide. Used for the standard definitions, material mechanisms or method limits developed in this lesson.
FY 2023 Facilities Survey. Used for the standard definitions, material mechanisms or method limits developed in this lesson.
Facilities public-use microdata. Used for the standard definitions, material mechanisms or method limits developed in this lesson.
The worked examples and transfer checks are developed here and state their own assumptions. The linked sources support the underlying facilities survey, measurement and statistical principles.
Return to contents · Return to The Gold Standard of Universities