Physiology Question Quality Review
Executive Summary
This review covers 200 validated non-gold candidate questions sampled from a Physiology pool of 10,474 items, analyzed across eight shards of 25 questions each. The findings are consistent and mutually reinforcing across shards: the dominant quality problem is a severe overrepresentation of low-cognitive-demand recall items that add no discriminatory value at PG entrance level, compounded by a secondary layer of structural defects, factual errors, and topic misclassification that affect a meaningful minority of questions.
The Bloom's distribution of the sampled set (93 questions at Level 1, 89 at Level 2, only 18 at Levels 3–5 combined) is the single most important quality signal. The benchmark and PYQ gold standard shows that even "easy" questions are expected to test comprehension of a mechanism or direction of change — not isolated name-to-function matching. The candidate pool inverts this expectation: the majority of questions are definitional lookups, bare numerical recall, or single-fact retrieval with no clinical hook.
Beyond the Bloom's problem, the reviewed set contains confirmed factual errors in the keyed answer (at least 6–8 questions across shards), structural defects that make questions non-functional (duplicate options, stem-option mismatches, undefined variables in formulas), and a recurring pattern of questions that belong to Biochemistry, Pathology, Surgery, or Blood Banking rather than Physiology.
A smaller but important positive signal also emerges: a cluster of well-constructed applied and integrative questions exists in the pool and represents what the subject should look like at scale. These items — ABG interpretation, Starling force calculations, cardiac cycle integration, reproductive pathway sequencing, and integrative clinical scenarios — are the quality ceiling of the sampled set and should serve as the template for remediation.
Headline numbers from the sampled set:
- Estimated proportion requiring no action beyond keep: ~25–30%
- Estimated proportion requiring a targeted fix before use: ~25–30%
- Estimated proportion warranting disable: ~40–45%
What Good Looks Like
The benchmark and PYQ gold standard items share a consistent profile that the content team should use as the quality bar for all remediation decisions.
Mechanism over name. The best items do not ask "what is X" — they ask "what happens when X changes" or "which of the following is NOT a mechanism of X." The ADH question (ae89195f) does not ask what ADH does; it asks which of four specific mechanistic statements is false, requiring the candidate to hold the full mechanism in mind. The pulsus alternans question (b56c8be9) does not ask what Frank-Starling law is; it asks which clinical sign is explained by it.
All necessary data are in the stem. The tetanization frequency question (e9f16e48) gives twitch duration = 40 ms and asks for the minimum frequency — a clean Bloom's-3 calculation where the answer cannot be retrieved from memory alone. The net filtration pressure question (fbb7b4a4) gives all four Starling variables. No assumption is required.
Distractors require knowledge to reject. In the insulin resistance question (0c1a2ed3), the distractors (RBCs, brain, kidney) are all tissues that genuinely do not require insulin for glucose uptake — a candidate who does not know the GLUT transporter distribution cannot eliminate them by common sense. In the Valsalva question (a11e0aed), the distractors are mechanistically adjacent and require understanding of the full pressure sequence.
Clinical or physiological consequence is the frame. Even the simplest benchmark items are anchored to a consequence: surface tension increases during expiration (c53f956c) — not "what is surfactant." Respiratory alkalosis at altitude (59e1e176) — not "what happens to breathing at altitude."
Negative stems are used sparingly and precisely. When NOT/EXCEPT format is used (ae89195f, 9e6314e3), the correct answer is a clean false statement and the distractors are all unambiguously true. There is no ambiguity about which option is the odd one out.
The candidate pool's best items — ABG interpretation (a7c9d255), DI and RAAS compensation (660698e2), Starling force calculation (c5ca219b), estrogen synthesis pathway ordering (9332a0cf), marathon hyponatremia vignette (42a9dd21) — all share these properties and should be treated as internal benchmarks for the subject.
Main Issue Categories
1. Bloom's-1 Recall Saturation: The Dominant Pool-Wide Problem
Why this pattern is bad
PG entrance examinations at INICET and NEET-PG level do not reward isolated fact retrieval. A question that asks "which organelle synthesizes steroids" or "what is the half-life of T4" can be answered correctly by a candidate who has no functional understanding of physiology. These items fail to discriminate between candidates who understand mechanisms and those who have only memorized lists. The benchmark standard makes this explicit: even the easiest gold-standard items (Bloom's-1 tagged, difficulty-1) are anchored to a clinical or mechanistic consequence. The candidate pool systematically violates this standard.
The Bloom's distribution data confirms the scale: 93 of 200 sampled questions are Level 1, and a further large proportion of the 89 Level-2 items function as Level 1 in practice (single-step comprehension with no genuine application). This is not a marginal skew — it is a structural feature of the pool.
How it shows up
The pattern appears across every topic area without exception. It is not concentrated in one subtopic. Representative examples from the reviewed set include:
- Organelle function recall: "Protein synthesis occurs in rough ER" (76d67563), "Steroid synthesis takes place at smooth ER" (537908f8), "H₂O₂ catabolism by peroxisomes" (c24d5051)
- Hormone source recall: "Erythropoietin is secreted by the kidney" (dfe3f508), "MIS is secreted by Sertoli cells" (06335b8c), "LH triggers ovulation" (64af93b0)
- Numerical fact recall: "T4 half-life = 6 days" (0a35bed1), "Normal GFR = 125 ml/min" (0c0db53b), "Insensible water loss = 50 mL/hour" (benchmark PYQ — acceptable because PYQ-verified; the candidate pool equivalents are not)
- Anatomical label recall: "Baroreceptors are in the carotid sinus" (8b941872), "Biceps jerk = C5, C6" (89090fc1), "Sleep centre = hypothalamus" (523b8960), "Hunger centre = hypothalamus" (ed00a10c)
- Definition recall: "QRS complex = ventricular depolarization" (27076298), "Sarcomere = area between two Z lines" (acceptable as PYQ; the candidate pool equivalents are not PYQ-verified), "Connexons are hexamers" (2f33de01)
The neurophysiology and sensory physiology subtopics are particularly saturated: questions on hypothalamus, olfaction, taste, sleep centres, and reflex arc components appear repeatedly as pure name-to-function matching with no applied framing (ed00a10c, 6c845126, b4b5f2f0, 7060c6c9, 523b8960, b943b73f).
Recommended disposition
Items in this category that have no PYQ tag, no template membership, and no clinical framing should be disabled. The pool already contains PYQ-verified Bloom's-1 items (e.g., sarcomere definition, insensible water loss, cystic fibrosis channel) that serve the legitimate function of anchoring basic recall. Untagged duplicates of the same conceptual territory add no value and dilute the pool's discriminatory power.
Items that test a genuinely important concept but are framed as pure recall should be fixed by adding a clinical or mechanistic stem (see Section 6 for examples). Items where the concept itself is too trivial for PG level (e.g., atmospheric pressure in psi, sperm velocity in mm/min) should be disabled regardless of framing.
Example IDs: 76d67563, 537908f8, dfe3f508, 64af93b0, 0a35bed1, 523b8960, cfc7f52c, 5f22f55c, 0c0db53b, a4635baa, 27076298, 89090fc1, b943b73f, ed00a10c, ea3ea83f, 8b941872, 2f33de01, aaffc563, f6ad138b, 31bc419e
2. Confirmed Factual Errors in the Keyed Answer
Why this pattern is bad
A question with a wrong answer key is worse than no question at all. It actively misteaches candidates, generates reports, and erodes platform credibility. This is the highest-priority remediation category because it requires correction before any use, regardless of the question's other qualities. In the reviewed set, factual errors in the keyed answer appear at a rate of approximately 1 in 25 questions — a non-trivial prevalence that, if representative of the full pool, implies several hundred incorrect answer keys across the 10,474-question bank.
How it shows up
The errors observed in this sample fall into three subtypes:
Outright factual inversion: The keyed answer states the opposite of the correct physiology.
- b89719bf keys "Increase in blood volume" as the effect of ANP. ANP causes natriuresis, diuresis, and decreased blood volume. The correct answer should be "Decrease in renin secretion."
- e48be40f keys HCO₃⁻ as the "predominant extracellular ion." The predominant extracellular cation is Na⁺ and the predominant anion is Cl⁻. HCO₃⁻ is the second most abundant anion.
- feaa8081 marks "synovial fluid is transcellular fluid" as NOT true. Synovial fluid is, by standard classification, a transcellular fluid. The answer key is inverted.
Sequence or cycle errors: The keyed answer places a physiological event in the wrong position in a sequence.
- bbe52d08 marks "isovolumetric relaxation precedes ventricular ejection" as correct. In the cardiac cycle, IVR follows ejection (after aortic valve closure), not precedes it.
Mechanistic conflation: The keyed answer is defensible under one interpretation but wrong under the standard teaching framing.
- cf7b0860 keys "Arterial CO₂" as what the central chemosensitive area responds to. Standard teaching (Ganong, Guyton) specifies that the direct stimulus is CSF H⁺; CO₂ is the indirect trigger that crosses the BBB and lowers CSF pH. The question as written conflates the mechanism.
- a41d02ba marks "amount of air that can be forcibly exhaled after normal exhalation" as the definition of FRC. This is the definition of ERV. FRC = ERV + RV.
Undefined or non-standard formula: The keyed answer uses a formula that does not correspond to any standard textbook formulation.
- 2f96aa7a marks T = WP/R as representing Laplace's law. The variable W is undefined and this formula does not appear in standard physiology texts. The standard vessel wall tension formula is T = Pr.
Recommended disposition
All items in this category require fix before any use. For outright inversions (b89719bf, e48be40f, feaa8081, bbe52d08, a41d02ba), the fix is straightforward: correct the answer key and verify all distractors remain valid. For mechanistic conflation items (cf7b0860), the fix requires either rewriting the stem to specify the level of mechanism being asked about, or changing the correct answer and adjusting distractors accordingly. For the formula error (2f96aa7a), the fix requires replacing the formula with the standard textbook version and rewriting the stem to specify the physiological context (vessel wall tension or alveolar surface tension).
Example IDs: b89719bf, e48be40f, feaa8081, bbe52d08, a41d02ba, cf7b0860, 2f96aa7a
3. Structural Defects That Make Questions Non-Functional
Why this pattern is bad
Structural defects — duplicate options, stem-option mismatches, undefined variables, broken image dependencies — make a question impossible to answer correctly regardless of the candidate's knowledge. These are not quality problems in the usual sense; they are functional failures. A candidate encountering a question with two identical options, or a stem that asks about organs but provides time periods as options, cannot be assessed at all. These items must be identified and removed from active use immediately.
How it shows up
Several distinct structural defect subtypes appear in the reviewed set:
Duplicate options: 5f52e05d has two options both reading "Right to left shunt" — one marked correct, one not. The question is also factually inverted (right-to-left shunt causes low arterial PO₂, not high alveolar O₂ tension), compounding the defect. a8c28951 (NMJ events) has two options with identical text for options B and C.
Stem-option mismatch: d2c9c9f5 asks which organs are responsible for fetal RBC production, but the four options are time periods (prenatal life, first year of life, up to five years, lifelong). The options answer a completely different question. This appears to be a scrambled or incorrectly assembled item.
"All of the above" as keyed correct answer: This is a structural flaw rather than a factual error. When "All of the above" is the correct answer, a test-wise candidate can identify it by finding any two options that are true, without evaluating the third. This format is explicitly prohibited in quality MCQ construction guidelines. It appears as the keyed correct answer in at least four questions in the reviewed set: 212a0f1e (Fick method components), c19494ab (gastric acid secretion), 77f9a4f7 (cytokines/monokines/lymphokines), and 9332b1f3 (pneumothorax management). The prevalence across eight shards suggests this is a bank-wide pattern requiring a systematic sweep.
Incomplete stem for calculation questions: 10ff8622 asks for cardiac output at HR 70 beats/min but provides no stroke volume in the stem. The answer (5.25 L/min) implies SV ≈ 75 mL, which must be assumed. A calculation question without all necessary data is not a calculation question — it is a recall question about a standard assumed value, which is a different and weaker item type.
True/False/Partially true/Partially false option format: 15e1ff52 uses this four-way format for a cardiac valve question. This format is not acceptable for PG MCQ standards and provides no meaningful discrimination.
Recommended disposition
Items with duplicate options (5f52e05d, a8c28951) and stem-option mismatches (d2c9c9f5) should be disabled immediately — they cannot be used in any assessment context. Items using "All of the above" as the keyed correct answer (212a0f1e, c19494ab, 77f9a4f7) should be fixed by restructuring as single-best-answer questions; if the underlying concept is sound, the fix is straightforward. Items with incomplete calculation stems (10ff8622) should be fixed by adding the missing data to the stem. The True/False format item (15e1ff52) should be disabled in current form and rewritten as a standard four-option MCQ if the content is worth preserving.
Example IDs: 5f52e05d, a8c28951, d2c9c9f5, 212a0f1e, c19494ab, 77f9a4f7, 9332b1f3, 10ff8622, 15e1ff52
4. Topic Misclassification and Subject Boundary Drift
Why this pattern is bad
Questions filed under the wrong topic or wrong subject create two problems. First, they appear in the wrong assessment context — a candidate preparing for a Physiology test encounters a Pathology or Blood Banking question, which is disorienting and unfair. Second, they inflate the apparent coverage of a topic while actually testing something else, making topic-level quality metrics unreliable. In the reviewed set, misclassification appears in two forms: wrong topic within Physiology (e.g., a cardiovascular question filed under Acid-Base Balance), and wrong subject entirely (e.g., a Biochemistry or Surgery question filed under Physiology).
How it shows up
Wrong subject — Biochemistry bleed-in: 46553bd2 (selenocysteine in deiodinase) is a molecular biochemistry fact with no physiological reasoning component. c24d5051 (H₂O₂ catabolism by peroxisomes) is a cell biology/biochemistry item. de934fa0 (calcium-binding proteins) is borderline Biochemistry. e03ef7ba (lysozymes in eukaryotes) tests microbiology taxonomy.
Wrong subject — Clinical Medicine/Surgery: 48637953 (Zollinger-Ellison syndrome triad) is a clinical medicine/surgery fact. 9332b1f3 (spontaneous pneumothorax management) is a clinical management question belonging to Surgery or Respiratory Medicine.
Wrong subject — Pathology: 1d7b6246 (SIRS pathogenesis) is a well-constructed Bloom's-4 question but belongs to Pathology/General Surgery, not Physiology.
Wrong subject — Blood Banking/Transfusion Medicine: 4fd152c3 (glucose added to stored blood) is a blood banking question. The correct answer is also debatable in a physiology framing (glucose primarily prevents ATP depletion and shape change, not simply "nutrition").
Wrong topic within Physiology: b1038f1c is filed under "Acid-Base Balance" but its content is entirely about sympathetic nervous system activation and arteriolar norepinephrine release — a cardiovascular/autonomic topic. 9ef63e7a (memory cells and nerve growth factor) is tagged under "Cardiovascular System." 763dd469 (erythropoietin inhibited by estrogen) is placed under Renal Physiology but is primarily an endocrine/haematology concept.
Recommended disposition
Questions that belong to a different subject entirely (Biochemistry, Surgery, Pathology, Blood Banking) should be disabled from the Physiology bank and routed to the appropriate subject if the question quality is otherwise acceptable. Questions that are correctly within Physiology but filed under the wrong topic should be fixed by updating the topic tag — no content change is needed. The content team should treat this as a systematic audit task rather than a case-by-case fix, given the prevalence observed across shards.
Example IDs: 46553bd2, c24d5051, e03ef7ba, 48637953, 9332b1f3, 1d7b6246, 4fd152c3, b1038f1c, 9ef63e7a, 763dd469, de934fa0
5. Difficulty and Bloom's Metadata Miscalibration
Why this pattern is bad
Metadata tags for Bloom's level and difficulty are used downstream for test assembly, adaptive learning sequencing, and quality filtering. When a question is tagged Bloom's-5 but tests a single recalled fact, or tagged difficulty-2 but is trivially easy, the metadata actively misleads the system. Candidates receive items that do not match their intended challenge level, and quality filters that rely on Bloom's tags to exclude low-level items will fail to catch mislabeled questions.
How it shows up
Bloom's level over-assigned: e9547e9d (bitter taste localized to posterior tongue) is tagged Bloom's-5 (Evaluate/Create). This is a pure recall fact — Bloom's-1 at most. The question is otherwise usable but the tag is wrong by four levels. b51f0b24 (cardiac output increases in exercise) is tagged at a level implying application but functions as recall.
Bloom's level under-assigned: Several questions tagged Bloom's-1 actually require comprehension of a mechanism (e.g., 84f0e8b3 on renin secretion inhibition by increased Na⁺ in DCT functions as Bloom's-2). These are less harmful but still create calibration noise.
Difficulty string vs. numeric inconsistency: A non-trivial subset of questions in the reviewed set uses string labels ("easy," "medium") rather than numeric codes (1–5), inconsistent with the benchmark standard. Questions caaccf7c, ed043a2b, 728f8bd5, 8e39b7b5, and e0e0d2c4 all carry string difficulty values. This is a metadata normalization problem that affects downstream filtering.
Difficulty rating inconsistent with actual item demand: 35e74585 (surfactant reduces surface tension) is tagged difficulty-2 but is straightforward Bloom's-1 recall. 9b256074 (epidermal turnover = 4 weeks) is tagged at a difficulty level that implies some challenge, but the question is pure numerical recall.
Recommended disposition
Bloom's and difficulty metadata errors do not by themselves make a question unusable, but they must be corrected before the question enters any adaptive or difficulty-stratified assessment. Items with Bloom's tags that are wrong by more than one level (e.g., e9547e9d at Bloom's-5) should be fixed with corrected metadata. Items with string difficulty labels should be fixed by normalizing to numeric codes. This category is operationally distinct from content fixes and should be handled as a metadata normalization pass rather than a content review pass.
Example IDs: e9547e9d, 35e74585, caaccf7c, ed043a2b, 728f8bd5, 8e39b7b5, e0e0d2c4, 9b256074
6. Worthwhile Concepts Undermined by Weak Execution
Why this pattern is bad
This category is distinct from pure recall saturation. These questions test concepts that are genuinely high-yield and exam-relevant, but the execution — stem phrasing, distractor quality, or clinical framing — is weak enough that the question either fails to discriminate or creates ambiguity. Unlike the Bloom's-1 recall items (which should mostly be disabled), these questions are worth fixing because the underlying concept is sound and the remediation path is clear.
How it shows up
Stem fragment or grammatically incomplete stem: ff737f36 (leptin linking obesity and puberty) has a stem that reads as a fragment: "The hormone that links obesity and puberty." This is not a question. The content is correct and the item is NEET-PG tagged, but it cannot be used until the stem is rewritten as a complete question.
Misleading preamble that obscures the actual question: b79ffef6 opens with "K⁺ is the most abundant intracellular cation" — a true statement that is entirely irrelevant to the question being asked (which GI secretion has the highest luminal K⁺ concentration). The preamble misleads candidates into thinking the question is about intracellular vs. extracellular distribution.
Ambiguous stem that makes the correct answer defensible only under a specific unstated interpretation: 3091d670 asks about conditions where oxygen content is reduced, with methemoglobinemia as the "except." Methemoglobin does reduce true arterial oxygen content; the question is only defensible if it means SpO₂ as measured by pulse oximetry (which reads falsely normal in methemoglobinemia). The stem does not specify this, making the question factually problematic as written.
Distractor that is too close to correct to serve as a clean wrong option: 0227ade9 (gastric emptying regulation) marks "enteric reflexes" as correct over "local hormones in the duodenum." Both are major regulators; the distractor is not clearly wrong. The question needs a qualifier ("primary" or "predominant") and a reference-anchored answer.
Clinically relevant concept buried in a trivial stem: 99cee417 (sodium fluoride and the oxygen dissociation curve) tests a genuinely non-trivial concept (NaF inhibits glycolysis → prevents 2,3-DPG synthesis → left shift, but at blood tube concentrations has no net effect on the ODC). The question as written provides no clinical context, making it appear to be trivia rather than applied physiology.
Distractor construction errors that create internal inconsistency: d8ee064a (leptin function) includes "All of the options" as a distractor, but one of the other options ("Decrease lipolysis") is factually wrong for leptin. A candidate who knows leptin increases lipolysis will correctly reject that option — but then "All of the options" becomes an impossible distractor, which signals the correct answer by elimination rather than knowledge.
Recommended disposition
All items in this category should be fixed. The fixes are targeted and do not require conceptual reconstruction: rewrite the stem as a complete question (ff737f36), remove the misleading preamble (b79ffef6), add a clinical qualifier to the stem (3091d670, 99cee417), add "primary/predominant" to the stem (0227ade9), replace the internally inconsistent distractor (d8ee064a). These are editorial fixes, not content rewrites.
Example IDs: ff737f36, b79ffef6, 3091d670, 99cee417, 0227ade9, d8ee064a, 45ba19bf, cd409db9, cf8029ec, b51f0b24
7. Underrepresentation of Higher-Order and Integrative Question Types
Why this pattern is bad
This is a gap problem rather than a defect problem, but it has direct operational consequences. The Bloom's distribution of the sampled set (93 Level-1, 89 Level-2, 18 at Levels 3–5) means that the pool is structurally unable to support high-discrimination assessments. INICET and NEET-PG increasingly test integrative reasoning — multi-system scenarios, quantitative application, pathway sequencing, and clinical vignettes. A pool dominated by Bloom's-1 and Bloom's-2 items cannot generate tests that match this standard.
The gap is not uniform across topics. In the reviewed set, certain topic areas are almost entirely unrepresented at Bloom's-3 and above:
- Reproductive physiology has almost no pathway-application or comparative-physiology questions; the few items present are Bloom's-1 recall (64af93b0, 0a35bed1, 31bc419e).
- Exercise physiology has only one Bloom's-3 item in the entire reviewed set (b23502aa on a-vO₂ difference during maximal exercise).
- Sensory physiology (auditory, visual, olfactory, gustatory) is represented almost entirely by name-to-function matching items with no pathway or consequence framing.
- Neurophysiology — autonomic is similarly dominated by Bloom's-1 items (ea3ea83f, a8685163).
By contrast, the topic areas that do have higher-order items — cardiovascular (ECG axis, Valsalva, Frank-Starling), renal (filtration pressure calculations, ADH mechanisms), and endocrinology (Wolff-Chaikoff, steroidogenesis enzyme deficiencies) — are the areas where the pool most closely matches the benchmark standard.
How it shows up
The positive examples in the reviewed set define what is missing at scale: ABG interpretation requiring numerical reasoning (a7c9d255), DI and RAAS compensation requiring multi-system integration (660698e2), Starling force calculation requiring formula application (c5ca219b, fbb7b4a4), estrogen synthesis pathway sequencing requiring understanding of the two-cell model (9332a0cf), marathon hyponatremia requiring clinical vignette reasoning (42a9dd21), equilibrium potential calculation (428cbdc2), ECG axis determination (faad35aa).
These items exist in the pool but are outnumbered approximately 6:1 by recall items. The content team should treat this as a seeding priority for new question development, not just a curation problem.
Recommended disposition
This category does not generate disable or fix calls on existing items. It generates a development priority signal: the topics most deficient in Bloom's-3/4 coverage (reproductive physiology, exercise physiology, sensory physiology, autonomic neurophysiology) should be targeted for new question development using the integrative and calculation-based templates already present in the pool's best items. Existing Bloom's-1 items in these topics that are disabled should be replaced, not simply removed.
Representative high-quality items to use as development templates: a7c9d255, 660698e2, c5ca219b, fbb7b4a4, 9332a0cf, 42a9dd21, 428cbdc2, faad35aa, b23502aa, 53e8189f
Prioritization
The following priority ordering is recommended for the content operations team, based on urgency and operational impact:
Priority 1 — Immediate action required (before next assessment cycle)
Disable all questions with structural defects that make them non-functional: duplicate options (5f52e05d, a8c28951), stem-option mismatches (d2c9c9f5), and the True/False format item (15e1ff52). These cannot be used in any assessment context and should be removed from active pools immediately.
Fix all confirmed factual errors in keyed answers: b89719bf (ANP increases blood volume — wrong), e48be40f (HCO₃⁻ as predominant extracellular ion — wrong), feaa8081 (synovial fluid not transcellular — wrong), bbe52d08 (IVR precedes ejection — wrong), a41d02ba (FRC definition — wrong), 2f96aa7a (Laplace formula — undefined variable). These require content expert review and answer key correction before any use.
Priority 2 — Fix pass (next content sprint)
Fix all "All of the above" keyed-correct items (212a0f1e, c19494ab, 77f9a4f7) by restructuring as single-best-answer questions. Fix incomplete calculation stems (10ff8622) by adding missing data. Fix stem fragments and misleading preambles (ff737f36, b79ffef6, 3091d670, d8ee064a). Fix metadata miscalibration for Bloom's and difficulty tags (e9547e9d, caaccf7c, ed043a2b, 728f8bd5, 8e39b7b5, e0e0d2c4).
Priority 3 — Curation pass (systematic thinning)
Disable the large cluster of untagged, zero-template-membership, Bloom's-1 recall items with no PYQ provenance and no clinical framing. This is the highest-volume action and will have the largest impact on pool quality. The threshold for disable in this category should be: Bloom's-1 + no PYQ tag + no template membership + concept already covered by a better-framed item in the pool. Representative IDs across shards: 76d67563, 537908f8, dfe3f508, 64af93b0, 0a35bed1, 523b8960, cfc7f52c, 5f22f55c, 0c0db53b, a4635baa, 27076298, 89090fc1, b943b73f, ed00a10c, ea3ea83f, 2f33de01, aaffc563, f6ad138b, 31bc419e, 8eb340dd, 977b9729, 48637953, b13f4930, 4f5e0b6a, 0283a610.
Priority 4 — Topic audit (cross-cutting)
Conduct a systematic topic-tag audit to identify and reclassify questions that belong to Biochemistry, Pathology, Surgery, or Blood Banking (46553bd2, c24d5051, e03ef7ba, 48637953, 9332b1f3, 1d7b6246, 4fd152c3). Separately, audit for within-Physiology topic misfiling (b1038f1c, 9ef63e7a, 763dd469).
Priority 5 — Development signal (medium term)
Use the gap analysis from Issue Category 7 to commission new Bloom's-3/4 questions in reproductive physiology, exercise physiology, sensory physiology, and autonomic neurophysiology. Use the existing high-quality items (a7c9d255, 660698e2, c5ca219b, 9332a0cf, 42a9dd21, 428cbdc2) as templates.
Example Keep / Fix / Disable Calls
The following table summarizes representative disposition calls from across the reviewed set, organized to illustrate the reasoning applied in each category.
KEEP — No action required
| Question ID | Topic | Reason |
|---|---|---|
| 660698e2 | Integrative Physiology | Bloom's-4 multi-system scenario (DI + RAAS + hypernatremia); INICET/NEET-PG tagged; matches benchmark quality; consider elevating to benchmark pool |
| a7c9d255 | Acid-Base Balance | Bloom's-3 ABG interpretation with numerical values; clean stem, unambiguous answer, no structural flaws |
| c5ca219b | Cardiovascular | Bloom's-3 Starling forces calculation; all four variables given in stem; correct arithmetic; good distractors |
| fbb7b4a4 | Renal Physiology | Bloom's-3 net filtration pressure calculation; INICET PYQ-tagged; matches benchmark style of e9f16e48 |
| 42a9dd21 | Renal/Integrative | Bloom's-4 clinical vignette (marathon hyponatremia); clean mechanism question; aligns with benchmark standard |
| 428cbdc2 | Cellular Physiology | Bloom's-4 equilibrium potential calculation; genuine quantitative reasoning; well-constructed distractors |
| faad35aa | Cardiovascular | Bloom's-3 ECG axis determination; NEET-PG tagged; requires actual knowledge of quadrant method |
| 9332a0cf | Reproductive Endocrinology | Bloom's-2 pathway sequencing (estrogen synthesis two-cell model); well-constructed distractors |
| 53e8189f | Nerve and Muscle | Bloom's-2 NOT question on biphasic action potential; NEET-PG 2023 PYQ; requires conceptual understanding |
| 222d445b | Nerve and Muscle | Bloom's-2 smooth muscle contraction mechanism; NEET-PG PYQ; meaningful mechanistic distinction |
| 00437efb | Endocrinology | Bloom's-3 vignette (intracellular receptor for stress hormone); clean stem; unambiguous correct answer |
| b3f59d2c | Endocrinology | Bloom's-2 Wolff-Chaikoff effect applied to pre-surgical iodide; named mechanism → clinical application |
| 7a743c68 | Sensory Physiology | Bloom's-2 auditory hair cell K⁺ influx mechanism; non-trivial concept; well-constructed distractors |
| afce73ad | Neurophysiology | Bloom's-2 prostaglandins as sensitizers vs. direct excitors; NEET-PG tagged; clean discrimination |
| b5baa8e1 | Cardiovascular | Bloom's-2 phonocardiogram S₂ phase identification; PYQ-tagged; integrates cardiac cycle with auscultation |
FIX — Usable after targeted correction
| Question ID | Issue | Required Fix |
|---|---|---|
| b89719bf | Factual error: ANP keyed as increasing blood volume | Correct answer key to "Decrease in renin secretion"; verify all distractors |
| e48be40f | Factual error: HCO₃⁻ keyed as predominant extracellular ion | Correct answer key to Na⁺ (cation) or Cl⁻ (anion); reframe question |
| a41d02ba | Factual error: ERV definition keyed as FRC | Swap correct answer and first distractor; verify all options |
| bbe52d08 | Sequence error: IVR keyed as preceding ejection | Correct the cardiac cycle sequence; fix answer key |
| 2f96aa7a | Undefined variable in Laplace formula (T = WP/R) | Replace with standard formula T = Pr; rewrite stem to specify physiological context |
| 212a0f1e | "All of the above" keyed correct (Fick method) | Restructure as single-best-answer; identify one most-tested component as correct answer |
| c19494ab | "All of the above" keyed correct (gastric acid secretion) | Restructure as single-best-answer; fix imprecise H2 blocker option |
| 10ff8622 | Incomplete stem: cardiac output calculation missing stroke volume | Add "stroke volume = 75 mL" to stem; converts to valid Bloom's-3 calculation item |
| ff737f36 | Stem is a fragment, not a question | Rewrite as complete question: "Which hormone links nutritional status with onset of puberty?" |
| b79ffef6 | Misleading preamble obscures actual question | Remove first sentence; rewrite stem as "Which GI secretion has the highest potassium concentration?" |
| d8ee064a | "All of the above" distractor with internally contradictory options (leptin) | Remove "All of the above" distractor; replace with a plausible incorrect leptin function |
| a8c28951 | Duplicate option text in NMJ question | Replace one duplicate option with a distinct distractor |
| e9547e9d | Bloom's-5 tag on a pure recall item (bitter taste localization) | Reclassify to Bloom's-1; no content change needed |
| caaccf7c | String difficulty label; distractor states GH decreases IGF-1 (wrong) | Normalize difficulty to numeric; fix distractor to "GH increases IGF-1 and somatostatin" |
| b1038f1c | Topic misfiled under Acid-Base Balance (content is autonomic/CVS) | Refile to Cardiovascular System or Neurophysiology — Autonomic; no content change |
| 3091d670 | Ambiguous stem: methemoglobinemia and oxygen content | Reframe to specify pulse oximetry context, or restructure with anemia as the "except" |
| 0227ade9 | Distractor too close to correct (gastric emptying regulation) | Add "primary/predominant" qualifier to stem; review answer key against Guyton |
| cf7b0860 | Mechanistic conflation: central chemoreceptor stimulus | Rewrite stem to specify "direct chemical stimulus" and key CSF H⁺, or specify "indirect" and key CO₂ |
| 45ba19bf | Distractors too obviously wrong (pregnancy respiratory changes) | Replace weak distractors with clinically confused alternatives (FRC vs. RV, TV vs. RR) |
DISABLE — Remove from active pool
| Question ID | Topic | Reason |
|---|---|---|
| 5f52e05d | Respiratory | Duplicate options + factual inversion + incoherent premise; non-functional |
| d2c9c9f5 | Blood | Stem asks about organs; options are time periods; fundamental stem-option mismatch |
| 15e1ff52 | Cardiovascular | True/False/Partially true format; unacceptable for PG MCQ standards |
| 9ef63e7a | Cardiovascular (misfiled) | Immunology/cell biology question; contestable key (NGF for memory cell survival); wrong subject |
| 46553bd2 | Endocrinology (misfiled) | Biochemistry fact (selenocysteine in deiodinase); no physiological reasoning; no tags |
| 4fd152c3 | Blood (misfiled) | Blood banking question; not core Physiology; contestable answer framing |
| 1d7b6246 | Blood (misfiled) | Well-constructed Bloom's-4 item but belongs to Pathology/Surgery; disable from Physiology bank |
| e03ef7ba | General Physiology (misfiled) | Microbiology taxonomy; factually imprecise; wrong subject |
| 48637953 | GI (misfiled) | Clinical medicine/surgery fact (ZE syndrome triad); not a physiology concept |
| 9332b1f3 | Altitude/Diving (misfiled) | Clinical management question (pneumothorax); wrong subject; "All of the above" key |
| 76d67563 | Cellular Physiology | Bloom's-1 organelle recall; no tags; no template; no clinical framing; below PG floor |
| 537908f8 | Cellular Physiology | Bloom's-1 organelle recall; identical problem to 76d67563; disable or merge |
| dfe3f508 | Renal Physiology | Bloom's-1 hormone source recall; trivially known; no discriminatory value |
| 64af93b0 | Reproductive | Bloom's-1; "LH triggers ovulation"; no tags; below minimum quality bar |
| 0a35bed1 | Endocrinology | Bloom's-1; T4 half-life number; no clinical context; no tags |
| 523b8960 | Neurophysiology | Bloom's-1; "sleep centre = hypothalamus"; no tags; trivial |
| 0c0db53b | Renal Physiology | Bloom's-1; "GFR = 125 ml/min"; no discriminatory value; disable or replace with calculation |
| a4635baa | Renal Physiology | Bloom's-1; "renin from JGA"; no tags; no template |
| 27076298 | Cardiovascular | Bloom's-1; "QRS = ventricular depolarization"; most basic ECG fact; below PG floor |
| ea3ea83f | Neurophysiology | Bloom's-1; "norepinephrine = fight-or-flight"; tagged TELEGRAM only; no PYQ lineage |
| 94d68e53 | Altitude/Diving | Physics trivia (atmospheric pressure in psi); no physiological reasoning; no PYQ lineage |
| c24d5051 | General Physiology | Biochemistry/cell biology item (peroxisomes and H₂O₂); wrong subject |
| 977b9729 | Altitude/Diving | Bloom's-1; "hypoxic hypoxia at altitude"; redundant with better benchmark items on same topic |
| 8eb340dd | Neurophysiology | Bloom's-1; CSF protein range at basal cistern; rote numerical recall; no template |
| b13f4930 | GI | Bloom's-1; gene name trivia (CLPS for colipase); no physiological reasoning; no tags |
| 4f5e0b6a | Renal Physiology | Bloom's-1; neonatal concentrating ability by age; obscure developmental fact; debatable answer |
| 31bc419e | Reproductive | Bloom's-1; sperm velocity in mm/min; numerical trivia; no mechanistic relevance |
| aaffc563 | Neurophysiology | Bloom's-1; "nightmares in REM sleep"; no clinical context; disable or upgrade |
| f6ad138b | Cardiovascular | Bloom's-1; upper/lower limb BP difference = 5 mm; isolated numerical fact; no application |
| ed00a10c | Neurophysiology | Bloom's-1; "hunger centre = hypothalamus"; "stria nigra" is not a real anatomical term |