@prefix this: . @prefix sub: . @prefix np: . @prefix rdf: . @prefix prov: . @prefix npx: . @prefix dc: . @prefix xsd: . sub:Head { this: a np:Nanopublication; np:hasAssertion sub:assertion; np:hasProvenance sub:provenance; np:hasPublicationInfo sub:pubinfo . } sub:assertion { a ; dc:description "The download and derivation are shared with claim ch03-18 (../thompson1917-ch03-18/fetch.py; fetch.py here delegates to it): per species a preliminary OLS of log10 W on log10 L drops 258 fish with |residual| > 0.5 (pre-registered), then fish are binned per species x quarter x 1-cm length class with n, sum and sum of squares of log10 W (derived/ns_ibts_lw_bins.csv). analysis.py keeps taxa with >= 500 fish and a 97.5th/2.5th percentile length ratio >= 3 (20 taxa: 12 teleosts, 5 elasmobranchs, 3 scyphozoan jellyfish), fits log10 W = a + b1 x + b2 x^2 (x = log10 L) by weighted least squares on class means (exact individual-level OLS since all fish in a class share L), tests it against the cube law (b1 = 3, b2 = 0) by F-test, and computes the ratio of fitted k = W/L^3 between its maximum and minimum over the 2.5-97.5 percentile length range; 'variable k' = P < 0.001 and ratio >= 1.10. Within-quarter fits and a fish-only count are reported as sensitivity checks; the plaice k profile by 1-cm class is the pre-registered secondary."; "Measured: 0.7 fraction of eligible taxa (14 of 20) in which k = W/L^3 varies by >= 10% across the growing range and the cube law is rejected at P < 0.001 (n = 20, exactly at the pre-registered 70% threshold; fish only (17 taxa, jellyfish excluded): 11/17 = 0.647; within-quarter fits: 21 of 30 variable). Interpretation: The first half of the claim (W = k L^3 for an animal of fixed form and density) is geometry and is not tested. The empirical content is Thompson's expectation that k 'is by no means constant' while growth continues, because form keeps changing; the 10% threshold is the amount of k variation that would be clearly visible on his Fig. 21 (plaice k rises about 15% from small fish to the 30-cm maximum). Changes of k can also come from density, gut fullness or gonad state rather than shape (assessors' caveat); the test cannot separate these, so a supporting result shows that k varies, not that form is the reason. Pooling years and quarters mixes seasonal condition cycles into the k-vs-length relation; the analysis therefore also reports results within each quarter as a robustness check (pre-registered, not part of the verdict). Length conversion: LngtCode '1' means LngtClass in cm; '.' or '0' means mm. A single probe request (NS-IBTS 2019 Q1 CA, HTTP status and first record only) was made before this pre-registration to confirm that the web service is reachable."; "Claim test of thompson1917-ch02-01: supports, confidence 0.55"; ; prov:wasAttributedTo ; "Planned datasets: ices-datras-ns-ibts-ca-hh"; "claim-test/0.1.0"; "biological"; "Pre-registered (claim refuted). At most 30% of eligible species have variable k (k is effectively constant across the growing range in most fishes, contrary to Thompson's expectation). Between 30% and 70% is 'mixed'. If fewer than 10 species are eligible the verdict is 'inconclusive' (data-unavailable)."; "A varying k is consistent with, but does not demonstrate, a change of form: condition, gut fullness, gonad state and ontogenetic changes in density or fat content also change W at fixed L (assessors' caveat). The test therefore addresses Thompson's empirical expectation that k will be found to vary, not his inference about form.", "Eutrigla appears twice (species-level and genus-level records); neither set has variable k (the genus-level set is the only taxon in which the cube law is not even rejected), so merging them would give 14/19 = 0.74.", "Lengths are 1-cm (or coarser) classes, weights are from survey scales at sea (whole-gram or coarser), sexes and years are pooled, and fish are treated as independent; the F-test P-values are therefore nominal. The 10% k-range threshold, not the P-value, is what separates taxa here.", "The result sits exactly on the pre-registered threshold (14/20 = 70.0%), so the verdict is fragile: excluding the three jellyfish taxa (bell diameter is not a fish length, though they are 'other growing animals' within the claim's scope) gives 11/17 = 65% and the verdict would be 'mixed'. The robust findings are weaker than the verdict label: the cube law is rejected statistically in 19 of 20 taxa (huge samples make any departure significant), the median exponent b is 3.07 (range 2.35-3.38), but only 9 of 20 taxa have |b - 3| > 0.1, and k varies by >= 10% across the central 95% of lengths in 14. So k is 'by no means constant' in a majority, but in a sizeable minority of fishes (plaice, dab, grey gurnard, starry ray, starry smooth-hound) it is constant to within 5-10% across a 3-fold length range.", "Thompson's specific plaice observation (k rising to a maximum near 30 cm and then declining) is not reproduced: in 54,375 North Sea plaice k is lowest at 5-10 cm (0.90) and at 20-25 cm (0.91) and rises monotonically to 1.01 at 45-50 cm, with no decline at large sizes; the overall exponent is 3.02 and k varies by only 5% between 13.5 and 40.5 cm. His 1909 data were from one locality and season; the pooled survey sample mixes seasons (Q1 spawning condition is about 10% lower than Q3, see claim ch03-18) and areas, which could flatten or shift a peak. Within-quarter fits were still 'variable k' in 21 of 30 species-quarter combinations."; "Pre-registered (claim supported). In ICES NS-IBTS CA records (Q1 and Q3, 2015-2024, individual weights, CANoAtLngt = 1), for every species with at least 500 weighed fish whose 97.5th-percentile length is at least 3 times its 2.5th-percentile length, fit log10 W = a + b1 log10 L + b2 (log10 L)^2 after dropping fish with |residual| > 0.5 log10 units from a preliminary linear fit (pre-registered cleaning). A species has 'variable k' if (i) an F-test rejects the cube law (model with b1 = 3, b2 = 0) at P < 0.001 and (ii) the fitted k = W / L^3 varies by at least 10% (max/min >= 1.10) across the species' 2.5th-97.5th percentile length range. The claim is supported if at least 70% of eligible species have variable k. Secondary, descriptive: in plaice, the 1-cm-bin mean of k is maximal somewhere between 20 and 40 cm and is lower at both smaller and larger lengths."; "Pre-registered (claim refuted). At most 30% of eligible species have variable k (k is effectively constant across the growing range in most fishes, contrary to Thompson's expectation). Between 30% and 70% is 'mixed'. If fewer than 10 species are eligible the verdict is 'inconclusive' (data-unavailable)."; "0.55"^^xsd:double; ; "supports"; "Weight scales as the cube of length (W = k L^3) only while form and density are unchanged, and Thompson expects form to keep changing as long as growth continues; so in real growing fishes k = W / L^3 should NOT be constant across the size range: it should vary systematically with length in most species (in plaice, by his Fig. 21, rising to a maximum near 30 cm and then declining)."; "The download and derivation are shared with claim ch03-18 (../thompson1917-ch03-18/fetch.py; fetch.py here delegates to it): per species a preliminary OLS of log10 W on log10 L drops 258 fish with |residual| > 0.5 (pre-registered), then fish are binned per species x quarter x 1-cm length class with n, sum and sum of squares of log10 W (derived/ns_ibts_lw_bins.csv). analysis.py keeps taxa with >= 500 fish and a 97.5th/2.5th percentile length ratio >= 3 (20 taxa: 12 teleosts, 5 elasmobranchs, 3 scyphozoan jellyfish), fits log10 W = a + b1 x + b2 x^2 (x = log10 L) by weighted least squares on class means (exact individual-level OLS since all fish in a class share L), tests it against the cube law (b1 = 3, b2 = 0) by F-test, and computes the ratio of fitted k = W/L^3 between its maximum and minimum over the 2.5-97.5 percentile length range; 'variable k' = P < 0.001 and ratio >= 1.10. Within-quarter fits and a fish-only count are reported as sensitivity checks; the plaice k profile by 1-cm class is the pre-registered secondary."; "Pre-registered (claim supported). In ICES NS-IBTS CA records (Q1 and Q3, 2015-2024, individual weights, CANoAtLngt = 1), for every species with at least 500 weighed fish whose 97.5th-percentile length is at least 3 times its 2.5th-percentile length, fit log10 W = a + b1 log10 L + b2 (log10 L)^2 after dropping fish with |residual| > 0.5 log10 units from a preliminary linear fit (pre-registered cleaning). A species has 'variable k' if (i) an F-test rejects the cube law (model with b1 = 3, b2 = 0) at P < 0.001 and (ii) the fitted k = W / L^3 varies by at least 10% (max/min >= 1.10) across the species' 2.5th-97.5th percentile length range. The claim is supported if at least 70% of eligible species have variable k. Secondary, descriptive: in plaice, the 1-cm-bin mean of k is maximal somewhere between 20 and 40 cm and is lower at both smaller and larger lengths."; "false"^^xsd:boolean . . a ; dc:description "individual length and weight of 520,328 single fish (CANoAtLngt = 1, IndWgt > 0) of all species"; dc:license "ICES Data Policy: public, CC BY 4.0 with attribution to ICES and the national institutes"; "Dataset: ICES DATRAS, North Sea International Bottom Trawl Survey (NS-IBTS): CA (age/weight/maturity) and HH (haul) exchange records, Q1 and Q3, 2015-2024"; prov:atLocation , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , ; prov:wasDerivedFrom ; "sha256:0220be5227334b5ac1e849b7237cef375edc15a8d3b67d56689158f1313db4bc", "sha256:04e3ae57cc935865c5616fb14c73766d771cc48d90012a9272a85fe8f3470b61", "sha256:13c16a8d51338df6c745680c3d056ec00212b13f64b635e2e6a2b65a874b01bb", "sha256:15a02698aca785afac51f11a363abd8e81b0b8ba30bbdc2010f38093f3804063", "sha256:18389cf4b732c6bb47f66f69c587d1dcc0e8b74638bfc1b66bbbf1cc71937042", "sha256:1e3d5111411f54279509cf3987d090a5f70dfbd9220d562211dc189a7d43e273", "sha256:3833edfea643ba4f7702ec0dfc3e6ea35a64ae6f118c58202ad365e1c0ee908f", "sha256:4676ed65c4b6e8bdeb59b8edd2e02ed4f0beff6c7c40ea12fe3aba42f5926eb8", "sha256:49d58d2db82d76ae834421faaa3c718f38f97006563c116303fb3024c53d4719", "sha256:51a37676e9932249109931ef1c640d6f55344eeeb1a9a6e804b710ff957e78cf", "sha256:547b234d5f5790f63d83eb11c59405a054dd93249c05d9b72d1b9b9f3e04f568", "sha256:56c5dd3cab313748205e7e9c52615f5084553b7728a26e3f7e40338fbed1e2c9", "sha256:7276babf2009ec5f0c574c577c947288c2f0c3f4eedcad5f5d18a5d014b7fba3", "sha256:7ed23f41de37b6403f851b838795d041c955e97b9805efa0f8d39f1f77e8acc0", "sha256:838187ea28d0c581e6971d5cf344d68c83c88bba9deee8275dc4edb119613259", "sha256:86b81c99343dccce1048e22226dc6fd12ddc309a348dcad3a18cb2c7d0ed63dd", "sha256:97031868e87ff39a7459b333007054fa021738a86b8f6985f7eed3123dccc2af", "sha256:972a36186026f1d89115d4930fbabc134733b73c81f554d3935dd625fbcd950d", "sha256:9ee3cba07264172f6b1fc64ae2aabe2946f3d113cd0383baf54d5082118f219d", "sha256:a0fe6d95774ca55452ee7828a36bb87602425b117050e75bd26e93c83415d591", "sha256:a945a0e63bafb7b6592ddf73c4513a61c8c4e923f54e80bba34052c5be9ae887", "sha256:aa78f3bb12a46533b8b9c5fc8346e7c0bf87f5f4c685a491197af9eb897c1b19", "sha256:b2e064d2e8fa77c5e4df9f8fe7a838f7f2c417eb759fdd0afc54127c07e0046c", "sha256:b6d6ac629eb4d3493c7328764027c12b56c20eb0884278b7a6994f31cabf65cc", "sha256:bd82ae1f7612973a5d5b29ad54a87091a0470e809be24b943a717cd1ab4ce217", "sha256:c0831ca122a9233cf82360ae258b75812c47d82b86f422eedfc31403948998e9", "sha256:c0b42e2a1af6d8e7def78b232f047a428c896b76ffc914b5bcd9acbab4f4af02", "sha256:cb32f780e6b4f3443c1480d9266aa734cad49f2c1de16dc649e1fd5c7cd302ca", "sha256:cffa346d1a6457864ab55da80b6ea5c8c054f6e4e0fcb2acba57719b2b8318ae", "sha256:d0c30d13a10aaeb03f3cd30c76e5a8a283a2c969623ace73f657d1e2c774bab1", "sha256:d2bbff10ab17df6b29b84e1ea896e441f61ec774bf91fbe2a90e66211ba45ba9", "sha256:d5f37b83c495a00b6307073a01770633c615a771e395e4b4385117ae89dd91c9", "sha256:dbd475737936f5fd6ddf36d560dcab2ca50326df5d5bba51fdd61fa0b33b4efe", "sha256:e75067656ec5d16abd34cf435442046bf3eebd98625cbc5e18ac3a5d2d3475ab", "sha256:e85e1ec7bcfebc655992ddf11f34dc9deb361526bb381804ff890988746a1b69", "sha256:ec162445b4b028ee4123ee083b693c98ebb99397c2d3ff9c237b250b9e9ea27a", "sha256:ef2a0e55d83efa4b19e8aec725892a854e2835176695e7d68c64d869028bace4", "sha256:f3e8aec1df4a3d32d9da1742be8b29a8d644cd4b37e7714071712e5cbc3f1983", "sha256:f3f4919255fcf0f593e542a20a9e015a52fbb4dd7b72a72031343a1798e3aa9a", "sha256:f9363d745de26cf6f602e6495eea077ee9a1799b8d04b5e393ffad58b7fe34b6" . } sub:provenance { a prov:SoftwareAgent; "delphisci claim tester (claude-fable-5-1-thinking-high; cursor/thompson-claim-tests-6e84)"; prov:actedOnBehalfOf ; . a ; dc:identifier "claude-fable-5-1-thinking-high" . sub:assertion prov:wasAttributedTo ; prov:wasDerivedFrom ; prov:wasGeneratedBy sub:test . sub:plan prov:hadPlan . sub:test a prov:Activity; prov:qualifiedAssociation sub:plan; prov:wasAssociatedWith . } sub:pubinfo { this: dc:created "2026-11-08T16:24:31.889Z"^^xsd:dateTime; dc:creator ; dc:license ; npx:hasNanopubType ; npx:introduces ; "Claim test of thompson1917-ch02-01: supports, confidence 0.55"; . sub:sig npx:hasAlgorithm "RSA"; npx:hasPublicKey "MIIBIjANBgkqhkiG9w0BAQEFAAOCAQ8AMIIBCgKCAQEAyFHtPB6p0mCjDozxlOY9CmO3nBm7Te3jFbXyLmM+G1tGe47Da0jYIbhkl9Oo1ZjxmCa3LCwlEqK8gj6hEIcL4ahW30JXh9HycfKVk1bugkzMEV15tJdTzAU3sGMFnbNsn43xY2LFLy45tApkRm5byEiM7oRRrooIdwoQKnP4Wzq/bfAoJ9qvW+MGkelyOTd1jbSt+KfmT7xKmDfWzAHddUlWwMX2QJOLxDaffOteTop6ZRgPiNmPJKA/S62Gac9XtSYT3oDkZ6/nr5cxxNLpOyxTf5FPISiCe11xYO/unYjY+mwnOfA8yhJIyx1l0O9I0vpOpn+E5TvyHJV+oSTrvwIDAQAB"; npx:hasSignature "ibLkffnBXeEbYM+nfo+UTZShO+k0rlhhex19TalkW5SnslDNYFuDD87CyQFkFcT8VGjUIDmEDjs1enF2vxXfXZokIZFWKRzdVwGu4Y3c+Yj2DvL1XeQIsKrCuXtYLupsOzBUSyd6v59Apq2g5tmmC+qhAML4UJCjDziY710UaIx8C8phLv+VDZFMruYN31CKkH+Dihf1xKaa+xoByXdZ3naanMoHhkIO4NBBMAJFPxRBQXgS39+Jx8KV3OVtV5KYoxkpr9gS49s0HVn7DyWtyVqniS66oJ4K3AzWQGHUSjJWN609pPi94NbITI1kBfK6b+HxEvpaSEm4DAVrMEv6+w=="; npx:hasSignatureTarget this:; npx:signedBy . }