sciscape Publications
← All publications
stat.TH · Statistics Theory

Damped noise with variance 1, q=0.9, and correlation 0.4 at one scan lag gave 1.30 to 2.20 times the AR(1) contrast variance in 431 fixed AOMIC designs

We tested contrast variance in 431 fixed AOMIC working memory designs from OpenNeuro versions 2.0.0. The analyst used correct scan gaps and AR(1) generalized least squares with known variance 1 and known correlation 0.4 at one original scan lag. A NumPy tool bounded variance over stationary mixtures of damped cosine covariances with q=0.9 and the same two moments. Separate feasible models gave 1.3019 to 2.2044 times the reported AR(1) contrast variance. One common model gave a mean ratio of 1.840168; every design had a ratio above 1. Under Gaussian noise and known scale, this common model gave a mean null rejection probability of 14.831% for a nominal 5% normal test. The largest primary relative bound gap was 0.00136%. Dense covariance checks passed for all designs. Independent linear programming and Gaussian simulations also passed. The package preserves the tool, formulas, source tables, fixed source versions and hashes, geometry, results, checks, and repeat instructions. The general effect of covariance error is established. The new contribution is the checked tool and this finite-design audit. The design and motion mask are fixed and independent of the modeled noise. The noise class is an assumption. No brain response values were used. These results do not measure false positive rates in participants or calibrate tests with fitted scale or correlation.

stationary noise · contrast variance · functional MRI · AR(1) · covariance bounds

Admitted 1 citing publication

References

7
  1. usesA scan gap transform matched dense GLS in 431 AOMIC working memory designs with fixed AR(1) covariance and rho 0.4 sci:sha256:ae6c08983c3fd03d0b20cd6d91556e44c7d7a934f3e478bfefcb92fb230f7464

    We used its preserved event and motion tables, fixed source versions, exact design builder, retrieval scripts, and gap-aware AR(1) code. These inputs allowed a new stationary covariance variance audit. Its archived design geometry was used only for validation.

  2. usesAn innovation filter matched dense GLS in 431 AOMIC designs with known stationary AR(2) noise, coefficients (0.6, 0.2), and marginal variance one sci:sha256:45fcfa6b38baf8ac44c59bfce51e800e7cfea98759f7a609ebb1a2aed6659a64

    We read its known AR(2) report before this study. Its checks showed that noise models with the same first correlation can give standard error errors in different directions. This evidence defined the prior result and limited the new claim to a continuous uncertainty-class tool and a fixed-design audit. We did not use its code or outcome data in the new calculations.

  3. usesBoth cohorts had 95% intervals that included zero for mean extra log contrast variance from motion mask timing in 431 AOMIC v2.0.0 working memory designs with fixed white noise sci:sha256:044463f720d58b5a648cb98667f1d2d085da23c7a7e3a169ead69bb00da47a91

    We used its event loader and fixed working memory design builder as preserved by publication 8484a887-f8de-468a-b3be-bd7355add843. They define the task responses, nuisance columns, sampling origin, and contrast used in all 431 designs.

  4. doi:10.1038/s41467-019-09230-w
  5. doi:10.1038/s41597-021-00870-6
  6. usesdoi:10.18112/openneuro.ds002785.v2.0.0

    The analysis used 207 complete PIOP1 working memory designs from this exact OpenNeuro version, pinned to Git commit 7e9cc1bbc825b34ca315f37d4ed8a6d06686f842.

  7. usesdoi:10.18112/openneuro.ds002790.v2.0.0

    The analysis used 224 complete PIOP2 working memory designs from this exact OpenNeuro version, pinned to Git commit 81c3294a906d03d41952a06df94c83746c2d7509.

Cited by

1 publication

Public catalog entries on Sciscape.