Class: Analysis

A distinct reported inferential map or tested effect, linking its source acquisitions, sample, paradigm context, tested effect, preprocessing, model estimation, inference settings, and source result tables.

Create one Analysis for each distinct tested effect. Separate Analyses are required when normalized direction, participant-group comparison, method, the pattern of effect.cells, seed or connectivity model, decoded variable, component identity, or spatial_scope differ, and when an omnibus effect and its directional post-hoc contrast are both reported. ROI definitions, acquisition summaries, descriptive sample tables, and masks without inferential effects are not Analyses. Nor is a map reported with no inferential test at all -- an ICA or PCA component map presented descriptively, a connectivity matrix given without a test -- there being nothing for such a map to be the effect of.

URI: neurostudy:Analysis

 classDiagram
    class Analysis
    click Analysis href "../Analysis/"
      Analysis : acquisitions





        Analysis --> "*" Acquisition : acquisitions
        click Acquisition href "../Acquisition/"



      Analysis : assessments





        Analysis --> "*" Assessment : assessments
        click Assessment href "../Assessment/"



      Analysis : coordinate_space

      Analysis : defines_regions





        Analysis --> "*" Region : defines_regions
        click Region href "../Region/"



      Analysis : definition

      Analysis : details





        Analysis --> "1" AnalysisDetails : details
        click AnalysisDetails href "../AnalysisDetails/"



      Analysis : effect





        Analysis --> "1" Effect : effect
        click Effect href "../Effect/"



      Analysis : groups





        Analysis --> "1..*" AnalysisGroup : groups
        click AnalysisGroup href "../AnalysisGroup/"



      Analysis : id

      Analysis : inference_settings





        Analysis --> "0..1" InferenceSettings : inference_settings
        click InferenceSettings href "../InferenceSettings/"



      Analysis : interpretations

      Analysis : measure





        Analysis --> "1" Measure : measure
        click Measure href "../Measure/"



      Analysis : mirror_of





        Analysis --> "0..1" Analysis : mirror_of
        click Analysis href "../Analysis/"



      Analysis : model_estimation





        Analysis --> "1" ModelEstimation : model_estimation
        click ModelEstimation href "../ModelEstimation/"



      Analysis : model_representation_notes

      Analysis : name

      Analysis : outcome





        Analysis --> "0..1" AnalysisOutcome : outcome
        click AnalysisOutcome href "../AnalysisOutcome/"



      Analysis : prespecification

      Analysis : regions





        Analysis --> "*" Region : regions
        click Region href "../Region/"



      Analysis : source_table_analysis

      Analysis : spatial_scope

      Analysis : statistical_maps





        Analysis --> "*" StatisticalMap : statistical_maps
        click StatisticalMap href "../StatisticalMap/"



      Analysis : tables





        Analysis --> "*" Table : tables
        click Table href "../Table/"



      Analysis : tasks





        Analysis --> "*" Task : tasks
        click Task href "../Task/"

Slots

Name	Cardinality and Range	Description	Inheritance
id	1 String		direct
name	1 String	Study-local label for the analysis or inferential result, in the source's wor...	direct
definition	1 String	Precise statement of the tested effect in the source's own words, minimally c...	direct
prespecification	1 String or  Prespecification	Whether this contrast was planned before the data were seen or arrived at aft...	direct
interpretations	* String	The source's own statement of what this analysis found and what it means, typ...	direct
outcome	0..1 AnalysisOutcome	Whether anything survived the threshold this analysis was tested at: a cluste...	direct
spatial_scope	1 String or  SpatialScope	Spatial domain of the analysis modelled	direct
regions	* Region	The regions the analysis ran over -- its search space, what `spatial_scope: r...	direct
defines_regions	* Region	Regions this analysis produced: a localizer's clusters, or peaks of a contras...	direct
measure	1 Measure	The scientific quantity of the map, by identifier, independent of the statist...	direct
groups	1..* AnalysisGroup	The participant cohorts the analysis ran on, and how many participants each c...	direct
effect	1 Effect	What was tested about the map	direct
details	1 AnalysisDetails	The method that produced the map and its method-specific payload	direct
acquisitions	* Acquisition	Acquisition protocols supplying data to this analysis	direct
tasks	* Task	The participant paradigms that contextualize this analysis	direct
model_estimation	1 ModelEstimation	The model this analysis's contrast was taken from: its family, estimator, sta...	direct
inference_settings	0..1 InferenceSettings	The thresholding and correction this analysis was reported under, by identifi...	direct
tables	* Table	The source tables reporting results for this analysis	direct
mirror_of	0..1 Analysis	Present only on an analysis the builder derived by reversing another one, and...	direct
source_table_analysis	0..1 String	The key of the coordinate-parse analysis this analysis's results were read fr...	direct
statistical_maps	* StatisticalMap	Statistical maps made available for this analysis	direct
coordinate_space	0..1 String	Coordinate reference space or template the spatial results are reported in, s...	direct
assessments	* Assessment	Assessments whose measurements were used in or selected the sample for this a...	direct
model_representation_notes	0..1 String	Model components that could not be represented or were represented only appro...	direct

Usages

used by	used in	type	used
Analysis	mirror_of	range	Analysis
CoordinateSet	analysis	range	Analysis
Study	analyses	range	Analysis

Rules

Rule Applied	Preconditions	Postconditions	Elseconditions
slot_conditions	{'spatial_scope': {'any_of': [{'equals_string': 'whole_brain'}, {'equals_string': 'searchlight'}]}}	{'regions': {'value_presence': 'ABSENT'}}

Rule Applied	Preconditions	Postconditions	Elseconditions
slot_conditions	{'spatial_scope': {'equals_string': 'roi'}}	{'regions': {'value_presence': 'PRESENT'}}

Identifier and Mapping Information

Schema Source

• from schema: https://neurostuff.org/schemas/neuroimaging-study-storage

Mappings

Mapping Type	Mapped Value
self	neurostudy:Analysis
native	neurostudy:Analysis

LinkML Source

Direct

name: Analysis
description: 'A distinct reported inferential map or tested effect, linking its source
  acquisitions, sample, paradigm context, tested effect, preprocessing, model estimation,
  inference settings, and source result tables.

  Create one Analysis for each distinct tested effect. Separate Analyses are required
  when normalized direction, participant-group comparison, method, the pattern of
  `effect.cells`, seed or connectivity model, decoded variable, component identity,
  or spatial_scope differ, and when an omnibus effect and its directional post-hoc
  contrast are both reported. ROI definitions, acquisition summaries, descriptive
  sample tables, and masks without inferential effects are not Analyses. Nor is a
  map reported with no inferential test at all -- an ICA or PCA component map presented
  descriptively, a connectivity matrix given without a test -- there being nothing
  for such a map to be the effect of.'
from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage
rank: 1000
attributes:
  id:
    name: id
    in_subset:
    - mvp
    - deterministic
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    identifier: true
    domain_of:
    - Arm
    - Timepoint
    - Group
    - Task
    - Condition
    - Assessment
    - Region
    - Acquisition
    - Device
    - Analysis
    - Measure
    - Preprocessing
    - ModelEstimation
    - ModelTerm
    - InferenceSettings
    - StatisticalMap
    - Table
    - CoordinateSet
    - Study
    range: string
    required: true
  name:
    name: name
    description: Study-local label for the analysis or inferential result, in the
      source's words.
    in_subset:
    - mvp
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    domain_of:
    - Arm
    - Timepoint
    - Group
    - Task
    - Condition
    - Assessment
    - Region
    - Analysis
    - ModelTerm
    - ExternalDataset
    - PerformanceMetric
    range: string
    required: true
  definition:
    name: definition
    description: Precise statement of the tested effect in the source's own words,
      minimally cleaned. Always source-grounded. For an interaction it must state
      the direction of the interaction as the source states it; for an ordered contrast
      it must keep the ordering that the cell directions cannot express.
    in_subset:
    - mvp
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    domain_of:
    - Analysis
    - ConjunctionComponent
    range: string
    required: true
  prespecification:
    name: prespecification
    description: 'Whether this contrast was planned before the data were seen or arrived
      at afterwards.

      Open vocabulary: write the source''s own wording in the field when neither value
      fits. The two are not a spectrum and the words a source uses for them are many
      -- `post-hoc`, `confirmatory`, `planned`, `hypothesis-driven` -- so a closed
      range would discard the whole analysis over a synonym. Normalization maps the
      wording onto the two values.'
    in_subset:
    - mvp
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    domain_of:
    - Analysis
    required: true
    any_of:
    - range: Prespecification
    - range: string
  interpretations:
    name: interpretations
    description: 'The source''s own statement of what this analysis found and what
      it means, typically in Results and Discussion, verbatim or minimally cleaned.
      Always source-grounded, and the counterpart of `definition`: that field is the
      Methods-side statement of what was tested, this one is the Results/Discussion-side
      statement of what came of it.'
    in_subset:
    - mvp
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    domain_of:
    - Analysis
    range: string
    multivalued: true
  outcome:
    name: outcome
    description: 'Whether anything survived the threshold this analysis was tested
      at: a cluster, a peak, a significant region. The answer to "did this contrast
      find anything", which `Effect.cells` cannot give -- a cell''s direction is the
      sign of the contrast that was tested, and is written the same whether the test
      found an effect or not. Meta-analyses exclude null effects and pool only analyses
      that reported foci, so a tested-and-null contrast and a reported effect must
      not extract to the same record. Applies only to an analysis whose coordinate
      set is a `result` (or has no role yet): an ROI, seed, reference or other set
      was not tested here, so it has no outcome.'
    in_subset:
    - mvp
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    domain_of:
    - Analysis
    range: AnalysisOutcome
  spatial_scope:
    name: spatial_scope
    description: Spatial domain of the analysis modelled. The spatial_scope refers
      to what voxels were modelled, not which voxels/brain regions had a correction
      applied.
    in_subset:
    - mvp
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    domain_of:
    - Analysis
    required: true
    any_of:
    - range: SpatialScope
    - range: string
  regions:
    name: regions
    description: 'The regions the analysis ran over -- its search space, what `spatial_scope:
      roi` restricts it to. Empty for whole-brain and searchlight, where the emptiness
      asserts that inference was not restricted.

      Not the analysis''s seeds: a whole-brain seed-based analysis has a seed and
      no search-space region, so putting one here says the opposite of what it did.
      Seeds are `ConnectivityDetails.seed_regions`.'
    in_subset:
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    domain_of:
    - Analysis
    - FactorLevel
    - CoordinateSet
    - Study
    range: Region
    multivalued: true
    inlined: false
  defines_regions:
    name: defines_regions
    description: 'Regions this analysis produced: a localizer''s clusters, or peaks
      of a contrast later used as seeds or ROIs. Populated on the analysis a region
      came *from*, not on those that use it, so a region whose `definition_method`
      is `same_study_analysis` and which another analysis lists in `regions` or `seed_regions`
      is a circularity the record states rather than asserts.'
    in_subset:
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    domain_of:
    - Analysis
    range: Region
    multivalued: true
    inlined: false
  measure:
    name: measure
    description: 'The scientific quantity of the map, by identifier, independent of
      the statistic used to test it. Analyses of one quantity reference one record;
      an analysis of a different quantity takes its own. Required: a paper vague about
      what it measured still takes a record, with `source_label` as given and `extraction_status:
      not_reported` elsewhere.'
    in_subset:
    - mvp
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    domain_of:
    - Analysis
    range: Measure
    required: true
    inlined: false
  groups:
    name: groups
    description: 'The participant cohorts the analysis ran on, and how many participants
      each contributed. Membership only: this slot says which people were analysed,
      never what was compared. A cohort comparison is a pair of directional `Cell`s
      on the model term whose levels name the cohorts, exactly as a condition or occasion
      comparison is.'
    in_subset:
    - mvp
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    domain_of:
    - Analysis
    - FactorLevel
    - Study
    range: AnalysisGroup
    required: true
    multivalued: true
    inlined_as_list: true
  effect:
    name: effect
    description: What was tested about the map. Universal across every method.
    in_subset:
    - mvp
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    domain_of:
    - Analysis
    range: Effect
    required: true
    inlined: true
  details:
    name: details
    description: The method that produced the map and its method-specific payload.
      The populated subclass identifies the method. A voxelwise, vertexwise, or ROI-wise
      GLM, mixed-effects model, ANOVA, ANCOVA, or robust regression takes MassUnivariateDetails;
      a method outside the first-class families takes OtherAnalysisDetails and names
      itself in method_label.
    in_subset:
    - mvp
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    domain_of:
    - Analysis
    range: AnalysisDetails
    required: true
    inlined: true
  acquisitions:
    name: acquisitions
    description: Acquisition protocols supplying data to this analysis. Structural,
      diffusion, and other non-paradigm analyses are linked directly through this
      field.
    in_subset:
    - mvp
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    domain_of:
    - Task
    - Analysis
    - Study
    range: Acquisition
    multivalued: true
    inlined: false
  tasks:
    name: tasks
    description: 'The participant paradigms that contextualize this analysis. An analysis
      can span paradigms, as a cross-task conjunction or a localizer-plus-main-task
      analysis does.

      A resting-state analysis names its Task here like any other: the resting-state
      protocol is a paradigm, and a Task record exists for it. What leaves this empty
      is an analysis with no paradigm at all -- morphometry, diffusion, a static PET
      scan -- which is linked through `acquisitions` instead.'
    in_subset:
    - mvp
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    domain_of:
    - Analysis
    - Study
    range: Task
    multivalued: true
    inlined: false
  model_estimation:
    name: model_estimation
    description: 'The model this analysis''s contrast was taken from: its family,
      estimator, stage, software, and the terms it was specified with. What was selected
      out of those terms is `effect`. Required: a paper that describes a contrast
      and never names an estimator still takes a record, with the terms it does state
      and `extraction_status: not_reported` elsewhere. Analyses sharing one design
      reference one record; an analysis specified or estimated differently takes its
      own.'
    in_subset:
    - mvp
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    domain_of:
    - Analysis
    range: ModelEstimation
    required: true
    inlined: false
  inference_settings:
    name: inference_settings
    description: The thresholding and correction this analysis was reported under,
      by identifier. Analyses sharing one scheme reference one record; an analysis
      thresholded or corrected differently takes its own. Not the model's -- one fit
      thresholded two ways is two records here and one `model_estimation`. A contrast
      reported at two search volumes is two analyses sharing one `model_estimation`,
      not one analysis with the stricter threshold.
    in_subset:
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    domain_of:
    - Analysis
    - Study
    range: InferenceSettings
    inlined: false
  tables:
    name: tables
    description: The source tables reporting results for this analysis.
    in_subset:
    - mvp
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    domain_of:
    - Analysis
    - Study
    range: Table
    multivalued: true
    inlined: false
  mirror_of:
    name: mirror_of
    description: 'Present only on an analysis the builder derived by reversing another
      one, and naming the analysis it reverses.

      A results table with both signs in one column reports one contrast read two
      ways. The paper''s prose describes one direction, so only that half is extracted
      and the other is rebuilt arithmetically -- same cells, directions flipped. This
      slot is what says so. Without it the two halves look like independent findings,
      and a meta-analysis that counts both counts one sample twice.'
    in_subset:
    - mvp
    - deterministic
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    domain_of:
    - Analysis
    range: Analysis
    inlined: false
  source_table_analysis:
    name: source_table_analysis
    description: 'The key of the coordinate-parse analysis this analysis''s results
      were read from: `ParsedAnalysis.key` in neuroimaging-paper-parse, `analysis#<h>`,
      where `h` hashes the parse analysis''s name, space and points. Derived from
      what was extracted, not where it was read, so a re-parse that extracts the same
      analysis reaches the same key. When pondie revised the parse, the key is the
      revision''s.

      This schema stores no coordinates, so reaching an analysis''s foci means joining
      through the parse, and `tables` alone cannot do it: a table usually reports
      several contrasts and several analyses usually cite the same table. This key
      is what makes the join exact.

      Copy it from the bracketed key the stage-1 listing prints beside each parsed
      analysis. One key and not a list, because one analysis is read from one row
      group; an analysis whose results span two tables cites both in `tables` and
      names here the row group that defined it.'
    in_subset:
    - mvp
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    domain_of:
    - Analysis
    range: string
  statistical_maps:
    name: statistical_maps
    description: Statistical maps made available for this analysis.
    in_subset:
    - deterministic
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    domain_of:
    - Analysis
    range: StatisticalMap
    multivalued: true
    inlined_as_list: true
  coordinate_space:
    name: coordinate_space
    description: 'Coordinate reference space or template the spatial results are reported
      in, such as MNI152 or Talairach. **The authoritative of the two places this
      can live**: where this and a `Table.coordinate_space` disagree, this one is
      the analysis''s space, since a table can be parsed as one space while the analysis
      it reports states another.

      Both may be filled, and usually are: the table parser fills the tables and the
      same parse supplies this when every table behind the analysis agrees, leaving
      the model to answer only where they do not. Never inferred from the model''s
      `spatial_unit` -- an analysis on a surface reports a surface template.'
    in_subset:
    - mvp
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    domain_of:
    - Analysis
    - StatisticalMap
    - Table
    range: string
  assessments:
    name: assessments
    description: Assessments whose measurements were used in or selected the sample
      for this analysis.
    in_subset:
    - mvp
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    domain_of:
    - Analysis
    - Study
    range: Assessment
    multivalued: true
    inlined: false
  model_representation_notes:
    name: model_representation_notes
    description: Model components that could not be represented or were represented
      only approximately, for an analysis whose method is otherwise first-class. Examples
      include random slopes, latent variables, dynamic connectivity, and dependencies
      between terms. An analysis whose method has no stable structured decomposition
      at all takes NotStructurableDetails instead.
    in_subset:
    - mvp
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    domain_of:
    - Analysis
    range: string
rules:
- preconditions:
    slot_conditions:
      spatial_scope:
        name: spatial_scope
        any_of:
        - equals_string: whole_brain
        - equals_string: searchlight
  postconditions:
    slot_conditions:
      regions:
        name: regions
        value_presence: ABSENT
  description: Whole-brain and searchlight analyses restrict inference to no region.
- preconditions:
    slot_conditions:
      spatial_scope:
        name: spatial_scope
        equals_string: roi
  postconditions:
    slot_conditions:
      regions:
        name: regions
        value_presence: PRESENT
  description: An ROI analysis must name the regions it ran over.

Induced

name: Analysis
description: 'A distinct reported inferential map or tested effect, linking its source
  acquisitions, sample, paradigm context, tested effect, preprocessing, model estimation,
  inference settings, and source result tables.

  Create one Analysis for each distinct tested effect. Separate Analyses are required
  when normalized direction, participant-group comparison, method, the pattern of
  `effect.cells`, seed or connectivity model, decoded variable, component identity,
  or spatial_scope differ, and when an omnibus effect and its directional post-hoc
  contrast are both reported. ROI definitions, acquisition summaries, descriptive
  sample tables, and masks without inferential effects are not Analyses. Nor is a
  map reported with no inferential test at all -- an ICA or PCA component map presented
  descriptively, a connectivity matrix given without a test -- there being nothing
  for such a map to be the effect of.'
from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage
rank: 1000
attributes:
  id:
    name: id
    in_subset:
    - mvp
    - deterministic
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    identifier: true
    owner: Analysis
    domain_of:
    - Arm
    - Timepoint
    - Group
    - Task
    - Condition
    - Assessment
    - Region
    - Acquisition
    - Device
    - Analysis
    - Measure
    - Preprocessing
    - ModelEstimation
    - ModelTerm
    - InferenceSettings
    - StatisticalMap
    - Table
    - CoordinateSet
    - Study
    range: string
    required: true
  name:
    name: name
    description: Study-local label for the analysis or inferential result, in the
      source's words.
    in_subset:
    - mvp
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    owner: Analysis
    domain_of:
    - Arm
    - Timepoint
    - Group
    - Task
    - Condition
    - Assessment
    - Region
    - Analysis
    - ModelTerm
    - ExternalDataset
    - PerformanceMetric
    range: string
    required: true
  definition:
    name: definition
    description: Precise statement of the tested effect in the source's own words,
      minimally cleaned. Always source-grounded. For an interaction it must state
      the direction of the interaction as the source states it; for an ordered contrast
      it must keep the ordering that the cell directions cannot express.
    in_subset:
    - mvp
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    owner: Analysis
    domain_of:
    - Analysis
    - ConjunctionComponent
    range: string
    required: true
  prespecification:
    name: prespecification
    description: 'Whether this contrast was planned before the data were seen or arrived
      at afterwards.

      Open vocabulary: write the source''s own wording in the field when neither value
      fits. The two are not a spectrum and the words a source uses for them are many
      -- `post-hoc`, `confirmatory`, `planned`, `hypothesis-driven` -- so a closed
      range would discard the whole analysis over a synonym. Normalization maps the
      wording onto the two values.'
    in_subset:
    - mvp
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    owner: Analysis
    domain_of:
    - Analysis
    range: string
    required: true
    any_of:
    - range: Prespecification
    - range: string
  interpretations:
    name: interpretations
    description: 'The source''s own statement of what this analysis found and what
      it means, typically in Results and Discussion, verbatim or minimally cleaned.
      Always source-grounded, and the counterpart of `definition`: that field is the
      Methods-side statement of what was tested, this one is the Results/Discussion-side
      statement of what came of it.'
    in_subset:
    - mvp
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    owner: Analysis
    domain_of:
    - Analysis
    range: string
    multivalued: true
  outcome:
    name: outcome
    description: 'Whether anything survived the threshold this analysis was tested
      at: a cluster, a peak, a significant region. The answer to "did this contrast
      find anything", which `Effect.cells` cannot give -- a cell''s direction is the
      sign of the contrast that was tested, and is written the same whether the test
      found an effect or not. Meta-analyses exclude null effects and pool only analyses
      that reported foci, so a tested-and-null contrast and a reported effect must
      not extract to the same record. Applies only to an analysis whose coordinate
      set is a `result` (or has no role yet): an ROI, seed, reference or other set
      was not tested here, so it has no outcome.'
    in_subset:
    - mvp
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    owner: Analysis
    domain_of:
    - Analysis
    range: AnalysisOutcome
  spatial_scope:
    name: spatial_scope
    description: Spatial domain of the analysis modelled. The spatial_scope refers
      to what voxels were modelled, not which voxels/brain regions had a correction
      applied.
    in_subset:
    - mvp
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    owner: Analysis
    domain_of:
    - Analysis
    range: string
    required: true
    any_of:
    - range: SpatialScope
    - range: string
  regions:
    name: regions
    description: 'The regions the analysis ran over -- its search space, what `spatial_scope:
      roi` restricts it to. Empty for whole-brain and searchlight, where the emptiness
      asserts that inference was not restricted.

      Not the analysis''s seeds: a whole-brain seed-based analysis has a seed and
      no search-space region, so putting one here says the opposite of what it did.
      Seeds are `ConnectivityDetails.seed_regions`.'
    in_subset:
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    owner: Analysis
    domain_of:
    - Analysis
    - FactorLevel
    - CoordinateSet
    - Study
    range: Region
    multivalued: true
    inlined: false
  defines_regions:
    name: defines_regions
    description: 'Regions this analysis produced: a localizer''s clusters, or peaks
      of a contrast later used as seeds or ROIs. Populated on the analysis a region
      came *from*, not on those that use it, so a region whose `definition_method`
      is `same_study_analysis` and which another analysis lists in `regions` or `seed_regions`
      is a circularity the record states rather than asserts.'
    in_subset:
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    owner: Analysis
    domain_of:
    - Analysis
    range: Region
    multivalued: true
    inlined: false
  measure:
    name: measure
    description: 'The scientific quantity of the map, by identifier, independent of
      the statistic used to test it. Analyses of one quantity reference one record;
      an analysis of a different quantity takes its own. Required: a paper vague about
      what it measured still takes a record, with `source_label` as given and `extraction_status:
      not_reported` elsewhere.'
    in_subset:
    - mvp
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    owner: Analysis
    domain_of:
    - Analysis
    range: Measure
    required: true
    inlined: false
  groups:
    name: groups
    description: 'The participant cohorts the analysis ran on, and how many participants
      each contributed. Membership only: this slot says which people were analysed,
      never what was compared. A cohort comparison is a pair of directional `Cell`s
      on the model term whose levels name the cohorts, exactly as a condition or occasion
      comparison is.'
    in_subset:
    - mvp
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    owner: Analysis
    domain_of:
    - Analysis
    - FactorLevel
    - Study
    range: AnalysisGroup
    required: true
    multivalued: true
    inlined: true
    inlined_as_list: true
  effect:
    name: effect
    description: What was tested about the map. Universal across every method.
    in_subset:
    - mvp
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    owner: Analysis
    domain_of:
    - Analysis
    range: Effect
    required: true
    inlined: true
  details:
    name: details
    description: The method that produced the map and its method-specific payload.
      The populated subclass identifies the method. A voxelwise, vertexwise, or ROI-wise
      GLM, mixed-effects model, ANOVA, ANCOVA, or robust regression takes MassUnivariateDetails;
      a method outside the first-class families takes OtherAnalysisDetails and names
      itself in method_label.
    in_subset:
    - mvp
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    owner: Analysis
    domain_of:
    - Analysis
    range: AnalysisDetails
    required: true
    inlined: true
  acquisitions:
    name: acquisitions
    description: Acquisition protocols supplying data to this analysis. Structural,
      diffusion, and other non-paradigm analyses are linked directly through this
      field.
    in_subset:
    - mvp
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    owner: Analysis
    domain_of:
    - Task
    - Analysis
    - Study
    range: Acquisition
    multivalued: true
    inlined: false
  tasks:
    name: tasks
    description: 'The participant paradigms that contextualize this analysis. An analysis
      can span paradigms, as a cross-task conjunction or a localizer-plus-main-task
      analysis does.

      A resting-state analysis names its Task here like any other: the resting-state
      protocol is a paradigm, and a Task record exists for it. What leaves this empty
      is an analysis with no paradigm at all -- morphometry, diffusion, a static PET
      scan -- which is linked through `acquisitions` instead.'
    in_subset:
    - mvp
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    owner: Analysis
    domain_of:
    - Analysis
    - Study
    range: Task
    multivalued: true
    inlined: false
  model_estimation:
    name: model_estimation
    description: 'The model this analysis''s contrast was taken from: its family,
      estimator, stage, software, and the terms it was specified with. What was selected
      out of those terms is `effect`. Required: a paper that describes a contrast
      and never names an estimator still takes a record, with the terms it does state
      and `extraction_status: not_reported` elsewhere. Analyses sharing one design
      reference one record; an analysis specified or estimated differently takes its
      own.'
    in_subset:
    - mvp
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    owner: Analysis
    domain_of:
    - Analysis
    range: ModelEstimation
    required: true
    inlined: false
  inference_settings:
    name: inference_settings
    description: The thresholding and correction this analysis was reported under,
      by identifier. Analyses sharing one scheme reference one record; an analysis
      thresholded or corrected differently takes its own. Not the model's -- one fit
      thresholded two ways is two records here and one `model_estimation`. A contrast
      reported at two search volumes is two analyses sharing one `model_estimation`,
      not one analysis with the stricter threshold.
    in_subset:
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    owner: Analysis
    domain_of:
    - Analysis
    - Study
    range: InferenceSettings
    inlined: false
  tables:
    name: tables
    description: The source tables reporting results for this analysis.
    in_subset:
    - mvp
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    owner: Analysis
    domain_of:
    - Analysis
    - Study
    range: Table
    multivalued: true
    inlined: false
  mirror_of:
    name: mirror_of
    description: 'Present only on an analysis the builder derived by reversing another
      one, and naming the analysis it reverses.

      A results table with both signs in one column reports one contrast read two
      ways. The paper''s prose describes one direction, so only that half is extracted
      and the other is rebuilt arithmetically -- same cells, directions flipped. This
      slot is what says so. Without it the two halves look like independent findings,
      and a meta-analysis that counts both counts one sample twice.'
    in_subset:
    - mvp
    - deterministic
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    owner: Analysis
    domain_of:
    - Analysis
    range: Analysis
    inlined: false
  source_table_analysis:
    name: source_table_analysis
    description: 'The key of the coordinate-parse analysis this analysis''s results
      were read from: `ParsedAnalysis.key` in neuroimaging-paper-parse, `analysis#<h>`,
      where `h` hashes the parse analysis''s name, space and points. Derived from
      what was extracted, not where it was read, so a re-parse that extracts the same
      analysis reaches the same key. When pondie revised the parse, the key is the
      revision''s.

      This schema stores no coordinates, so reaching an analysis''s foci means joining
      through the parse, and `tables` alone cannot do it: a table usually reports
      several contrasts and several analyses usually cite the same table. This key
      is what makes the join exact.

      Copy it from the bracketed key the stage-1 listing prints beside each parsed
      analysis. One key and not a list, because one analysis is read from one row
      group; an analysis whose results span two tables cites both in `tables` and
      names here the row group that defined it.'
    in_subset:
    - mvp
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    owner: Analysis
    domain_of:
    - Analysis
    range: string
  statistical_maps:
    name: statistical_maps
    description: Statistical maps made available for this analysis.
    in_subset:
    - deterministic
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    owner: Analysis
    domain_of:
    - Analysis
    range: StatisticalMap
    multivalued: true
    inlined: true
    inlined_as_list: true
  coordinate_space:
    name: coordinate_space
    description: 'Coordinate reference space or template the spatial results are reported
      in, such as MNI152 or Talairach. **The authoritative of the two places this
      can live**: where this and a `Table.coordinate_space` disagree, this one is
      the analysis''s space, since a table can be parsed as one space while the analysis
      it reports states another.

      Both may be filled, and usually are: the table parser fills the tables and the
      same parse supplies this when every table behind the analysis agrees, leaving
      the model to answer only where they do not. Never inferred from the model''s
      `spatial_unit` -- an analysis on a surface reports a surface template.'
    in_subset:
    - mvp
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    owner: Analysis
    domain_of:
    - Analysis
    - StatisticalMap
    - Table
    range: string
  assessments:
    name: assessments
    description: Assessments whose measurements were used in or selected the sample
      for this analysis.
    in_subset:
    - mvp
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    owner: Analysis
    domain_of:
    - Analysis
    - Study
    range: Assessment
    multivalued: true
    inlined: false
  model_representation_notes:
    name: model_representation_notes
    description: Model components that could not be represented or were represented
      only approximately, for an analysis whose method is otherwise first-class. Examples
      include random slopes, latent variables, dynamic connectivity, and dependencies
      between terms. An analysis whose method has no stable structured decomposition
      at all takes NotStructurableDetails instead.
    in_subset:
    - mvp
    - model_extracted
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage/analysis
    rank: 1000
    owner: Analysis
    domain_of:
    - Analysis
    range: string
rules:
- preconditions:
    slot_conditions:
      spatial_scope:
        name: spatial_scope
        any_of:
        - equals_string: whole_brain
        - equals_string: searchlight
  postconditions:
    slot_conditions:
      regions:
        name: regions
        value_presence: ABSENT
  description: Whole-brain and searchlight analyses restrict inference to no region.
- preconditions:
    slot_conditions:
      spatial_scope:
        name: spatial_scope
        equals_string: roi
  postconditions:
    slot_conditions:
      regions:
        name: regions
        value_presence: PRESENT
  description: An ROI analysis must name the regions it ran over.