Class: FactorLevel

One level of a categorical model term and the entities that carry it. The ModelTerm records which design cell a cohort or condition occupies because that placement belongs to the design, not to a contrast drawn from it.

The five entity slots are what make the entity view of a contrast recoverable: an Effect's cells name levels, and joining those levels here says which cohorts, conditions, occasions, arms or regions were compared. A factor normally ranges over one of the five -- conditions for a task factor, cohorts for a participant factor, occasions for a time factor, arms for a treatment factor, regions for one comparing places in the brain -- so the others stay empty; a level realized by a crossing, such as a drug given at follow-up, fills more than one.

A factor whose levels are not entities of the study at all -- hemispheres, tasks, frequency bands, acquisition sessions -- leaves all five empty and is carried by level alone. Such a factor is complete, and takes Cells like any other.

URI: neurostudy_ex:FactorLevel

 classDiagram
    class FactorLevel
    click FactorLevel href "../FactorLevel/"
      FactorLevel : arms





        FactorLevel --> "*" Arm : arms
        click Arm href "../Arm/"



      FactorLevel : conditions





        FactorLevel --> "*" Condition : conditions
        click Condition href "../Condition/"



      FactorLevel : groups





        FactorLevel --> "*" Group : groups
        click Group href "../Group/"



      FactorLevel : level





        FactorLevel --> "1" ExtractedString : level
        click ExtractedString href "../ExtractedString/"



      FactorLevel : order





        FactorLevel --> "0..1" ExtractedInteger : order
        click ExtractedInteger href "../ExtractedInteger/"



      FactorLevel : regions





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



      FactorLevel : timepoints





        FactorLevel --> "*" Timepoint : timepoints
        click Timepoint href "../Timepoint/"

Slots

Name	Cardinality and Range	Description	Inheritance
level	1 ExtractedString	The level's label, as the source words it	direct
order	0..1 ExtractedInteger	Position of this level in the factor's sequence, counting from 1, for a facto...	direct
conditions	* Condition	The conditions carrying this level, by identifier	direct
groups	* Group	The participant cohorts carrying this level, by identifier	direct
timepoints	* Timepoint	The occasions carrying this level, by identifier	direct
arms	* Arm	The arms carrying this level, by identifier	direct
regions	* Region	The brain regions carrying this level, by identifier, for a factor comparing ...	direct

Usages

used by	used in	type	used
ModelTerm	levels	range	FactorLevel

Identifier and Mapping Information

Schema Source

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

Mappings

Mapping Type	Mapped Value
self	neurostudy_ex:FactorLevel
native	neurostudy_ex:FactorLevel

LinkML Source

Direct

name: FactorLevel
description: 'One level of a categorical model term and the entities that carry it.
  The ModelTerm records which design cell a cohort or condition occupies because that
  placement belongs to the design, not to a contrast drawn from it.

  The five entity slots are what make the entity view of a contrast recoverable: an
  Effect''s cells name levels, and joining those levels here says which cohorts, conditions,
  occasions, arms or regions were compared. A factor normally ranges over one of the
  five -- conditions for a task factor, cohorts for a participant factor, occasions
  for a time factor, arms for a treatment factor, regions for one comparing places
  in the brain -- so the others stay empty; a level realized by a crossing, such as
  a drug given at follow-up, fills more than one.

  A factor whose levels are not entities of the study at all -- hemispheres, tasks,
  frequency bands, acquisition sessions -- leaves all five empty and is carried by
  `level` alone. Such a factor is complete, and takes `Cell`s like any other.'
from_schema: https://neurostuff.org/schemas/neuroimaging-study-extraction
rank: 1000
attributes:
  level:
    name: level
    description: 'The level''s label, as the source words it. Unique within its term:
      a level appearing twice is one level recorded twice.'
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-extraction/analysis
    domain_of:
    - Cell
    - FactorLevel
    - PaperSection
    range: ExtractedString
    required: true
  order:
    name: order
    description: 'Position of this level in the factor''s sequence, counting from
      1, for a factor whose levels are ordered. It carries an ordering `Effect.cells`
      cannot: a contrast signs only its extremes and omits what lies between, so without
      this 1-back, 2-back and 3-back are an unordered set.

      Populate it whenever the levels have an order the source states or plainly implies
      -- a dose series, a load or difficulty series, an ordinal severity grading.
      Leave it unset for a nominal factor, where imposing an order would assert a
      structure the design does not have. Timepoint.order is the same device for occasions.'
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-extraction/analysis
    domain_of:
    - Timepoint
    - FactorLevel
    range: ExtractedInteger
  conditions:
    name: conditions
    description: The conditions carrying this level, by identifier.
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-extraction/analysis
    domain_of:
    - Task
    - FactorLevel
    range: Condition
    multivalued: true
    inlined: false
  groups:
    name: groups
    description: The participant cohorts carrying this level, by identifier.
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-extraction/analysis
    domain_of:
    - Analysis
    - FactorLevel
    - Study
    range: Group
    multivalued: true
    inlined: false
  timepoints:
    name: timepoints
    description: 'The occasions carrying this level, by identifier. A factor whose
      levels are baseline, week 6 or follow-up ranges over occasions and fills this:
      left empty, the level stands on its label alone and no query can tell which
      scan preceded the intervention.'
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-extraction/analysis
    domain_of:
    - StudyDesign
    - FactorLevel
    range: Timepoint
    multivalued: true
    inlined: false
  arms:
    name: arms
    description: The arms carrying this level, by identifier. A factor whose levels
      are drug and placebo, or active and sham, ranges over arms. In a parallel trial
      each level is both an arm and a cohort and both slots are filled; in a crossover
      one cohort receives every arm, so `arms` alone distinguishes the levels and
      filling only `groups` loses the allocation.
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-extraction/analysis
    domain_of:
    - StudyDesign
    - FactorLevel
    range: Arm
    multivalued: true
    inlined: false
  regions:
    name: regions
    description: The brain regions carrying this level, by identifier, for a factor
      comparing places. A three-region factor is one `ModelTerm` with three levels,
      one region on each.
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-extraction/analysis
    domain_of:
    - Analysis
    - FactorLevel
    - CoordinateSet
    - Study
    range: Region
    multivalued: true
    inlined: false

Induced

name: FactorLevel
description: 'One level of a categorical model term and the entities that carry it.
  The ModelTerm records which design cell a cohort or condition occupies because that
  placement belongs to the design, not to a contrast drawn from it.

  The five entity slots are what make the entity view of a contrast recoverable: an
  Effect''s cells name levels, and joining those levels here says which cohorts, conditions,
  occasions, arms or regions were compared. A factor normally ranges over one of the
  five -- conditions for a task factor, cohorts for a participant factor, occasions
  for a time factor, arms for a treatment factor, regions for one comparing places
  in the brain -- so the others stay empty; a level realized by a crossing, such as
  a drug given at follow-up, fills more than one.

  A factor whose levels are not entities of the study at all -- hemispheres, tasks,
  frequency bands, acquisition sessions -- leaves all five empty and is carried by
  `level` alone. Such a factor is complete, and takes `Cell`s like any other.'
from_schema: https://neurostuff.org/schemas/neuroimaging-study-extraction
rank: 1000
attributes:
  level:
    name: level
    description: 'The level''s label, as the source words it. Unique within its term:
      a level appearing twice is one level recorded twice.'
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-extraction/analysis
    owner: FactorLevel
    domain_of:
    - Cell
    - FactorLevel
    - PaperSection
    range: ExtractedString
    required: true
  order:
    name: order
    description: 'Position of this level in the factor''s sequence, counting from
      1, for a factor whose levels are ordered. It carries an ordering `Effect.cells`
      cannot: a contrast signs only its extremes and omits what lies between, so without
      this 1-back, 2-back and 3-back are an unordered set.

      Populate it whenever the levels have an order the source states or plainly implies
      -- a dose series, a load or difficulty series, an ordinal severity grading.
      Leave it unset for a nominal factor, where imposing an order would assert a
      structure the design does not have. Timepoint.order is the same device for occasions.'
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-extraction/analysis
    owner: FactorLevel
    domain_of:
    - Timepoint
    - FactorLevel
    range: ExtractedInteger
  conditions:
    name: conditions
    description: The conditions carrying this level, by identifier.
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-extraction/analysis
    owner: FactorLevel
    domain_of:
    - Task
    - FactorLevel
    range: Condition
    multivalued: true
    inlined: false
  groups:
    name: groups
    description: The participant cohorts carrying this level, by identifier.
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-extraction/analysis
    owner: FactorLevel
    domain_of:
    - Analysis
    - FactorLevel
    - Study
    range: Group
    multivalued: true
    inlined: false
  timepoints:
    name: timepoints
    description: 'The occasions carrying this level, by identifier. A factor whose
      levels are baseline, week 6 or follow-up ranges over occasions and fills this:
      left empty, the level stands on its label alone and no query can tell which
      scan preceded the intervention.'
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-extraction/analysis
    owner: FactorLevel
    domain_of:
    - StudyDesign
    - FactorLevel
    range: Timepoint
    multivalued: true
    inlined: false
  arms:
    name: arms
    description: The arms carrying this level, by identifier. A factor whose levels
      are drug and placebo, or active and sham, ranges over arms. In a parallel trial
      each level is both an arm and a cohort and both slots are filled; in a crossover
      one cohort receives every arm, so `arms` alone distinguishes the levels and
      filling only `groups` loses the allocation.
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-extraction/analysis
    owner: FactorLevel
    domain_of:
    - StudyDesign
    - FactorLevel
    range: Arm
    multivalued: true
    inlined: false
  regions:
    name: regions
    description: The brain regions carrying this level, by identifier, for a factor
      comparing places. A three-region factor is one `ModelTerm` with three levels,
      one region on each.
    from_schema: https://neurostuff.org/schemas/neuroimaging-study-extraction/analysis
    owner: FactorLevel
    domain_of:
    - Analysis
    - FactorLevel
    - CoordinateSet
    - Study
    range: Region
    multivalued: true
    inlined: false