Slot: sample_source¶
Where this group's participants came from. recruited when they were enrolled for this study and no other source is stated; public_dataset when they were drawn from a named public or consortium dataset; previously_reported when the same or an overlapping sample was already reported in another publication.
On the group and not the study, because overlap is routinely partial: a study may scan a fresh patient group against controls taken from an existing dataset, and a study-level flag cannot say which. Study.external_datasets remains the place for a named dataset the study as a whole used.
A protocol, preprocessing pipeline or analysis "described previously" is a shared METHOD and leaves this recruited; so does comparing findings to earlier literature in the discussion. Only a statement that these same people were scanned, studied or reported before counts, and a source that means it normally says so in one sentence.
recruited is an assertion and not the absence of one. A meta-analysis excluding overlapping samples has to distinguish a study that recruited freshly from one nobody checked, and an unfilled field cannot.
Applicable Classes¶
| Name | Description | Modifies Slot |
|---|---|---|
| Group | An aggregate participant group and its reported demographics | no |
Properties¶
Type and Range¶
| Property | Value |
|---|---|
| Range | SampleSource |
| Domain Of | Group |
| ### Slot Characteristics |
| Property | Value |
|---|---|
| Owner | Group |
In Subsets¶
Identifier and Mapping Information¶
Schema Source¶
- from schema: https://neurostuff.org/schemas/neuroimaging-study-storage
Mappings¶
| Mapping Type | Mapped Value |
|---|---|
| self | neurostudy:sample_source |
| native | neurostudy:sample_source |
LinkML Source¶
name: sample_source
description: 'Where this group''s participants came from. `recruited` when they were
enrolled for this study and no other source is stated; `public_dataset` when they
were drawn from a named public or consortium dataset; `previously_reported` when
the same or an overlapping sample was already reported in another publication.
On the group and not the study, because overlap is routinely partial: a study may
scan a fresh patient group against controls taken from an existing dataset, and
a study-level flag cannot say which. `Study.external_datasets` remains the place
for a named dataset the study as a whole used.
A protocol, preprocessing pipeline or analysis "described previously" is a shared
METHOD and leaves this `recruited`; so does comparing findings to earlier literature
in the discussion. Only a statement that these same people were scanned, studied
or reported before counts, and a source that means it normally says so in one sentence.
`recruited` is an assertion and not the absence of one. A meta-analysis excluding
overlapping samples has to distinguish a study that recruited freshly from one nobody
checked, and an unfilled field cannot.'
in_subset:
- mvp
- model_extracted
from_schema: https://neurostuff.org/schemas/neuroimaging-study-storage
rank: 1000
owner: Group
domain_of:
- Group
range: SampleSource