Dimensions Research Integrity on GBQ, Version 2.0.0

Summary

Version 2.0.0 of Dimensions Research Integrity on Google BigQuery (GBQ) features the following updates

• Improvements to all of our models for recognising Trust Markers, so you may see some changes in the data you see in Google BigQuery or the Dimensions Research Integrity app.

• A new "location" field for every Trust Marker record showing where in a publication the Trust Marker was found.

• data.data_locations.categories, which categorises publications according to whether and where the research data has been made available, may now take more than one value. This allows us to more accurately represent cases where a publication's data is available in more than one place.

• The top-level funding record is now called research_funding, to make it clear that it refers to funding of the current publication and help distinguish it from funding the author(s) might have received for other reasons, such as funding for separate projects that is declared in a paper as a potential source of conflicts of interest.

• The same data as in authors.conflict_of_interest_statement can now be found under a new conflicts_of_interest record, i.e. conflicts_of_interest.conflict_of_interest_statement, and this is now the recommended way to access this data.

New Features

Location field

A new location field is now available within every Trust Marker Record, indicating the type of section in a publication where we found a Trust Marker.

It is still possible to check the presence or absence of a Trust Marker in a publication using an approach like:

SELECT
  code.code_availability_statement.present
FROM dimensions-ai-integrity.data.trust_markers
WHERE id = "pub.1139691916"

But it is also now possible to see where in the publication the Trust Marker was found by using location:

SELECT
  code.code_availability_statement.location
FROM dimensions-ai-integrity.data.trust_markers
WHERE id = "pub.1139691916"

Schema changes

data.data_locations.categories may now take more than one value

data.data_locations.categories, which categorises publications according to whether and where the research data has been made available, may now take more than one value. This allows us to more accurately represent cases where a publication's data is available in more than one place.

For example:

SELECT data.data_locations.categories
FROM dimensions-ai-integrity.data.trust_markers
WHERE ARRAY_LENGTH(data.data_locations.categories) > 1
LIMIT 1

funding is now called research_funding

The top-level funding record is now called research_funding, to make it clear that it refers to funding of the current publication and help distinguish it from funding the author(s) might have received for other reasons, such as funding for separate projects that is declared in a paper as a potential source of conflicts of interest.

New conflicts_of_interest record

The conflict_of_interest_statement Trust Marker, which has so far been available as a record under the authors record in the data set

is now also available under a new record called conflicts_of_interest: