1. Sign our petition calling on Cochrane to withdraw their review of Exercise Therapy for CFS here.
    Dismiss Notice
  2. Guest, the 'News in Brief' for the week beginning 18th March 2024 is here.
    Dismiss Notice
  3. Welcome! To read the Core Purpose and Values of our forum, click here.
    Dismiss Notice

Data Sharing & Reanalysis

Discussion in 'Research methodology news and research' started by James, Feb 18, 2018.

  1. James

    James Established Member (Voting Rights)

    Messages:
    90
  2. Adrian

    Adrian Administrator Staff Member

    Messages:
    6,478
    Location:
    UK
    The conclusions are something I was wondering about. Data sharing is really quite minimal hence it allows calculations to be checked but what is really needed is sufficient data to explore other questions and to combine data sets.
     
    MErmaid, Louie41, Esther12 and 4 others like this.
  3. James

    James Established Member (Voting Rights)

    Messages:
    90
    The data shared is a check on research integrity by anyone motivated and qualified that questions research reports in a journal.
    The NIH have been pushing for common data elements since around 2006 that would enable combination of data sets although the words "data parasite" have been bandied about by those unhappy about such openness / sharing /transparency.
     
    Invisible Woman likes this.
  4. Adrian

    Adrian Administrator Staff Member

    Messages:
    6,478
    Location:
    UK
    I find it a really offensive term. People take part in clinical trials (with associated risks) because they want new treatments and to help the science move on to. When people talk of "data parasites" they are basically arguing that they own the data and should have rights to it so that they can further their careers by publishing from it. But people didn't take part in the trial to further academics careers but to further science.

    Given analysis plans should be prepublished I don't see why it takes long to publish. The code for the analysis should be written and tested prior to the data being unblinded. (although journals are slow). Hence I wouldn't give much time at all. The aim should be to maximize value from the data as quickly as possible.
     
  5. Barry

    Barry Senior Member (Voting Rights)

    Messages:
    8,385
    [My bold]

    Oh well, that argument is obviously completely rational, given how they fully justify their reasoning for it not being appropriate ... not! :rolleyes::rolleyes:.
     
  6. James

    James Established Member (Voting Rights)

    Messages:
    90
    The MRC list that objective on their website:
    Every year, the MRC invests around £850 million of public money in research, the primary output of which is data. We want to maximise the research opportunities that such a diversity, richness and quantity of data provides. One of the best ways of achieving this is to ensure that data are properly preserved for sharing and informed use beyond the originating research teams
    https://www.mrc.ac.uk/research/policies-and-guidance-for-researchers/data-sharing/

    As guidance on the use of publicly funded research and ethical responsibility for the implicit contract between research participants and researchers that demonstrable benefit has come from public spiritedness, the MRC page reads as though it were a letter to Santa wish list.
     
  7. Adrian

    Adrian Administrator Staff Member

    Messages:
    6,478
    Location:
    UK
    Neither does the statement appeal to authority in terms of saying who they got advice from. They could simply have said Peter 'what do you think about releasing data?' and got the response 'Well Peter, I think it would be inappropriate'.
     

Share This Page