Transparency as it relates to power differentials
Posted: Wed Nov 30, 2016 4:59 pm
This working group is also considering how different transparency norms have differential accts given the power inequities in our profession. I am an assistant-level faculty member at a good (though not top) liberal arts college. I received less than ten thousand dollars in start-up funds. My co-PIs and I recently launched a set of survey experiments, which are costing upwards of five figures. We will apply for external grants -- but, in the short term, these initial experiments have decimated my start-up funds. We plan to write an article on the results, and then develop a book-length project.
I bring these numbers up because I believe DA-RT poses serious costs on even quantitative scholars. If my co-PIs and I publish an initial article on our results, we must make all our data public. This rule turns the dollar cost of our entire project into the cost of a single article, and we would lose exclusive access to the data as we develop the book-length project. How are under-resourced scholars supposed to compete at this rate? Junior scholars--or scholars at any level--should not be asked to invest thousands of dollars of limited funds, as well as perhaps several years of work , to simply publish one article. This point especially holds where graduate students or junior scholars have spent several years constructing datasets by hand (rather than simply purchasing them via an on-line experiment).
Either the peer review process sufficiently vets scholarly work without having access to the original data or replication files--or the peer review system is broken. If the peer review system is broken, then we need more fixes than simply making the data accesible, transparent, and replicable to the global community. (For instance-- one compromise solution-- require the researchers to turn over the data to graduate students working for the journal, for a replication check prior to publication, but do not require the dataset to be made available for free, on-line.)
Resource constraints will always dictate the kinds of scholarly work we can and cannot do. For instance, I could choose different projects, one that did not decimate my start-up funds as quickly. However, DA-RT should not be exacerbating these inequities. Perhaps asking scholars with access to five or six-figure research funds to un-gate their data upon publication does not impose a high burden, as funds exist for well-resourced scholars to be endlessly creative in the next iteration of their designs. (And scholars from well-resourced institutions also fare disproportionately better in access to large-dollar research grants, which further allowes them to "pay to play.") But, for under-resourced scholars, DA-RT only sharpens their inability to compete-- or compete for long-- when it comes to publishing their work in top journals.
I bring these numbers up because I believe DA-RT poses serious costs on even quantitative scholars. If my co-PIs and I publish an initial article on our results, we must make all our data public. This rule turns the dollar cost of our entire project into the cost of a single article, and we would lose exclusive access to the data as we develop the book-length project. How are under-resourced scholars supposed to compete at this rate? Junior scholars--or scholars at any level--should not be asked to invest thousands of dollars of limited funds, as well as perhaps several years of work , to simply publish one article. This point especially holds where graduate students or junior scholars have spent several years constructing datasets by hand (rather than simply purchasing them via an on-line experiment).
Either the peer review process sufficiently vets scholarly work without having access to the original data or replication files--or the peer review system is broken. If the peer review system is broken, then we need more fixes than simply making the data accesible, transparent, and replicable to the global community. (For instance-- one compromise solution-- require the researchers to turn over the data to graduate students working for the journal, for a replication check prior to publication, but do not require the dataset to be made available for free, on-line.)
Resource constraints will always dictate the kinds of scholarly work we can and cannot do. For instance, I could choose different projects, one that did not decimate my start-up funds as quickly. However, DA-RT should not be exacerbating these inequities. Perhaps asking scholars with access to five or six-figure research funds to un-gate their data upon publication does not impose a high burden, as funds exist for well-resourced scholars to be endlessly creative in the next iteration of their designs. (And scholars from well-resourced institutions also fare disproportionately better in access to large-dollar research grants, which further allowes them to "pay to play.") But, for under-resourced scholars, DA-RT only sharpens their inability to compete-- or compete for long-- when it comes to publishing their work in top journals.