APA 7th, part I: Improvements in Data Transparency

Hi Ma! I’m in the APA 7th Edition publication manual.

apa 7

Thanks to Rose Sokol-Chang, Anne Woodworth and Emily Ayubi, I had a chance to give my input to a number of issues while it was in production.

For one, I’m happy that in line with my suggestions, they sharpened up the conflict of interest definitions (section 1.20, p. 23). Financial interests, which can involve authors, editors or reviewers, are distinguished from personal connections, which are of concern to editors and reviewers handling authors’ work. Differences of opinion are not per se conflicts, but advice is given to handle this issue objectively. Editors are encouraged to set policy on such issues as the time frame beyond which a past academic relationship is no longer a conflict.

Also, the new edition, with my input, clarifies the ethics of data sharing in previous APA policy (section 1.14, pp. 13-16). Let’s talk more about that.

For science to be self-correcting, scientists have to be able to check the data that published conclusions are based on. At the same time, we also want to limit the vexatious abuse of this process, and to give authors confidence that the data will only be used for checking the accuracy of their article. These principles guided me in advising how the policy could be improved, inspired by breakdowns in function of the previous policy and by experience with controversial cases as an editor. Here are some highlights that made it to the Manual.

  • Who counts as a “qualified professional?” (Accountability matters.)
  • What are reasonable uses of data to confirm the conclusions of the original paper? (Focus on verifying the conclusions, including the possibility of approaching them with alternative analyses).
  • What are special considerations when dealing with proprietary data, and data based on proprietary methodologies? (Editors need to decide how they act when they get a paper that would not exist if the data had to be shared).
  • How are the costs of data transmission to be assessed? (According to a reasonable local rate that covers the actual job.)
  • What are the consequences of failing to share data on the request of an editor? (Potentially, retraction or expressions of concern about the article).

The policy also asks the requester more specifically to bear the costs of sharing data over and above any preparations needed to make the data usable internally within an institution. In other words, it should always be the researcher’s responsibility to keep their data in good shape in case they get hit by a bus or decide to enter a hermitage. You, or I, may be slack on this responsibility at any given moment, but that doesn’t change whose responsibility it is.

What would this mean in practice? Well, if I ask for a data set, and you charge me $100 an hour to have someone change the variable name into something that matches up with the analyses you reported, that is just something you should have done in the first place. If I ask for you to translate open-ended responses from your source files from Chinese into English, that’s a different story. It’s more justifiable that I should be stuck with the bill, or just find someone to do the translation on my end.

Two more improvements I wasn’t involved in: encouragement for editors to adopt open data policies and data badges, and a section on sharing qualitative data. Overall, this expanded section gives guidance to make science published under APA rules more reliably transparent.The other section of the Manual I was consulted on concerned the implications of APA’s long-standing ethical policy about full reporting. The uptake of that advice in the final manual is more complicated, and leads me into some thoughts about how standards can be enforced in publishing. That’s another blog post, coming soon!