- conversation about clinical data management practices that generalize to non-clinical projects
- continue developing bibliometrics cover pages for faculty dossier workshop
- data management lit review – statistics texts
Clinical data management
I had an interesting conversation with the CTSI’s Director of Clinical Data Management a couple of weeks ago. We share the perspective that there are a lot of common elements to good data management practices across clinical and non-clinical research. I’m still surprised by how little literature there is on the topic. Perhaps because the Good Clinical Data Management Practices manual (produced by the Society for Clinical Data Management) seems to be heavily used and informed by industry folks, academic researchers have paid it little attention. In clinical research, there is more often a recognized role for data management, thus separation of data management responsibilities from the PI. That’s one challenge to improving the data management practices for fields or teams in which data management is embedded in many roles and no one person has clear responsibility – it’s not a priority and there is no accountability. Aside from identifying common good practices, this is the real issue in data management – how do we get researchers to incorporate good data management practices in their processes? First comes the awareness, then attitude changes, and finally the hardest of all – behavioral/cultural changes.
Anyhow, this led to an invitation to sit in on a call with a teaching/research hospital that has a robust data management group. It was fascinating to hear an industry-informed perspective on data management and the general messiness of academic research. There is so much valuable information to be shared here, so why isn’t this knowledge transfer happening on a larger scale? This is ripe for a network analysis study, which I don’t have the chops or time to do.
Bibliometrics cover pages
We’re offering our Gathering Evidence workshop for the second time this September. I’ve been tasked with developing some sample evidence pages for faculty to incorporate into their dossier. Setting aside the fact that I am not a designer, this has proven to be very challenging and time-intensive. Standardizing the presentation of even citation data is confusing because Web of Science, Scopus, and Google Scholar all have varying coverage. That doesn’t even begin to touch on the more ephemeral social data – views, downloads, shares, etc. for other types of impact and other types of products. Each researcher and field of study seems to have their own behaviors, social networks used, and expectations for information sharing. Is any of this documented? Nope! Should we – as users – have better access to these data? Yes! The reasons why are beyond my ability to articulate today, but we truly need easier access to these data for P&T dossiers, research impact metrics, and institutional evaluation.
Data management lit review – statistics texts
I’m getting a bit bogged down in the numerical perspective of data and analysis, so I’m interspersing reading from two qualitative texts (Dey and Creswell) to balance it out. As I read through these texts and having conversations with some computer and information science faculty, I’m realizing that much of this doesn’t apply to computational and modeling code and data. Guess I’ll be adding to the reading list for phase 3 of this lit review!