The registration continuum in clinical science: A guide toward transparent practices.

Abstract

Clinical scientists can use a continuum of registration efforts that vary in their disclosure and timing relative to data collection and analysis. Broadly speaking, registration benefits investigators by offering stronger, more powerful tests of theory with particular methods in tandem with better control of long-run false positive error rates. Registration helps clinical researchers in thinking through tensions between bandwidth and fidelity that surround recruiting participants, defining clinical phenotypes, handling comorbidity, treating missing data, and analyzing rich and complex data. In particular, registration helps record and justify the reasons behind specific study design decisions, though it also provides the opportunity to register entire decision trees with specific endpoints. Creating ever more faithful registrations and standard operating procedures may offer alternative methods of judging a clinical investigator's scientific skill and eminence because study registration increases the transparency of clinical researchers' work.

Making the black box transparent: A template and tutorial for (pre-)registration of studies using Experience Sampling Methods (ESM)

“…As previously discussed in this paper, and consistent with the idea of a registration continuum (Benning et al, 2019), studies using pre-existing data can be post-registered. In these cases, for maximal transparency researchers should record their prior knowledge of the datasets (van den Akker et al, 2019), as any knowledge of the data can lead researchers to make data dependent decisions, and consequently introduce further researcher degrees of freedom into the process.…”

“…For example, the term "post-registration" has been proposed for registrations of studies using pre-existing datasets, where all data have been collected but no data have been analyzed or where some data have already been analyzed and published, e.g. by other members of the research team (Benning, Bachrach, Smith, Freeman, & Wright, 2019).…”

“…Some issues, however, may only emerge after data have been accessed, explored, or analyzed. In conceptualizing pre-registration as a continuum (Benning et al, 2019), we feel there is still value in transparently documenting changes to analysis plans in supplementary registrations, even when data have been accessed. In this case, the function of the registration is altered and serves more as an open lab book.…”

“…Following reviewer comments, we have also incorporated elements from the pre-registration template for pre-existing data (van den Akker et al, 2019), to reflect the fact that many ESM studies yield large, rich datasets, which are also used by other researchers for primary analysis and, over time, for secondary analysis. The expanding conceptualization of the "registration continuum" to include "post-registrations" of studies using archival data (Benning et al, 2019) also brings new opportunities in this regard. Throughout the paper, we have taken the position that open and transparent plans are, even if imperfect, better than no plans (Nosek et al, 2019) and have made suggestions for maximizing transparency in cases where deviations from the (pre-)registration are necessary, e.g.…”

“…Throughout the paper, we have taken the position that open and transparent plans are, even if imperfect, better than no plans (Nosek et al, 2019) and have made suggestions for maximizing transparency in cases where deviations from the (pre-)registration are necessary, e.g. through "co-registration" (Benning et al, 2019).…”