Professional Documents
Culture Documents
RR_What's next for Registered Reports-2019
RR_What's next for Registered Reports-2019
TECHNOLOGY How the race to SOCIETY Malcolm Gladwell’s HEALTH Elder or old — which FILM India’s frugal Mars
cash in is closing off research latest tome traces trust and view of age is best for our triumph gets big-screen
into quantum computing p.190 prejudice to same roots p.192 greying globe? p.193 moment p.194
ILLUSTRATION BY DAVID PARKINS
W
hat part of a research study — to cherry-picked highlights that can bring Clearly, we need to change how we
hypotheses, methods, results, prominent articles, grant funding, promo- evaluate and publish research. For the
or discussion — should remain tion and esteem. past six years, I have championed Reg-
beyond a scientist’s control? The answer, The ‘results paradox’ is a chief cause of istered Reports (RRs), a type of research
of course, is the results: the part that mat- unreliable science. Negative, or null, results article that is radically different from con-
ters most for publishing in prestigious go unpublished, leading other research- ventional papers. The 30 or so journals
journals and advancing careers. This ers into unwittingly redundant studies. that were early adopters have together
paradox means that the careful scepticism Ambiguous or otherwise ‘unattractive’ published some 200 RRs, and more than
required to avoid massaging data or skew- results are airbrushed (consciously or not) 200 journals are now accepting submis-
ing analysis is pitted against the drive to into publishable false positives, spurring sions in this format (see ‘Rapid rise’).
identify eye-catching outcomes. Unbiased, follow-up research and theories that are When it launched in 2017, Nature Human
negative and complicated findings lose out bound to collapse. Behaviour became the first of the Nature
1 2 S E P T E M B E R 2 0 1 9 | VO L 5 7 3 | NAT U R E | 1 8 7
©
2
0
1
9
S
p
r
i
n
g
e
r
N
a
t
u
r
e
L
i
m
i
t
e
d
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
COMMENT
SOURCE: C. CHAMBERS
Since 2013, the number of journals offering Registered Reports (RRs) has risen to more than 200 titles.
RRs on the rise, now is a good time to take
stock of their potential and limitations.
200
BMC Medicine launches
first RRs for clinical trials.
HOW DO THEY WORK? First multidisciplinary
The Registered Report format splits journal launches RRs
across 200 sciences (Royal
conventional peer review in half. First, Society Open Science).
Number of journals
authors write an explanation of how they
will probe an important question. This
‘Stage 1’ manuscript includes an overview 100
of the background literature, preliminary First journal exclusively for
RRs (Comprehensive Results Publication of 100th
work, theory, hypotheses and proposed in Social Psychology). completed RR.
methods, including the study procedures
and analysis plan. Before researchers do
the studies, peer reviewers assess the value
and validity of the research question, the 0
rationale of the hypotheses and the rigour 2013 2014 2015 2016 2017 2018 2019*
of the proposed methods. They might (*As of June)
1 8 8 | NAT U R E | VO L 5 7 3 | 1 2 S E P T E M B E R 2 0 1 9
©
2
0
1
9
S
p
r
i
n
g
e
r
N
a
t
u
r
e
L
i
m
i
t
e
d
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
COMMENT
project in a short time might not be able to exploratory analyses are labelled trans- (see go.nature.com/2kb5s7v). This is now
wait for reviewer feedback, although some parently as post hoc and do not dominate used by many journals, including Cortex
teaching programmes have had success by conclusions. and Animal Behavior and Cognition. The
dividing up research-project design and Exploration is alive and well. Stage 2 publisher Wiley has opted to publish
execution in creative ways (see, for exam- submissions almost always include further accepted protocols. And venues such as
ple, K. Button Nature 561, 287; 2018). analyses. The difference is that researchers F1000Research offer the option to post
By contrast, RRs have distinct advantages cannot fool themselves or their readers Stage 1 articles before peer review, with
for longer-term students. The in-principle by presenting only the most interesting reviews and revisions made public as they
acceptance at Stage 1 allows them to list a analyses or imply that these were intended become available. A badging system shows
publication much sooner than they could from the outset. RRs are a plan, not a prison. that the Stage 2 article adhered to the criteria
for a conventional manuscript, and with A related misgiving is that researchers and can be labelled as a RR.
more certainty. There is emerging evidence will find themselves locked into a sub
that RRs are popular with early-career optimal protocol once experiments begin. Standardization. Improving the stand-
researchers. For example, at Cortex, 78% of In my experience, the opposite is more ardization of submitted protocols promises
RR first authors (n = 82) are PhD students likely: reviewers can prevent researchers to improve computational reproducibility.
or postdocs, compared with 67% in a con- from running less-informative experi- Currently, submitted manuscripts are often
trol sample (n = 57) of conventional articles. ments. And reviewers of Stage 2 manu- prepared in word-processing software and
Although RRs require researchers to scripts generally understand reasonable contain insufficient methodological detail
wait for review before starting experi- changes. It is not flexibility that is lost, but or linking between predictions and analyses.
ments, I suspect that the time to pub- the ability to airbrush both reasonable and The next generation of RRs — ‘Registered
lication probably declines overall. A questionable changes out of the picture. Reports 2.0’ — is likely to be template-based
conventional article might be rejected on and could integrate tools such as Code Ocean
the basis of results or because of method- MOVING FORWARD (see https://codeocean.com/researchers).
ological problems that can no longer be RRs are not a panacea — the format needs This would ensure that analyses are
fixed, leaving authors to submit their work constant refinement. It currently sits immutable within a stable, self-contained
to journal after journal, or to perform rather awkwardly between the old world of software environment. With standardized
extra experiments. Over the past six years, scientific publishing and the new. Innova- metadata and badging, RRs will become use-
dozens of authors have told me —and tions over the next few years should make ful for systematic reviews and meta-analyses.
written publicly — that they appreciate the this format even more powerful, and
more-predictable timeline of RRs (see, for stimulate wider reforms. Efficiency. The review process can be
example, go.nature.com/2kwnjuj). extended even further back in the research
Decreased flexibility is an oft-expressed Transparency. When RRs first launched, life cycle. Under the emerging RR grant
concern over the format. One early critic some journals published Stage 2 manu- model, reviewers award funding and sig-
said it would “put science in chains”. The scripts but not those for Stage 1, making it nal in-principle acceptance of a research
fear is that peer-reviewed preregistration impossible for readers to see whether the publication simultaneously or in rapid
dampens the creativity and serendipity completed protocol matched the planned succession. The Children’s Tumor Founda-
that could come from free-wheeling data one. In 2018, the Center for Open Science tion and PLoS ONE have pioneered such a
exploration. But preregistration imposes launched a simple tool that places submit- partnership (see go.nature.com/2kpjzat), as
no such limit: it merely requires that ted Stage 1 manuscripts in a public registry have Cancer Research UK and the journal
Nicotine & Tobacco Research5. More are in
the works.
F R O M P I LOT TO P R OTOC OL The lesson of RRs speaks to all areas of
science reform. Instead of forcing qual-
A brief history of Registered Reports ity to compete with success, partner them
up. Instead of pitting what is best for the
The potential of protocol registration to a variant focusing on replications at individual against what is best for all, create a
prevent publication bias and increase Perspectives on Psychological Science7. model that benefits everyone — the scientist,
rigour has been recognized for decades in The same year, psychologists Brian Nosek their community and the taxpayer — and
clinical-trials research. A format similar to and Daniël Lakens announced that a the rest will come naturally. ■ SEE TECHNOLOGY
Registered Reports (RRs) was also piloted special issue of Social Psychology would FEATURE P.297
at the now-defunct European Journal use the format to publish replications of
for Parapsychology in the 1970s to help important results8. Chris Chambers is a professor of cognitive
ensure publication of negative results. From 2014, more journals in neuroscience at Cardiff University, UK.
In 1997, The Lancet launched an article neuroscience and psychology began e-mail: chambersc1@cardiff.ac.uk
type similar to Stage 1 of RRs, which adopting and publishing RRs, and the
1. Nature Hum. Behav. 3, 763 (2019).
reviewed protocols of proposed research. format has now expanded across the 2. Allen, C. & Mehler, D. M. A. PLoS Biol 17,
Almost 150 were published before the life and social sciences. No specialized e3000246 (2019).
article type was discontinued in 2015, physical-science journals yet offer them. 3. Hummer, L. T., Singleton Thorn, F., Nosek, B.
A. & Errington, T. M. Preprint at OSF Preprints
ostensibly because other outlets served Some multidisciplinary journals — https://doi.org/10.31219/osf.io/5y8w7 (2017).
the same purpose6. including Royal Society Open Science 4. Ouares, K. A., Beurrier, C., Canepari, M.,
I began lobbying the editorial board of — have launched the format across Laverne, G. & Kuczewski, N. Eur. J. Neurosci. 49,
6–26 (2019).
Cortex to consider RRs almost as soon as I all subjects in science, technology,
5. Munafò, M. R. Nicotine Tob. Res. 19, 773 (2017).
joined as an editor. It gave the green light engineering and mathematics. I hope RRs 6. The Editors of The Lancet. Lancet 386, 2456–
in November 2012, and by March 2013 it will become an option in all mainstream 2457 (2015–16).
had adopted the full RR format. At around life- and social-science journals within 7. Simons, D. J., Holcombe, A. O. & Spellman, B. A.
Persp. Psychol. Sci. 9, 552–555 (2014).
the same time, a separate group launched ten years. C.C. 8. Nosek, B. A. & Lakens, D. Soc. Psychol. 44, 59–60
(2013).
1 2 S E P T E M B E R 2 0 1 9 | VO L 5 7 3 | NAT U R E | 1 8 9
©
2
0
1
9
S
p
r
i
n
g
e
r
N
a
t
u
r
e
L
i
m
i
t
e
d
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.