While I write my discussion I'm going to add to this thread.
1 #dissertation in numbers....
3 years. 4 countries. 1 topic.
The STROBE #reportingguideline
(drop the bass :P)
#phdlife #phdchat #AcWri

I'll be largely covering....
1 protocol http://dx.doi.org/10.1136/bmjopen-2017-019043
1 research letter http://dx.doi.org/10.1097/EDE.0000000000000899
2 articles https://doi.org/10.1016/j.jclinepi.2018.11.006
http://doi.org/10.1016/j.jclinepi.2019.07.019
{1 more in #peerreview that I presented at @METMeetings last week} and all the good stuff along the way...

All #research should start w/ a #protocol
https://bmjopen.bmj.com/content/7/10/e019043
13 STROBE extensions identified
- 7 published before Mar 17 (time buffer for journal uptake)
- 6 extensions published in 16 (busy year!)

2 objectives:
-investigate content (all 13)
- journal endorsement (7)

http://dx.doi.org/10.1097/EDE.0000000000000899 @EpidemiologyLWW
298 coded additions from extensions to STROBE #reportingguideline checklist
112 (37.6%) were general #epidemiology /methodology guidance. Should really be in STROBE, not in extensions...
Communication probs? Gaps in content? red flags?

Next, we focused on...
7 STROBE extensions. A #bibliometrics study investigated
*how* (the language and its strength --> recommend or require, vague or specific) and
*if* journals were endorsing them (and STROBE) https://linkinghub.elsevier.com/retrieve/pii/S0895435618305377 #reportingguidelines #AcWri

3172 journals were initially identified in
7 areas ( #antimicrobialresistance #infectiousdiseases #molecularepi #rheumatology #geneticassociation #healthdata & #responsedrivensampling)
After screening, we had
257 unique journals
https://osf.io/jxep8/  #dataset #opendata

Sidenote. #observationalstudy & field-specific search strategies were run in #medline to ensure that identified journals published relevant studies for a certain STROBE extension.
94,000 abstracts were identified
(And I found out #excel & medline have line/ download limits )

>24,000 *mostly* #observationalstudy abstracts are on @OSFramework https://osf.io/hp623/  along with
the coding #dataset for the
257 journals #openscience #opendata
https://osf.io/eqyg6/ 

5 major types of "endorsement" were identified.
Ranging from strict requirements, simple direct suggestions, vague mentions to #reportingguidelines & even vaguer notes of resources that encouraged use.
A lot of information was in #editorialpolicies not #authorguidelines

Endorsement literature largely does not make these distinctions or provide their definitions of "endorsement"
nor do they generally look at information other than #authorguidelines
All ref documents from the journals are also #opensource #journalology https://osf.io/hp623/files/ 

139/257 (54%) didn't endorse STROBE in any manner
171/310 (55%) didn't mention relevant extensions
Those endorsing STROBE were more likely to also endorse other #reportingguidelines (CONSORT, PRISMA) & ref @C0PE #ethics #publicationethics
Good pub practices should join forces!

So in sum, if you are going to study endorsement of #reportingguidelines
BE CLEAR WHAT YOU MEAN (suggest/require/vague references to external resources) and
LOOK EVERYWHERE (the information might not be in an obvious place --> journals, y'all should fix this...)

Those working on 1 specific #reportingguideline,
don't restrict yourself. Strength is in numbers. Endorsements often come ask a package deal. The ovearching ideals of transparency and reproducibility can get you far #openscience @CONSORTing @PRISMAStatement @C0PE

I previously did a #tweetorial that summed up the quantitative results from the #onlinesurvey so I'll just direct you to that while I keep writing #500wordsaday https://twitter.com/sharpmelk/status/1159418840121524224

@threader_app compile