How to Manage Data: Data Stewardship and FAIR Skills

This post has been orig­i­nal­ly pub­lished on Prof. Lau­rent Gat­to’s blog.

In this inter­ac­tive talk at the Hong Kong Uni­ver­si­ty of Sci­ence and Tech­nol­o­gy Research Data Man­age­ment Sym­po­sium 2021, Lau­rent Gat­to and Mar­ta Teperek will join forces to offer some con­crete exam­ples on improv­ing research repro­ducibil­i­ty and trans­paren­cy. Lau­rent will speak from his own per­spec­tive as a researcher and will share some tips and tricks on how one can become a bet­ter sci­en­tist by apply­ing open and repro­ducible research prac­tices. Mar­ta will speak from the per­spec­tive of sup­port staff and will offer sev­er­al exam­ples of how insti­tu­tions can part­ner with researchers to help make research more repro­ducible and more trans­par­ent.

In our talk we will cov­er aspects such as peo­ple sup­port, poli­cies, train­ing, rewards and com­mu­ni­ty build­ing. The com­mon theme will be col­lab­o­ra­tion and part­ner­ship between researchers and sup­port staff.

We promise a lot of con­crete exam­ples and inter­ac­tion, and plen­ty of time for ques­tions!

The slides are avail­able here at bit.ly/202110RDM.

What it’s really about

The actu­al title of this talk comes in two parts:

Becom­ing a bet­ter sci­en­tist with open and repro­ducible research and sup­port­ing the sci­en­tist on this jour­ney.

and

Or… the impor­tance of col­lab­o­ra­tion between researchers and sup­port staff for sus­tain­able open and repro­ducible research prac­tices.

The goal is to have a joint sem­i­nar: to mix some of Laurent’s expe­ri­ences as an open researcher/teacher and Marta’s expe­ri­ence from TU Delft in pro­vid­ing the sup­port to help researchers achieve these goals. We hope that by doing this we would appeal to both researchers and to sup­port staff in the audi­ence:

• Researchers will get some tan­gi­ble exam­ples from Laurent’s expe­ri­ence.
• Sup­port staff can see how to best sup­port researchers with mak­ing their research more Find­able, Acces­si­ble, Inter­op­er­a­ble and Re-usable (FAIR).

Parts of this post are based on pre­vi­ous Laurent’s notes/talks from talks on Becom­ing a bet­ter sci­en­tist with open and repro­ducible research from May 2019 and Dec 2020.

Who’s here?

Speak­ers: we are Dr Mar­ta Teperek (Head of Research Data Ser­vices at TU Delft Library and Direc­tor of 4TU.ResearchData, The Nether­lands) and Prof Lau­rent Gat­to (Com­pu­ta­tion­al Biol­o­gy and Bioin­for­mat­ics Unit, de Duve Insti­tute, UCLou­vain, Bel­gium)

Audi­ence: a week or so before the sem­i­nar, we had 158 reg­is­tered atten­dees. Out of which 114 are research post­grad­u­ate stu­dents, 26 are HKUST staff/faculty, and 9 are guest par­tic­i­pants (librar­i­ans from oth­er HK uni­ver­si­ties). The break-down of HKUST par­tic­i­pants’ school affil­i­a­tions gives: Sci­ence: 48 (32%), Busi­ness: 13 (9%), Admin offices (library, research office): 12 (8%), Human­i­ties and Social Sci­ences: 19 (12%) and Engi­neer­ing: 59 (39%).

How it all started

From 2010 in Cam­bridge to 2021 in Hong Kong (vir­tu­al­ly).

In 2010 Lau­rent was doing his post­doc in the Depart­ment of Bio­chem­istry at the Uni­ver­si­ty of Cam­bridge. At the same time, Mar­ta was doing her PhD at the Gur­don Insti­tute, also in Cam­bridge. We were lit­er­al­ly next door neigh­bours, worked on sim­i­lar types of research (in fact Mar­ta even col­lab­o­rat­ed with Laurent’s group on some pro­teomics analy­sis) and yet we have only met each oth­er prop­er­ly in 2016 when our shared pas­sion for open and repro­ducible research prac­tice allowed us to con­nect in con­text of the Data Cham­pi­ons ini­tia­tive. Lau­rent was then a Data Cham­pi­on, and Mar­ta part of the Research Data Ser­vices at Cam­bridge.

What is ‘Open’, ‘Reproducible’ and ‘Good’ science?

• Poll: Do you prac­tice open and/or repro­ducible research?

• Word­cloud: Ask for a cou­ple of words describ­ing how par­tic­i­pants see open/reproducible research and cre­ate a word cloud?

Laurent

We can start with a def­i­n­i­tion:

Open science/research is the process of trans­par­ent dis­sem­i­na­tion and access to knowl­edge, that can be applied to var­i­ous sci­en­tif­ic prac­tices: open data, open source, open access, …

and

Repro­ducible research: there exist sev­er­al lev­els, of increas­ing dif­fi­cul­ty, that describe the action of using exter­nal data/software/material/informations to attempt to observe the same or com­pa­ra­ble results. For­mal­ly, we repeat, repro­duce, repli­cate or re-use depend­ing on how much of the orig­i­nal mate­r­i­al we have access to.

What I dis­like about the pre­vi­ous open science/research def­i­n­i­tion is that it can give the mis­lead­ing impres­sion that open research is about col­lect­ing badges, and that the more badges you pos­sess, the bet­ter an open researcher you are. And rec­i­p­ro­cal­ly, not hav­ing any badge to dis­play excludes one from being an open researcher. And as soon as peo­ple start to believe this, we will stop prac­tic­ing open research and will start doing stamp col­lec­tion.

Also, Open science/research can mean dif­fer­ent things to dif­fer­ent peo­ple, in par­tic­u­lar when declined along its many tech­ni­cal, admin­is­tra­tive, legal and philo­soph­i­cal attrib­ut­es.

An impor­tant word above is excludes: as thriv­ing open researchers, we need to under­stand that it isn’t only the dis­tance towards better/open research that we have trav­elled that is rel­e­vant, but that the start­ing point mat­ters a lot. The way some­body prac­tices open research, whether that per­son has the pos­si­bil­i­ty to imple­ment this or that open (and repro­ducible) research prac­tice, or whether they can be vocal about it most­ly depends on their envi­ron­ment and the sup­port or push back they get.

Embrace open and repro­ducible research to the extent you want and you can. Seek allies and sup­port around you, but do not feel pres­sured. It isn’t open or closed. It is cer­tain­ly not the same open or close for every­body.

So my very first take-home mes­sages are:

• Open and repro­ducible aren’t bina­ry, they are gra­di­ents, mul­ti­dis­ci­pli­nary and mul­ti­di­men­sion­al.

• How to be an open sci­en­tist and imple­ment RR:

• Let’s be open and under­stand­ing of dif­fer­ent sit­u­a­tions and con­straints, includ­ing our own.

And also:

Open != repro­ducible

Open != good (by default)

Repro­ducible != good (by default)

Open research and repro­ducible research aren’t the same thing, and one doesn’t imply the oth­er. Even though in our mod­ern under­stand­ing of these terms and con­cepts, they are inti­mate­ly linked, his­tor­i­cal­ly, they are very dif­fer­ent. And research being open or repro­ducible doesn’t make it good (what­ev­er the def­i­n­i­tion of good).

But open and repro­ducible research are sup­port­ed by good data man­age­ment (the top­ic of this talk/post) and lead to trust, ver­i­fi­ca­tion and guar­an­tees:

• Trust in Report­ing — result is accu­rate­ly report­ed
• Trust in Imple­men­ta­tion — analy­sis code suc­cess­ful­ly imple­ments cho­sen meth­ods
• Sta­tis­ti­cal Trust — data and meth­ods are (still) appro­pri­ate
• Sci­en­tif­ic Trust — result con­vinc­ing­ly sup­ports claim(s) about under­ly­ing sys­tems or truths

which are a hall­mark of good research.

From Gabriel Beck­er An Imper­fect Guide to Imper­fect Repro­ducibil­i­ty, May Insti­tute for Com­pu­ta­tion­al Pro­teomics, 2019.

People support

Laurent

Work­ing open­ly and repro­ducibly is para­mount for my own and my close collaborators/students’ ben­e­fit. Hence, it was nat­ur­al for me, when start­ing my research group, to cen­ter the lab’s activ­i­ties around the prin­ci­ples of good data man­age­ment to enable open and repro­ducible research that you can trust.

This is reflect­ed in the CBIO lab state­ment:

Open Sci­ence and Repro­ducible Research We are com­mit­ted to the open, trans­par­ent and rig­or­ous prac­tice of sci­en­tif­ic enquiry. In par­tic­u­lar, we make every pos­si­ble effort to make our research repeat­able, repro­ducible and replic­a­ble, in the hope that it can be re-used and improved upon by as many as pos­si­ble. Con­comi­tant­ly, we release all our soft­ware and data under open per­mis­si­ble licences. Final­ly, we will also ensure that our research (such as, but not lim­it­ed to jour­nals arti­cles, pre­sen­ta­tions, and book chap­ters) is pub­lished under open access licences to allow every­body to freely read, re-use and mine it.

Marta

To a lot of researchers the effort, but also the skills need­ed to effec­tive­ly man­age their research data and soft­ware might be insur­mount­able bar­ri­ers. And the truth is that we should not be expect­ing researchers to be excel­lent at every­thing: doing research, apply­ing for grants, man­ag­ing peo­ple, writ­ing papers, teach­ing, man­ag­ing research data and soft­ware. Sci­ence should be more about team­work, not about hero­ic efforts of indi­vid­u­als. And work­ing as a team and bring­ing dif­fer­ent types of exper­tise togeth­er, teams can be much more effi­cient at answer­ing the big and chal­leng­ing ques­tions. So researchers need to be sup­port­ed by skilled pro­fes­sion­als.

At TU Delft we have two main groups of col­leagues who pro­vide researchers with ded­i­cat­ed sup­port for data and soft­ware man­age­ment:

• Data stew­ards — at TU Delft we have one data stew­ard at every fac­ul­ty. Each fac­ul­ty spe­cialis­es in its own area of research and data stew­ard pro­vide expert advice in this research area. They are not only the go to peo­ple for researchers who need sup­port with data man­age­ment, but they also advise the fac­ul­ties on poli­cies, pro­vide train­ing and raise aware­ness about data man­age­ment and repro­ducible research with­in the research com­mu­ni­ty.
• Dig­i­tal Com­pe­tence Cen­tre Sup­port team — this team con­sists of a cen­tral pool of data man­agers and research soft­ware engi­neers. Mem­bers of this team pro­vide hands-on sup­port to research groups. They teach researchers skills nec­es­sary to make their data and soft­ware Find­able, Acces­si­ble, Inter­op­er­a­ble and Reusable (FAIR) by work­ing with them on their data and soft­ware. They help researchers cre­ate effec­tive data man­age­ment pipelines, make datasets more inter­op­er­a­ble, intro­duce entire teams to ver­sion con­trol for data and code, improve the qual­i­ty and sus­tain­abil­i­ty of their research soft­ware. Researchers apply to receive their sup­port and if such appli­ca­tions are grant­ed (com­pet­i­tive process), team mem­bers join their research groups for up to 2 days per week and up to 6 months.

All of these col­leagues have research expe­ri­ence and most have a PhD degree.

And ded­i­cat­ed peo­ple are also essen­tial to sup­port infra­struc­ture — to be the human side of it. At TU Delft we also make use of 4TU.ResearchData, which is a data and code repos­i­to­ry shared between four tech­ni­cal uni­ver­si­ties in the Nether­lands. It pro­vides all the state-of-the-art func­tion­al­i­ties, such as DOIs for data, long-term preser­va­tion, ver­sion­ing, Git inte­gra­tion, but what makes the biggest dif­fer­ence to the researchers is the ded­i­cat­ed sup­port of our data cura­tors.

Practice and policy

• Poll: Does your insti­tu­tion have poli­cies in place on data man­age­ment?

Laurent

Here’s an exam­ple of prac­tice from my research:

From left to right, we have an exam­ple of prac­tice of open and repro­ducible research:

• A repro­ducible doc­u­ment, pro­duced with open source soft­ware and curat­ed data,
• a preprint and
• a peer review­er paper.

But one might ask: does it take more time to work repro­ducibly?

No, it is a mat­ter of relo­cat­ing time!

From Five things about open and repro­ducible sci­ence that every ear­ly career researcher should know.

From per­son­al expe­ri­ence, I can say that poli­cies aren’t the main moti­va­tion to prac­tice good data man­age­ment. It is impor­tant for data man­age­ment to be an intrin­sic moti­va­tion, which can (or should) eas­i­ly be trig­gered by the desire to pro­duce trust­ed research.

Local action at a lab’s lev­el and/or at the ini­tia­tive of one or a few moti­vat­ed researchers (ear­ly career researchers and fac­ul­ty) is pos­si­ble, easy and effi­cient. How­ev­er, it is extreme­ly dif­fi­cult to expand out­side of one’s direct envi­ron­ment. For this, poli­cies are impor­tant.

Marta

If poli­cies are intro­duced top-down and just for the sake of hav­ing them in place (com­pli­ance rea­sons or ways for research insti­tu­tions to demon­strate they care about the good prac­tice), there is indeed a high risk that they would turn into a box-tick­ing exer­cise. Or worse, they turn into mean­ing­less doc­u­ments with the major­i­ty of peo­ple they are sup­posed to apply to, not even aware they exist.

On the oth­er hand, if poli­cies are intro­duced through con­sul­ta­tions with the com­mu­ni­ty and if they go hand-in-hand with prac­tice, they can become pow­er­ful tools for dri­ving cul­tur­al change more wide­ly, at the insti­tu­tion­al lev­el. At TU Delft we have a sep­a­rate set of poli­cies (!) for research data man­age­ment and a recent­ly intro­duced pol­i­cy and guide­lines for research soft­ware.

Our data pol­i­cy con­sists of a frame­work pol­i­cy which sets out the basic roles and respon­si­bil­i­ties for every­one at TU Delft: from every researcher, through deans, sup­port staff, and all the way to the rec­torate. Fac­ul­ties devel­oped their own data poli­cies, based on the frame­work pol­i­cy, where they pro­vide dis­ci­pli­nary spe­cif­ic inter­pre­ta­tions of the frame­work.

It took years of con­sul­ta­tions with all the stake­hold­ers at TU Delft to arrive at a pol­i­cy text with which every­one was hap­py with, as well as with con­crete roles and respon­si­bil­i­ties. Inter­est­ing­ly, researchers were some­times the main dri­vers for strict poli­cies. For exam­ple, some of the pro­fes­sors were real­ly con­cerned about PhD stu­dents leav­ing TU Delft with all the data, or leav­ing messy data behind. Hence, they were very keen on intro­duc­ing a data man­age­ment plan as a com­pul­so­ry deliv­er­able for PhD stu­dents dur­ing their first year Viva (go/no go assess­ment) and also to ensure that they upload their data to a repos­i­to­ry before grad­u­a­tion.

These two actions are dri­vers to bet­ter data man­age­ment. Since the intro­duc­tion of the pol­i­cy we have seen huge demand for train­ing on data man­age­ment, but also keep hear­ing from PhD stu­dents who say that they man­aged to intro­duce their super­vi­sors and some­times their entire research teams to bet­ter data man­age­ment prac­tices.

Anoth­er exam­ple is TU Delft’s Research Soft­ware Pol­i­cy. That’s anoth­er pol­i­cy which took years to devel­op and imple­ment and one which was dri­ven by a researcher who was fed up with TU Delft’s copy­right stance. TU Delft’s offi­cial stance on copy­right, prob­a­bly the same as of most research insti­tu­tions, was large­ly that TU Delft owns copy­right on soft­ware pro­duced by TU Delft researchers, and if researchers wished to pub­lish soft­ware, they need­ed to ask TU Delft for a writ­ten per­mis­sion through fil­ing an inven­tion dis­clo­sure form. A lot of unnec­es­sary bureau­cra­cy, frus­trat­ing for every­one.

Under the new soft­ware pol­i­cy, researchers who wish to pub­lish their research soft­ware under open licences, are auto­mat­i­cal­ly allowed to do so and TU Delft trans­fers copy­right to them. This not only huge­ly reduces the admin­is­tra­tive bur­den on every­one, but also pro­motes open source soft­ware prac­tices across the entire insti­tu­tion.

Training

• Poll: have you had any ded­i­cat­ed train­ing on data man­age­ment and repro­ducible research?

Laurent

I start­ed as a Car­pen­tries instruc­tor and now apply these lessons and best prac­tices in my uni­ver­si­ty cours­es (bach­e­lor and mas­ters in bio­med­ical sci­ences (for exam­ple here, here and here) and teach­ing these in work­shops for grad­u­ate stu­dents and ECR.

Teach­ing data man­age­ment, open and repro­ducible research prin­ci­ples when run­ning a lab is ide­al to host stu­dents that are well trained in RDM and RR are read­i­ly up to speed to start their research.

Marta

I whole­heart­ed­ly believe that it is essen­tial that ade­quate train­ing is avail­able to sup­port researchers in man­ag­ing their data and soft­ware. While researchers should not need to do every­thing on their own (there is also the need for peo­ple sup­port), hav­ing a sol­id back­ground in data man­age­ment and soft­ware man­age­ment skills is often essen­tial to prop­er­ly ben­e­fit from such sup­port and to make sure it leads to long-term ben­e­fits by becom­ing embed­ded in the researchers’ work­flows and prac­tices.

At TU Delft our Data Man­agers and Research Soft­ware Engi­neers reg­u­lar­ly receive requests for hands-on sup­port. But often for these tech­ni­cal experts to even get start­ed, it is essen­tial to estab­lish a com­mon under­stand­ing with the research team about what needs to be done and how to do it best. For this to hap­pen, they need to speak the same lan­guage. Sol­id soft­ware and data man­age­ment skills are essen­tial to do effi­cient and effec­tive research and to save a lot of time and mon­ey.

As nice­ly artic­u­lat­ed in the recent OECD report Build­ing Dig­i­tal Work­force Capac­i­ty and Skills for Data-inten­sive Sci­ence:

Aca­d­e­m­ic libraries (…) are a nat­ur­al focus for dig­i­tal skills sup­port and capac­i­ty build­ing (…). [They] train oth­ers in data and soft­ware prac­tices, par­tic­u­lar­ly in rela­tion to foun­da­tion­al skills and data stewardship.(…)…Libraries can be an impor­tant resource for uni­ver­si­ties to increase their dig­i­tal work­force capac­i­ties, pro­vid­ed that the nec­es­sary invest­ment is made.

At TU Delft we have devel­oped a shared vision on what kind of data and soft­ware man­age­ment train­ing should be pro­vid­ed to our researchers. The vision was pub­lished in 2019. Since then we have been work­ing hard to imple­ment this vision. Feed­back we have been receiv­ing on our cours­es has been real­ly good and empha­sised the need for train­ing. How­ev­er, we have already hit a capac­i­ty gap. Our sup­ply for train­ing is unable to meet the demand. So, as the OECD report stat­ed, invest­ment is nec­es­sary.

But as Lau­rent said, the Car­pen­tries are an excel­lent exam­ple of where col­lab­o­ra­tion between sup­port staff and researchers is essen­tial. To deliv­er the Car­pen­tries, the Library pro­vides the frame­work, pays for the mem­ber­ship, organ­is­es instruc­tor train­ing and coor­di­nates the organ­i­sa­tions. The cours­es them­selves are deliv­ered by a com­mu­ni­ty of instruc­tors, con­sist­ing of data stew­ards and researchers. Soft­ware Car­pen­try work­shops are typ­i­cal­ly dis­ci­pline-agnos­tic, but to deliv­ery dis­ci­pline-spe­cif­ic data car­pen­tries, such as the Genom­ic Data Car­pen­try or Data Car­pen­try for Social Sci­ences, it is essen­tial to part­ner with researchers who work on these types of data.

Rewards

• Word­cloud: What would moti­vate you towards being more open/RR? More cita­tions, bet­ter chances of get­ting hired, …

Laurent

Ben­e­fits for your aca­d­e­m­ic career: How open sci­ence helps researchers suc­ceed and more exam­ples from the Open as a career boost para­graph:

• Open access arti­cles get more cita­tions.
• Open pub­li­ca­tions get more media cov­er­age.
• Data avail­abil­i­ty is asso­ci­at­ed with cita­tion ben­e­fit.
• Open­ly avail­able soft­ware is more like­ly to be used. (I don’t have any ref­er­ence for this, and there are of course many couterex­am­ples).
• Ben­e­fit from insti­tu­tion­al sup­port of open research prac­tices

Net­work­ing oppor­tu­ni­ties (this talk here today)

See also Why Open Research

• Increase your vis­i­bil­i­ty: Build a name for your­self. Share your work and make it more vis­i­ble.
• Reduce pub­lish­ing costs: Open pub­lish­ing can cost the same or less than tra­di­tion­al pub­lish­ing.
• Take back con­trol: Know your rights. Keep your rights. Decide how your work is used
• Pub­lish where you want: Pub­lish in the jour­nal of your choice and archive an open copy. (See The cost of knowl­edge boy­cott of Else­vi­er).
• Get more fund­ing: Meet fun­der require­ments, and qual­i­fy for spe­cial funds such as the Well­come Trust Open Research Fund.
• Get that pro­mo­tion: Open research is increas­ing­ly recog­nised in pro­mo­tion and tenure. See also Repro­ducibil­i­ty and open sci­ence are start­ing to mat­ter in tenure and pro­mo­tion July 14th, 2017, Bri­an Nosek) and the EU’s Eval­u­a­tion of Research Careers ful­ly acknowl­edg­ing Open Sci­ence Prac­tice defines an Open Sci­ence Career Assess­ment Matrix (OS-CAM).

And of course the Five self­ish rea­sons to work repro­ducibly!.

Marta

There are also changes hap­pen­ing in the reward and recog­ni­tion of researchers at insti­tu­tion­al, nation­al and inter­na­tion­al lev­els. Inter­na­tion­al­ly, DORA has been the agent of change. There seems to be the moment grow­ing recog­nis­ing that pub­li­ca­tions and impact fac­tors aren’t suit­able indi­ca­tors of research qual­i­ty or impact and that pub­li­ca­tion pres­sure can have seri­ous adverse effects. More and more organ­i­sa­tions want to dis­cour­age the use of impact fac­tors in research eval­u­a­tion. DORA will devel­op a dash­board track­ing hir­ing and pro­mo­tion cri­te­ria across the insti­tu­tions. Fun­ders are also tak­ing impor­tant steps for­ward. For exam­ple, the Well­come Trust in the UK has been award­ing ded­i­cat­ed funds for Open Research since 2018. NWO, the main Dutch sci­ence fun­der not only has been award­ing fund­ing to open sci­ence and repro­ducible research, but also intro­duced ‘nar­ra­tive CVs’ for researchers who apply for grants. Insti­tu­tions also become increas­ing­ly pro-active at recog­nis­ing that change in recog­ni­tion and rewards is nec­es­sary.

But in my opin­ion the most impor­tant changes are the ones done by indi­vid­u­als. Every­one mat­ters, every­one can con­tribute to mak­ing a dif­fer­ence. When I think about the work that Lau­rent does… he now has two Mas­ter stu­dents in his lab who got inspired by his prac­tice and want­ed to do projects with him. We can all lead by exam­ple and should con­tin­ue doing this.

Here we should also remem­ber that recog­ni­tion also means recog­nis­ing the cru­cial work of sup­port staff, who are often impor­tant dri­vers for Open Sci­ence and Repro­ducible Research prac­tice. The work of data man­agers, data stew­ards, research soft­ware engi­neers, com­mu­ni­ty man­agers, librar­i­ans is some­times not vis­i­ble, but it is essen­tial and should be recog­nised as such. Here we would also like to thank the organ­is­ers for bring­ing us all togeth­er today and facil­i­tat­ing today’s dis­cus­sion on open and repro­ducible research prac­tices.

Community

Com­mu­ni­ty, peer sup­port and meet­ing great peo­ple. Open not only as in shar­ing, but as inclu­sive and wel­com­ing. ♥

Laurent

There is

Open Sci­ence as in wide­ly dis­sem­i­nat­ed and open­ly acces­si­ble

and

Open Sci­ence as in inclu­sive and wel­com­ing

Cit­ing Cameron Ney­lon:

And since 2010, Mar­ta and Lau­rent still work togeth­er 🙂

Marta

I ful­ly agree with Lau­rent about the impact and impor­tance of com­mu­ni­ties. In many uni­ver­si­ties there are now Data Cham­pi­ons, Data Ambas­sadors, Open Sci­ence or Open Research Com­mu­ni­ties… So join one if you have one! And if you don’t have one, no prob­lem. These com­mu­ni­ties are bot­tom up. This means you can sim­ply start from meet­ing for a cof­fee with a like-mind­ed col­league or two. That’s how Lau­rent and col­leagues have start­ed the Open­Con­Cam­bridge back then. Entire­ly bot­tom-up. You don’t need to wait for some­one else to get start­ed. Any oppor­tu­ni­ty to con­nect with like-mind­ed indi­vid­u­als shar­ing your thoughts and curios­i­ty about open sci­ence and repro­ducible work­ing prac­tices is a great oppor­tu­ni­ty. And who knows, maybe this webi­nar will be your chance to get start­ed?

And to give you some con­crete ben­e­fit… The jour­ney that Lau­rent and I had togeth­er start­ed in 2010. Now, in 2021, vir­tu­al­ly in Honk-Kong we are still friends and we now we will always sup­port one anoth­er and can rely on each oth­er. Also when prepar­ing this sem­i­nar.

Credits

Lau­rent: One of my advice was to make allies. I have been lucky to meet won­der­ful allies and inspir­ing friends along the path towards open and repro­ducible research that works for me. Among these, I would like to high­light Cori­na Logan, Stephen Eglen, Mar­ta Teperek, Kirstie Whitak­er, Chris Hart­genink, Naomie Pen­fold, Yvonne Nobis.

Mar­ta: Would like to give cred­it to numer­ous col­leagues from TU Delft and beyond who were the dri­ving force behind all the work described in this post, and in par­tic­u­lar: Alas­tair Dun­ning, Anke Ver­steeg, Con­nie Clare, Data Stew­ards (Diana Popa, Esther Plomp, Heather Andrews, Jasper van Dijk, Jeff Love, Kees den Hei­jer, Nico­las Dintzn­er, Robert Egger­mont, San­tosh Ilam­paruthi, Shali­ni Kura­p­ati, Yan Wang, Yasemin Turky­il­maz-van der Velden), Dig­i­tal Com­pe­tence Cen­tre Sup­port Team (Amir Fard, Ash­ley Cryan, Jose Urra, Julie Beard­sell, Manuel Gar­cia, Mark Schenk, Mau­rits Kok, Meta Kei­jz­er-de Rui­jter, Niket Agraw­al, Susan Branchett), Eiri­ni Zorm­pa, Emmy Tsang, Irene Haslinger, Karel Luy­ben, Maria Cruz, Paula Mar­tinez Lavanchy.