Skip to main content
SAGE
Search form
  • 00:06

    [Vas Taras Discusses Research Crowdsourcing, Data Sharing& Large-scale Collaboration][Why is learning about research methods important?Why is it interesting?]

  • 00:19

    VAS TARAS: The human species has been aroundfor about 200,000 years.[Dr. Vas Taras.Assisstant Professor, Bryan Schoolof Business and Economics, Universityof North Carolina at Greensboro]Different accounts, but that's whatscientists seem to believe.For the whole 200,000 years we'vehad roughly the same amount of knowledgeabout how the world works.May be a little bit more at some timesbut roughly the same amount of knowledge.Then somehow, 300 years ago, or so, somethinghappened that led to an explosion

  • 00:41

    VAS TARAS [continued]: in the amount of knowledge we haveabout how the world, both natural and social, works.And that one thing was basically whatwe call the scientific method.And so research methods are the foundationof the scientific method.So research methods is this wonderful toolthat allows us to understand how the world works.And most of us are interested in how the world works.

  • 01:02

    VAS TARAS [continued]: So this wonderful gift is basicallythe most wonderful thing-- very important and very exciting.So by learning the research methods, all of a suddenwe can increase our knowledge about the world.What can be more beautiful, more wonderful?So that is statistics that will advance knowledge.

  • 01:22

    VAS TARAS [continued]: [How would you define and describe the value of researchcrowdsourcing, data sharing, and large-scale collaborationto a student or early career researcher?]Again, I'm going to give you a historic perspective here.When you look at the quality of lifeof people all the way back, you willsee that there have been several big jumps in our life

  • 01:42

    VAS TARAS [continued]: expectancy and the quality of life we have.The first one occurred about 150 years ago.And that's when the Industrial Revolution happened.And the reason it improved our lives so muchis because before that we could rely onlyon our own muscle strengths.So we could only do what our bodies can do.The Industrial Revolution brought about machines,

  • 02:04

    VAS TARAS [continued]: and all of a sudden our power became virtually infinite.The steam engine, internal combustion engine,electrical motor all of a sudden made us very, very strong.And when you can look at the life expectancyaround that time, comfort of life,you would see that people startedliving almost twice as long.And they would have much more comfortable lives--big, big change.Then the next change happened sometime

  • 02:25

    VAS TARAS [continued]: in the 60s, 70s, 80s, and that's when wehad the computing revolution.That's where our computers, all of a sudden,made our brains much stronger than they were before.All of a sudden we could do millionsof computations per second.All of a sudden we became much, much smarter.And that's when you saw the saw that second jump in our qualityof life in a sense.The third revolution happened a few years ago,

  • 02:47

    VAS TARAS [continued]: and is basically still happening right now.And that's what we called the social revolution.All of a sudden our brains, our computersbecame interconnected.With my cell phone, all of a sudden,I can access the entire human knowledge database.I can connect to any human being on this planetvirtually for free.And it already has some important applications.

  • 03:07

    VAS TARAS [continued]: We already see how, for example, Wikipedia is basicallykilling the encyclopedia businesswhere crowds beat experts.We see that in all kinds of other areas of life--to some extent, not so much in science.I mean, some of that is happening in natural sciences,like for example, in astronomy, in biology, but notso much in social sciences.So far, unfortunately, my field, we

  • 03:28

    VAS TARAS [continued]: use the social media, the interconnectedness,to, I don't know, tell the world what I had for breakfast today.Not very useful, but I predict that in the next five,maybe 10 years, we will see a huge revolutionin how social sciences and sciences in general are done.By being able to connect to the crowd, to the world,we will be able to do research much, much better, muchmore efficiently.

  • 03:49

    VAS TARAS [continued]: We will increase the rate of discovery dramatically.And we are on the verge of some serious changes in how researchis done, how science is done.And as a result, we will see many, many more discoveriesthan we've seen before.So crowdsourcing is basically the platform, the tool,that will allow us to do that.[What kind of research questions are data sharingand large-scale collaboration uniquely positioned to help

  • 04:10

    VAS TARAS [continued]: answer?]There are several types of tasks, or projects,where crowds will consistently beat expert researchers.I'll go from the simplest one to the more complex one.First of all, when the task is extremely simple,you don't really need the crowd to complete it.

  • 04:31

    VAS TARAS [continued]: Like, for example, tying your shoes,you don't need a crowd to do that.Individually, you can do it better.Likewise, sometimes the task may be so, so difficult,that the crowd will not be of much use.Like, for example, if you need to design a nuclear reactor.Chances are no matter how big your crowdis, an expert nuclear physicist will do it better.But there are some tasks where crowdswill do a much, much better job than professional scientists.

  • 04:53

    VAS TARAS [continued]: For example, sometimes you would haveprojects where a large amount of work needs to be done quickly.A very good example of crowd source project in that areawould be Galaxy Zoo.So there was this one graduate student at Oxford,and he needed to classify galaxies-- many galaxiesfor his research.He spent several months trying to classify them.And he was able to go through a few thousand

  • 05:13

    VAS TARAS [continued]: of those pictures taken by, I believe, Hubble Telescope.And then his friend said, well, why don't youtry to crowdsource this task?And so they built a platform where people-- like my kids,for example, participated-- can go to the platform--to their website.And they would see pictures of galaxies one by one.And they would use simple scales like is it blue, red,or yellow?And they would click on the right color.

  • 05:35

    VAS TARAS [continued]: Is it spinning clockwise or counterclockwise?They just click.And so they could do those very simple tasks mainly for fun.And in a matter of months they wereable to classify millions and millions of galaxies--much more than any team of researchers could do.And so that's where crowds come useful in that respect.A second type would be what we call citizen science.So that's where you have kind of semi professionals

  • 05:56

    VAS TARAS [continued]: who help with research.For example, eBird is a very good example,where you have all those birdwatchersall around the world.And so they enjoy watching birds.They know the types, and species, and whatever.And so they're happy to share that information.And so biologists, climatologists, theyrely on amateur citizen scientiststo collect data for them.

  • 06:16

    VAS TARAS [continued]: And so again, no single lab can ever collect and observethose birds-- their movement, migration around the world.But when you have the crowd, all of a sudden,you can collect those data together.The third type is where you have--I call it the brute force.That's where you have-- the truth is,that when you do research, the best guy for the jobtends to work for somebody else.

  • 06:37

    VAS TARAS [continued]: So, for example, you need to analyze the data,or you need to come up with a better theory.Chances are, no matter how strong your lab is,there will be somebody out there whoknows how to do that particular task better.And so, by crowdsourcing the task, by posting your task,and making it open for the world, and asking for helpallows you to reach those people whoare uniquely suited to complete the task that you do not

  • 06:58

    VAS TARAS [continued]: have in your own lab.And so crowdsourcing here allows to increase, kind of,performance of your lab by reaching someonewhose peak performance is better than anyone elses in your lab.But, up until this point, that's interesting-- potentiallyuseful, but still not the most effective method to use crowds.The most effective method to use crowds

  • 07:18

    VAS TARAS [continued]: is to allow crowds to communicate with one another.And then you can really have a big revolution.You know what brainstorming, for example, is, right?So studies show again and again that whenyou have 10 individuals trying to produce ideas and thenbrainstorm individually, they can produce only x ideas.But when you put those 10 individuals in the same room,and ask them to do the same task,they can produce twice, sometimes

  • 07:39

    VAS TARAS [continued]: 10 times as many ideas.Because your idea gives me an idea.And my idea gives him an idea.And so, all of a sudden, we kind of build upon each others work.And, all of a sudden, the final solutionis much better than anyone could have come up with individually.So the peak performance of the crowdbecomes not the peak performance of an individual.But, again, studies show sometimes 10 and sometimes even100 times as good.

  • 07:60

    VAS TARAS [continued]: Because when we kind of learn from each other--when we even sometimes, kind of, steal each others ideas, allof a sudden, we can come up with somethingthat was much better than what I had as an individual initially.And that's where I believe the real powerof crowdsourcing, open data sharing,large scale collaboration will manifest itselfin nice, beautiful discoveries.

  • 08:21

    VAS TARAS [continued]: [What do you see as the key strengths of this kindof research?]Well, as I was saying, we already have enough evidencefrom experimental studies.We haven't quite used it in the, kind of,in the wild-- in the real world.But the biggest advantage of crowdsourcing is that not onlycan you reach the best guys for the job--

  • 08:42

    VAS TARAS [continued]: I mean no matter how much money you have--no matter how many people work for your lab,you cannot have everyone and everybody.And even if you could buy everybody,it wouldn't be efficient because keeping people on the payrollis expensive.So it's much better to hire, in a sense,or invite people to work on that taskas the need arises for that particular task.And so sometimes it would be a researcher at a different lab.

  • 09:03

    VAS TARAS [continued]: Sometimes it can be a talented student.Or maybe someone I know-- stay at home mom-- whojust happens to know the answer to that question.But as I said, the greatest developmentswill be in crowdsourcing that relies on what wecall competitive collaboration.That's where you will have crowds that are notonly individually submitting many, many solutions--and some of them happen to be better than what you could have

  • 09:25

    VAS TARAS [continued]: come up with at your own lab-- but that'swhen the crowds will communicate with one another.So members of the crowds will communicate with one another,learn from each other, compete against one another.And as a result, the final solutionwill be one that is built, kind of, standingon the shoulders of them-- not giantsin this case-- shoulders of the other crowd members.And so where you will see the final product to be much,

  • 09:47

    VAS TARAS [continued]: much stronger than anybody could have evercome up on his or her own.And so that's where I think I think the greatestdevelopments will be happening.And wonderfully, we do have the resources for that.I mean now we have the social media.We have the internet.We have the platforms.We have ever component we need for that.We just need to change the way we think about it.And we will see that revolution soon enough.

  • 10:07

    VAS TARAS [continued]: [What do you see as the challenges of conducting thiskind of research?]Challenges to research crowd source-- and thisis probably the most important question.As I said, we do have all the elements in place.We just have this one little challenge leftthat prevents us from using it.And that is the paradigm that the researchers got stuck,

  • 10:30

    VAS TARAS [continued]: in terms of how we do research.The problem is that we live in the publish or perish world.We are promoted, we are rewarded for publishing.And for a long, long time we thoughtthat the way to publish more is to work in secrecy--is to hide your data, not share your data with anyone.

  • 10:51

    VAS TARAS [continued]: Because if you share your data, somebody can steal those dataand publish papers that you could have published yourself.And that is dead wrong.It is so counterproductive.So wrong that-- I mean, I'm furious when I think about it.And the reason it is wrong is because whenyou share your data, you will not onlybe able to produce more knowledge-- because we

  • 11:12

    VAS TARAS [continued]: are hired as-- our job as researchers is to generateknowledge, not papers.And so by working individually, we generate less knowledge.And while that may-- some people believe--advance our personal agendas, it's definitely notserving the field as a whole.But even that is wrong.Because by sharing our data, by working with peopleon a large scale, collaborate in open source,we will actually publish more individually

  • 11:33

    VAS TARAS [continued]: and thereby advance our careers more.I'll give you just a couple of examples.I tend to share my own data on my website freely for anyone.Usually I would share the data that Ifeel I have no longer use for.I thought that I published everythingI could publish based on the data,so why not share it with the world?I tried it a few times.Every time it led to many additional publicationsfor myself.The way it works is this.

  • 11:54

    VAS TARAS [continued]: You publish your data.And then sometime later, somebody maybeuses those data for their statistics class.And then a student looks at the dataand sees a very interesting relationship.They contact you back and ask youhave you thought about that.I mean, it seems like an interesting relationship.And you look at that like, wow, really,I never thought about that.And then they said, well, do you wantto write a paper about that?You know the data better.

  • 12:14

    VAS TARAS [continued]: You know the field.Let's do the paper together.So we got one paper.Then another professor looks at the data and finds an anomaly.And you think, yeah, maybe it's some sort of an outlier.But then you look closer and all of a sudden yourealize that it's not an outlier.It's a very interesting case.You write another paper maybe with that person.Then somebody worked with you because they liked your data,but now they have an interesting idea that is not

  • 12:36

    VAS TARAS [continued]: related to that data set.And they still invite you to work on a paperbecause they like you now.And they know you now.And all of a sudden, from a data set that you thoughttold you all the stories there are, you have additional three,four, five, sometimes, papers.And so it definitely strengthens your own regiment,improves your own standing in the scholarly community.But on top of that, it increases the number of discoveries

  • 12:57

    VAS TARAS [continued]: the field made as a whole.So by being open about that, people will notsteal stuff-- some might, I don't know.But in general, it probably just willlead to more publications for everybody--including for you-- for more discoveries,for better process, and probably a greater chance of youpromote being promoted after all.So we have just to change that mentality that, you know,we have to cling to outdated, never share them, and work

  • 13:19

    VAS TARAS [continued]: individually.No, that's our data.Worked for us for 300 years, but it's time to change now.[Who would you consider to be key thinkers in this researchspace, or thinkers that students should be familiar with?]The field is very new.And I can't really say that thereare people who now have completelychanged our understanding.

  • 13:40

    VAS TARAS [continued]: But there have been a number of books on the topic-- mainlyrelated to technology.But they do make references to how researchcan be done that same way.And so I'm talking about books like, Wecanomics, HereComes Everybody, The Wisdom of Crowds.Probably the most profound one is Reinventing Discovery.So that one specifically looks at crowdsourcing in research.

  • 14:01

    VAS TARAS [continued]: Those are the books that are now shapingour way of thinking about the whole discoveryand collaboration paradigm.In terms of specific individuals,the natural sciences are way ahead of us.So they have what they call the Bermuda agreement--for example, in genetics research.So years ago, there was a conference somewhere

  • 14:21

    VAS TARAS [continued]: in the Bermudas where scholars gathered up together.And somebody said, you know what, it would probablyserve our field if we made it mandatory to share dataif you want to publish in our scholarly journals.And they discussed it and thought sure,let's give it a try.And what happened in the subsequent yearsis you saw this huge gap between the human genome research

  • 14:44

    VAS TARAS [continued]: and animal genome research.Two fields, same methodology, but one requires the data to beopenly shared and the other doesn't.In a matter of a few short years,you could see a huge gap in the rate of discovery.Then the animal research, or genome researchers,caught up and made the same request.And so now they're more or less at the same level again.But for those few years when they had different policies,you could see a huge difference.

  • 15:05

    VAS TARAS [continued]: And the Bermuda agreement-- I'm not sureif it was a particular thinker who came up with that ideaor it was the field as a whole, but thatwas a very good example.In our field- in terms of the exact developments-- Dr. PiersSteel of the University of Calgary,he is working on a project-- that he calls metaBUS, whichis basically a huge database of all published

  • 15:27

    VAS TARAS [continued]: research and social sciences thattakes the findings of numerical empirical findingsfrom the published papers-- and we're talking about hundredsof thousands of papers-- and basically codesall those findings.And so the platform allows us to do, what they call,a one minute meta-analysis.And if the viewers know what a meta-analysis is--it basically takes the research from previous studies

  • 15:49

    VAS TARAS [continued]: and integrates it, kind of averages it.And usually it takes a long, long time to find the studiesand code the data from those studies.So metaBUS kind of allows us to bringall those infinite amounts of knowledge in one database.And, in a sense, it's not your crowdsourcingas people collaborating.It's a crowdsourcing as people have an accessto work with other people, and being

  • 16:09

    VAS TARAS [continued]: able to integrate it in minutes, and thereby [INAUDIBLE].So I think that is one of the few examples in our fieldwhere you can kind of see the first sproutsof this line of work.So there will be more soon but not so much so farunfortunately.[Can you tell us a little of how the fieldof large-scale collaboration emerged?What developments have most influenced you?]

  • 16:31

    VAS TARAS [continued]: I should probably give you my own,kind of, historic perspective as to howI've been exposed to the ideas.And it, in a sense, mirrors how the field has been developing.So about 10 years ago or so-- maybe even 15--I got into what we call meta-analysis.And meta-analysis is a technique where you find publications

  • 16:52

    VAS TARAS [continued]: on the topic of interest.And you take basically correlationsfrom those studies.And then you, in a sense, average them.You adjusted for sample size differences and thingslike that, and get in a sense an averageof the previous publications, and thereby integratewhatever's been done, you know, by researchers before you.Now, when you look for studies for meta-analogies,

  • 17:14

    VAS TARAS [continued]: you often find papers that-- basedon the description in the paper--have the data that you need for your meta-analysis.But the data are not reported in the paper itself.So what I had to do for my meta-analysis often,I would have to contact researchers,authors of those papers, asking for those missing correlations.To my surprise, many people wouldrespond with here is my entire data set.

  • 17:36

    VAS TARAS [continued]: Why don't you just take and find what you need in the data set?I don't have the time to look for the specific correlation.That got me thinking that if you can get all those datafrom researchers-- you know, raw data-- why not maybeask for the data, not just correlations?And so for one of my projects about a decade ago,I had 857 papers total.And I contacted authors, over 1,000 of them,

  • 17:58

    VAS TARAS [continued]: of all those papers-- some papers were multi-authored--and asked for the data.And many of them said, sure, I mean,I'm going to share with you, the data.So I was able to collect over 100 datasets on culture-- culture in organizations--put them together, and that gave me a large data set.And then I reached out to those peoplesaying, well, now we have this database that nobody does.Do you want to maybe analyze them and write some papers?

  • 18:19

    VAS TARAS [continued]: And many of them said sure, yeah, let's do that.So we published a few papers out of the database.One of the big challenges was the paperhad so many co-authors, that many journal submission systemscouldn't handle that number.So we had to contact the editor saying well,you know, it just happens to be sothat it's a collaborative big project wherewe share the data.And so there are like 50, 70 co-authors.

  • 18:40

    VAS TARAS [continued]: And so would that be OK to have all of them on the list?And the editors were a little puzzled.But then they said well, sure, why not?I mean, viable method.So you see that happening more and more now.You see more and more scholars sharing their data,teaming up, and working together.Then you have many more scholars these dayswho are willing to share their data just

  • 19:02

    VAS TARAS [continued]: for free with the world.Like, for example, a lead projectthat is called X Culture, we have thousands of peoplefrom around the world who, kind of, in a sense,take part in a competition that is aimed toward solvingbusiness challenges.And so we have about 4,000 peoplefrom over 100 universities in 40 something countriesin a given semester.And so we collect immense amounts

  • 19:23

    VAS TARAS [continued]: of data about how those teams work.How those people interact.You know, how they communicate.Do they have any conflicts?Background personality, all that kind of stuff.And while the team of researchers at the X Cultureproject is fairly large, we know that our data contain many morestories than we can ever discover.And so time ago we started freely sharing our data saying,

  • 19:43

    VAS TARAS [continued]: hey, if you-- anyone, anyone, if you see anything in those datathat you would like to discover or research, we will share it.In fact, here are the data.Take them without even letting us know.And it seems like we're not the only ones whoare doing that now.Like, we have the world values survey, for example,that is releasing the data.So they're collecting wonderful data every yearor every few years.

  • 20:04

    VAS TARAS [continued]: And, you know, it's freely available,hundreds of publications coming out of that data set.So there is a shift happening now.And so it seems like that's what we'll see more soon enough.And hopefully with your help the idea will be further promoted.[Can you give us an example of any research that you areparticularly proud of?]

  • 20:25

    VAS TARAS [continued]: Well, the mega analysis is the one that I really am proud of.In fact, I probably shouldn't be proud of itbecause my contribution here is just pure luck.I mean, I just happened to need those data,and people were willing to share.So it's not like I invented the method.It just happened.But we got a number of interesting publicationsout of that.Then the X Culture that I mentioned,

  • 20:46

    VAS TARAS [continued]: so it originally was envisioned as a teaching tool in a sense.The idea was that we would have students,graduate students, undergraduate students from around the world.We would put them in global virtual teams.And they will work in those virtual, international teams,and learn the challenges of international collaborationfirsthand.But once we started getting all those data,

  • 21:07

    VAS TARAS [continued]: and once we started sharing those data with the world,the research side became a very important componentof the project.In fact, probably even more important than the teachingcomponent.And I'd like to, I don't know, credit myself, I guess,for my team, and be proud of the developments we do.Because, as I said, lots of papersare being developed based on our data.

  • 21:27

    VAS TARAS [continued]: Lots of discoveries made also including by non-scientists.Our students have access to the data likewise.And many interesting ideas come from peoplewho never thought they could make scientific discoveries.So again, hopefully, by being so big and being so international,we will be able to inspire other researchersto follow the [INAUDIBLE] and do the same thing.And hopefully, again, somebody will eventually

  • 21:50

    VAS TARAS [continued]: make it-- if not, mandatory, then convince the rest of usto that that's the way to go forward.[What recent piece of research on crowdsourcing, data sharing,or large-scale collaboration has had an impact on you and why?]The biggest impact on me had been the World Value Surveyproject.I tend to study the effects of culture

  • 22:12

    VAS TARAS [continued]: in the organizations and the work settings.And there have been a number of large scale projectswhere people measured cultures around the world.We have obviously the IBM study by Hofstedefrom the 60s and 70s, and then published in the 80s.We have a wonderful project by the Globe Team,where they also measured cultures in 63 societies--

  • 22:33

    VAS TARAS [continued]: if I remember correctly.And then we have the World Values Survey.And the difference there is that Hofstede never reallyreleased his data.In fact, many people tried to get hold of those data.And they wouldn't be shared.The Globe study now recently kind of shares the data,but there is a very, very complicated processto request those data.And so apparently it's not as freely available

  • 22:53

    VAS TARAS [continued]: as it would like to be.And the project has been around for a long time.And for the first, at least, you know, ten yearsor so, the data were not available to the world.And then we have the World Values Surveyon the other hand, where the data hadbeen published-- you know, once collected,published right away.When you look at what those data produced as far as discoveries,as far as publications, there is a huge, huge gap.

  • 23:13

    VAS TARAS [continued]: The IBM data, Hofstede.Hofstede had published a few papers, books, I mean,that's it.I mean how much can an individualor a group of individuals do?The GLOBE study has more people on their team,so they have a few dozen scholars.And so they were able to publish a few dozen publications.Again, remarkable, but, again, you can expect only so muchfrom a limited group of researchers.

  • 23:34

    VAS TARAS [continued]: The World Values Survey opened up to the entire world--like literally entire world.Anyone can download those data at any time for free.And you see literally hundreds, if not,thousands of publications coming out based on those data.And so that's where you will see exactly the same effort goinginto collecting the data and such a huge differencein the outcome, in the quality of the outcomes,in the quantity of the outcomes.

  • 23:54

    VAS TARAS [continued]: And that's where I really started thinking hey, I mean,come on.We do the same thing, so why not use the World Values Surveymethod as opposed to the method or the approachto data sharing by the other teams?Because this works definitely better.It gives more credit to the creatorsof the data or collectors of the data,as well as it creates many more publications or discoveriesfor the field as the whole.

  • 24:15

    VAS TARAS [continued]: [What advice would you give to a student embarking on a projectinvolving these elements for the first time?]Do not be afraid.Do not be intimidated.I thought that when you approach people--especially big names in our field-- and ask for their data,they will not even reply to you.Or if they reply, they will say no.

  • 24:37

    VAS TARAS [continued]: I was very, very wrong.I was surprised how many people responded.Many people would respond literally within minutes.And they would say, oh sure, yeah, come on, yeah.You can use my data, or here is the data set.Or if there is anything I can help you with,sure, I'll go ahead.It seems like as long as you reach out-- eventhough the way the field works.The paradigm is outdated.Individual scholars recognize that it's outdated

  • 25:00

    VAS TARAS [continued]: and are willing to make the step forward.They just need that push.That is, just need someone to approach them and say,would you like to?And they would say yes.So students often are intimidatedby the size of the name or prestige some scholars carry.And they think, oh, I cannot approach them,ask for the data.Or I cannot offer them my idea when I read their papers.You know, maybe I have a better idea.

  • 25:22

    VAS TARAS [continued]: And so maybe don't be.People are very receptive of the whole way of thinking--of open collaboration.Try, some probably will not respond.Some probably will not agree.But there will be many, many, many who will.And so as a student-- myself, at that time--I was surprised how many people were

  • 25:43

    VAS TARAS [continued]: positive about my approach.And I think today it will be even more open and even morepositive, so don't be afraid.Give it a try, it will work.[What are the common mistakes made by new researcherswhen doing research crowdsourcing, data sharing,or large-scale collaboration for the first time?How might they be avoided?]I wouldn't call it, perhaps, a mistake.One of the errors we make-- in terms of assuming what's

  • 26:05

    VAS TARAS [continued]: going to happen is-- first, we believe if we share our data,somebody will steal them and will publish papersthat we could have published.I've had a number of papers with which youwould call to be crowdsourced.And I'm yet to see a single conflict among the co-authors.The main kind of thinking is that when you work either

  • 26:25

    VAS TARAS [continued]: in a large team-- or if you publish your dataor open your data-- somebody will comeand do something unethical-- will take your dataand will not give you credit.Or if you take somebody's data, they will get upset.And it actually seems to be the opposite.So people, especially researchers-- I mean,these are serious people with terminal degreesin their field.They tend to be, in my opinion so far--

  • 26:47

    VAS TARAS [continued]: that's what I see-- very ethical, very collegial.And when somebody asks for my data,they tend to offer something in exchange.They would say, well, maybe we should work together.Or when I want to use somebody's data,people seem to be fine with that.And so as long as we all act in good faith,it seems to work very well.And so the mistake, I guess, hereis to think that it will lead to problems.

  • 27:07

    VAS TARAS [continued]: That it will make matters more complicated.That it will lead to some sort of legal issues.Who owns the data?Who owns the publication?And so many people are just afraid to make that step.That is a mistake.I believe you should try it.My experience has been extremely positive.You occasionally encounter some free riders, perhaps,in a large team who are trying to get a co-authorship

  • 27:27

    VAS TARAS [continued]: without doing much.But that happen so rarely.And, you know, the team-- the crowdusually weeds those people out so quickly that I wouldn't evenworry about that.Most people, at least in my experience,are very ethical, very collegial, very open.And it just seems to work very, very well.So it's a mistake to think that it won't work because it will.

  • 27:48

    VAS TARAS [continued]: Give it a try, it will.[What does being an ethical researcher mean to you?Why are ethics important?]Ethics in research, especially whenyou have crowd-sourced research is probably the foundationthat may make it possible.In fact, the concerns about ethicsis what prevented this line of approach

  • 28:12

    VAS TARAS [continued]: to research from going forward.You have several issues that potentiallycould be problematic.One, once you open your data, the ownership of the databecomes an issue.So who owns the data?Can anyone take the data without letting you know?Can anyone publish papers without letting you know?Who's going to be co-author or of that paper?Second, when you have a crowd, you all of a sudden

  • 28:33

    VAS TARAS [continued]: have, what I call, vaguely defined co-author teams.When you work alone, it's clear you are the author.When you have a small team of researchers it's usuallyclear-- you know he, he, or she, she, she-- those five peopleare co-authors.When you have a crowd where some people makevery marginal contributions, some make larger contributions,it's very hard to decide where the co-authorship stops.

  • 28:54

    VAS TARAS [continued]: I mean, who is a co-author formallyand who already is not a co-author?Or who is not making enough contributionsto be a co-author?And so that sometimes potentially couldlead to conflicts, tensions, and other problems.And so that's where ethics becomes paramount.There are really no laws or rules

  • 29:14

    VAS TARAS [continued]: about how it should be governed.In fact, the situation is so complex,I don't think we will ever have precise guidelines on howto handle those matters.And so researchers have to rely on, basically, ethics.So we just think is it ethical to do it that way?Would it be right or wrong to include that personor not to include that person as a co-author?Would it be right or wrong to use those data

  • 29:35

    VAS TARAS [continued]: without letting everybody know or the author know?So that's where we have to rely on,basically, our ethical sense.Thinking on our good faith to do that kind of research.So being ethical is extremely important in that sortof research.[Can you give us some examples of any negative or positive

  • 29:56

    VAS TARAS [continued]: experiences that you have had while doing research?]I haven't really had any very negative experiences.So I can give you a few positive.And then maybe can talk about some limitationsthat we encountered along the way.So some of the positive experiences,as I said, in mega analysis, I was very surprisedthat people were so willing to share the data.

  • 30:16

    VAS TARAS [continued]: Including some of which I would callthe biggest names in our field.And in that time I was just a graduate student.I mean, I thought who would care about me?I also had a number of projects where I would use the crowdto complete certain tasks.Like, for example, in the X Culture project,we collect all kinds of data-- over 2,000 variablesthat we track longitudinally.Some of those variables are what you call qualitative.

  • 30:37

    VAS TARAS [continued]: So you would have open comments, experiencesharing by the participants of the project.And those are very difficult to quantifybecause you have those paragraphsand paragraphs of text, but how do you classify them?And so that's where you need a lot of manpowerto read those comments and classify them in terms ofis it a positive or a negative comment?Is it about the task or personal relationships?

  • 30:59

    VAS TARAS [continued]: All those kinds of things.I've had a very positive experience crowdsourcing.And so we have thousands of peoplewho took part in the project-- about 20,000 of them.I sent out an email to all of them asking,would you guys like to help us with classifying those data--quoting those data?You will see comments one by one.And you will have a bunch of questions.And you'll just need to click on and decide what it is.To my surprise, so many people replied, like,

  • 31:21

    VAS TARAS [continued]: literally thousands of people.Some classified only five or 10 comments.Some did hundreds.But as a result, we were able to classifyimmense amounts of data that I or my teamwould never, ever be able to do on our own.So that is a very positive experience.Same thing with multi-authored teams.I mean, when you have-- what is it?Two heads are better than one, but 2,000 headsare better than two.So when you have larger teams, you usually

  • 31:42

    VAS TARAS [continued]: tend to have a better product.The negative sides-- and it's a minor--I wouldn't call it a negative.It's just one of the challenges-- is free writing.So when you have a large team, it'shard to monitor performance of everyone.When you have a crowd, especially if it'sin hundreds or thousands, it's impossible to keep trackof everybody's performance.And occasionally you have people whosign up to participate and then drop out, which is OK,

  • 32:05

    VAS TARAS [continued]: but then come in the end and say,I did participate at some point.I still want to be a co-author.And so it doesn't really lead to big problems.But sometimes may lead to some slight tensionor disappointment of other coauthors who contributed moreand they feel like they could have gottenthe same credit for doing less.So again, it doesn't happen often, but sometimes it does.And that's where the issue of ethics comes into play.

  • 32:28

    VAS TARAS [continued]: So as long as people are ethical, it doesn't happen.Occasionally, it doesn't happen to be that way.But again, I wouldn't get discouraged justbecause of those little challenges.Happens everywhere in any approach,so not any worse here.[What are the practical benefits of studying this kindof research for a student's academic or professionalfuture?]

  • 32:49

    VAS TARAS [continued]: As I was saying before, we are on the vergeof a true revolution in how discoveries are made.And so students will become researchers or professionalsin their field soon enough.And probably their career advancementwill depend on their ability to generate new ideas,new solutions-- be it managerial ideas,

  • 33:12

    VAS TARAS [continued]: or technological ideas, or scientific ideas.And so if they can embrace crowdsourcing, opencollaboration, open data sharing early on,I know it will propel their careers to the new heightsmuch faster and much more effectively.They will become more known.They will benefit themselves as well as the whole field.

  • 33:32

    VAS TARAS [continued]: And so that's where I see the greatest developments.And I encourage the students to use Facebookand other social platforms-- not onlyto post pictures of what they had for breakfast,but also to reach out and find coauthors,find partners, crowd source their ideasor crowdsource their tasks.

  • 33:52

    VAS TARAS [continued]: Because the platform is suitable for that taskand can be used much more effectivelythan just sharing some useless pictures--funny cats for example.So the sooner they understand that,the better they will be off in their lives and careers.[What new research directions do you find most exciting?Where would you like to take your own research?]

  • 34:14

    VAS TARAS [continued]: I really would like to contribute to convincingthe leaders in the field.And by the leaders I mean the administrationof the Academy of Management, the Academy of InternationalBusiness, Psyops-- Society for Industrial OrganizationalPsychology-- other professional fields, organizationsand societies in my field to make

  • 34:36

    VAS TARAS [continued]: the step that the natural sciences made some time ago.And that is to change the policiesabout how research is done.I'd like to see-- and I hope it will happen soon.And I know when it happens it willchange the rate of discovery.It will improve the rate of discoverydramatically-- to encourage the field to accept the idea

  • 34:56

    VAS TARAS [continued]: and to change the policies that would require that scholarsshare their data openly.When you publish a paper, you have to release the data.So others can take a look at your data, check your math,try to replicate your study, perhaps findnew stories in those data.I'm not saying that the data should be stolen freely.I'm saying that the data should be published.And, in fact, that will protect your ownership

  • 35:18

    VAS TARAS [continued]: because everybody knows who the author of that published dataset is.It's like patents.Once you have the patent, nobody can steal your idea.Until you have the patent, people can steal your idea.So my hope is that the way forwardwill be changing the policies of journals,changing the policies of professional associations--that will favor open scale, large scale collaboration,

  • 35:39

    VAS TARAS [continued]: open data sharing, crowdsourcing.As well as, my hope is that therewill be more and more individuals or teams thatwill create platforms for that.Because, yes, you can use Facebook, Twitter, Dropboxfor that kind of research, but they're notspecifically tailored for research crowdsourcing.I'm hoping we will see some dedicatedplatforms for that kind of approach soon enough.

  • 36:00

    VAS TARAS [continued]: So that it will be not only possible, but very easyto do so.And hopefully, we'll soon enough have,like, a Craigslist of research crowdsourcing where peoplecan post challenges, and solve challenges for others,and form big, vaguely defined co-authorteams, and stuff like that.So I predict it will happen in the next few years.I'm not sure if it will be an individual team thatwill make that change, or it will

  • 36:21

    VAS TARAS [continued]: be several people in the same time doing the same typeof work around the world.But it will happen.And when that happens, we will see some wonderful things.

Abstract

Dr. Vas Taras describes his work in large-scale, crowdsourced research. He explains that this approach requires that researchers let go of old ideas about holding tightly to their data in order to brainstorm and collaborate more effectively with others.

Looks like you do not have access to this content.

Vas Taras Discusses Research Crowdsourcing, Data Sharing & Large-scale Collaboration

Dr. Vas Taras describes his work in large-scale, crowdsourced research. He explains that this approach requires that researchers let go of old ideas about holding tightly to their data in order to brainstorm and collaborate more effectively with others.

Copy and paste the following HTML into your website