 Good afternoon and hello from Hong Kong, first of all thanks to the Australian National Data Service and the Council of Australian University Librarians for giving me the opportunity to speak to you all today, I do hope that you find it useful. Before I start, a bit about HKBU, you can see here our University Library, the Auschwitz Hong Kong Memorial Library, we were founded back in 1956, we are one of the eight publicly funded universities in Hong Kong, we're relatively small, we're certainly the smallest comprehensive university in Hong Kong, we've got about 10,000 students spread across seven faculties in schools and about 800 faculty and about half of those faculty are expected to engage in research as part of their official duties, now in common with tertiary education systems around the world, there's increasing pressure in Hong Kong for universities to demonstrate high levels of research performance, so research is highlighted in our strategic plan, so hence interest in initiatives like ORCID, so here's my outline for this afternoon, just I'm going to talk first about the origins of the ORCID project at HKBU and what the planning of that involved, I'll then explain in a bit of detail the create on demand workflow and why we chose this method without ORCID adoption and rejected the alternative which is a create on behalf or batch creation approach, then I'll talk about the rollout, talk about how we promoted ORCID to our faculty and the support that we offer and continue to offer them and I'll round off by talking about the current status of our ORCID rollout and our plans for moving forward, so about the origins of our project, it's actually ORCID or the ORCID initiative at HKBU, it's part of a package of other initiatives that we call the research visibility project and we can trace this back to, came out of a really informal conversation actually that we had with a couple of key faculty members and our graduate school back in December 2013, so we just sort of mentioned, the library mentioned ORCID in that conversation and to be frank we didn't know that much about it at the time but our colleagues from, our academic colleagues were very interested in it, there was initial, there was an initial sort of voluntary call to sign up, that didn't have too much impact, it was very informal, the vice president for research sort of just sent an email out and said sign up for ORCID but that didn't have much of an impact at the time but as a result of the interest you know shown by our vice president for research and development, the library put together a more robust proposal for this research visibility project and apart from ORCID this included data cleanup in major citation databases, scopus and web of science and also the further development of our institutional repository. Now I apologize, at this computer I didn't have, I don't have a webcam so I wasn't able to be alongside Paul there for the introduction but I did include a photo here, you can see me, there I am on the right hand side so this is our RVP team, Angela and myself, Angela is second from the left there, we are both information services librarians in the university library and our work on this project is in addition to our regular duties and this was a bit of a challenge for us because this project has been substantially different from our usual reference and instruction work, the other two people in this picture are our research assistants Simon and Maggie who are actually recruited for this project using funding allocated by our vice president for research and they work more or less full time on this project. So we are a small team but we had a pretty big idea, so how did we go about planning for this implementation? We did spend a really significant amount of time planning our rollout, once we got our research assistants in place and trained by July 2014 and our Orchid invites didn't go out until January of this year so there was essentially a full six months just preparing for the rollout of Orchid. I personally think this level of planning was necessary and appropriate at least for our institution, I've heard of others who've acted more quickly rolling out within one month after the decision had been made to adopt Orchid but I personally again have reservations over the effectiveness of rolling out Orchid so quickly. We were able to be much more methodical and this allowed us to change direction when it made sense and I'll highlight a particular instance of that which is the choice of workflow and I'll elaborate more on that later. So of course the first thing we needed when planning is some clear goals for the project so this is what we set out to do at the beginning. First to a pretty straightforward, we wanted all of our faculty to have an Orchid ID. We also wanted the university to have a record of what that ID was. The last part was for Orchids to be pre-populated and this is where we saw a real opportunity to add value as an institution to the Orchid creation process because it's not strictly speaking necessary to add past publication data to Orchid. Faculty usually want this information included but at the same time, if I'm sure you've had these conversations with your own faculty, many of them are sick and tired of being asked to input data into different systems again and again so doing this on their behalf using data already held by the university was a really major selling point for this project for us. So how exactly were we going to go about doing this? What infrastructure did we need for this? Admittedly our initial conception of the project, we weren't too clear on this in our initial proposal. It was only really after we began the detail planning it became obvious that we would need some sort of internal system to organize the myriad data that we would be collecting as part of the project. So we're very grateful to the system section within the university library. They were fantastic in providing their help and expertise in creating a database to manage these data about our researchers and their publications and we call this a very creative name, we call it the RVP database. So this is what it looks like, there are basically two main sources of data. One is our personnel office, this is where we get basic information about our researchers like names, position titles, etc. The other major source of data is our graduate school, that might be a bit confusing for those not familiar with the institution. At HKBU our graduate school is actually responsible for a lot of research stuff as well. So they're responsible for collecting publication data from the faculty and they've done this for many years but they don't really do much with it, it's really just for reporting purposes so it's to report back to the research, the government research funding body in Hong Kong. So apart from that not much else is done with the data. We did find that these data were often very messy or incomplete because some faculty or usually their research assistants or administrators were all doing that independently with not too much oversight. So one major task for our RVP team was to check and clean up all that data to ensure that it's complete and accurate. So at this stage our RVP database contains records for around 420 researchers and almost 5,000 publications so that took time to set up but again I think it was essential to this project providing the necessary added value. So on to that dilemma we faced in the beginning which was the first really big decision we had to make was how to approach the ORCID account creation. This is a really crucial question to get right early on because it's difficult to change your approach once you've chosen which way you're going and the decision needs to be made before you work with the ORCID team on setting up your systems integration with the ORCID system. So our original intent was to create accounts on behalf of the faculty because one of our guiding principles for this project is to reduce as far as possible demands on faculty time. So our natural first thought was to create ORCID accounts on behalf of the faculty. I'm sorry if this photo is confusing I was looking for a photo of someone giving an ORCID to someone. I don't think those are ORCIDs but that was the closest I could get. So we wanted to offer the gift of ORCID to faculty members. So what would that approach look like? This batch creation creating accounts on behalf of faculty. So in this model the request to create would be initiated by our system which is embedded in the IRVP database. So we would make that request to create to the ORCID system and the ORCID account would be created and the account number would be returned to our database. At that point the ORCID does not go live because the ORCID system at the same time will then ask the faculty with using the email address that we provide to them to verify the accounts. And if they do not claim that ORCID within 10 days it will become public but then the institution is left responsible for maintaining the ORCID record. And there are a lot of risks with this approach. The key one being the low verification or low response rate. So during our planning phase we spoke with colleagues at the University of Hong Kong who had done this already and they had used this approach. And despite their best efforts you know in reaching out to the faculty and explaining what was happening they reported a verification rate of just 30 percent. So you know the rate was very low. And this approach can easily result in problems with the duplicate ORCIDs because the system will not be able to tell if faculty have already created ORCID IDs using a different email account. Maybe they use their personal email account to create an ORCID. So you might be creating ORCIDs for people who already have ORCID IDs. And then you have to get into merging those accounts and it's very messy. And last but not least the approach is rather passive in my opinion and does not encourage engagement with the faculty. So for most organizations ORCID themselves do not recommend this approach. They do support it. For some organizations it might make sense but it's generally not recommended. And certainly that's been our experience talking to other universities and from our own experience. So based on those conversations that we had with Hong Kong U colleagues as well as advice received from ORCID we decided to take a different approach which is the create-on-demand workflow. So here's a similar sort of diagram but with the create-on-demand approach. So there's three players in the system, the RVP database, the researcher and the ORCID system. The difference here is that the creation process is initiated by the researcher through a link that they will receive from us. So we customize a link for them and the email is sent to researchers and they're asked to actually initiate the creation process. So though there's a link in that email which I will be able to show you later in some later slides what exactly that looked like. The request to create comes from the researcher to the ORCID system. The ORCID account is then created and the account information is sent back to us along with an access token which allows us to add and upload publication information on the researcher's behalf and if the faculty doesn't change the default settings this permission is given via a so-called long-life token that lasts 20 years is my understanding so we can keep updating their record for that long. But that permission can be revoked by the researcher at any time. So once we have that permission we can upload the information to the ORCID record. There are lots of advantages to this approach. The key one being that the researcher has to be engaged in the process otherwise no account is going to be created. So without them actively verifying and saying yes I want this account nothing will be created for them. So this is a contrast to the create on behalf approach which can result in a large number of unverified ORCIDs that will never be used. The risk of duplicates is largely eliminated as well. If the faculty already has an ORCID ID they can simply sign in with their existing credentials and we will still get their ORCID ID and we will still ask them for permission to upload their publications to that ORCID ID. So I think by requiring faculty to take positive action this approach is more likely to result in faculty understanding and invest them more in the ORCID initiative. So that's just a quick overview of our setup. I'd like to talk now a little bit about how we got the word out about ORCID because faculty are always busy with their teaching, research, service activities and they'll probably not pay very much attention to new initiatives if they've not been made aware of the potential benefits. So before sending out the invitations we spent a lot of time and energy promoting ORCID and that was another benefit of the relatively lengthy gestation period that we had for this project. So one thing we did was attend a board meeting of every faculty in school at the university, not just myself but the university librarian also came along just to emphasize the senior management support for this project. We learned a lot from the questions that the faculty asked at these meetings. They made clear we're under pressure in terms of time, we don't have time for more administrative tasks. We also had an opportunity to maybe dispel some misconceptions about ORCID. A lot of people looked at ORCID, they sort of looked at a screenshot of an ORCID account and they said, well, I've already got a Google Scholar profile I already use academia.edu or research gate. I don't need another research profile system. So we hammered that message that ORCID is not another research profile system. It's becoming part of the global research infrastructure. It's being used by publishers and research funders to facilitate the scholarly communication workflows to get that message out. ORCID is not just a research profile system. Other forms of publicity, normal sorts of things, souvenirs, newsletters, information, sessions. Most of these efforts really emphasize the support of the senior university managers for this initiative. One thing I want to spend a bit of time on is the video tutorial that we created for this project because that'll give you a chance to see exactly how what this process looked like for faculty. So although the creative and demand workflow is relatively straightforward, we decided that the screencasts would be helpful and also serve to emphasize how quick the process would be and allow us to highlight the important points in the process. So this is the screenshot from the first part of the tutorial. So we emphasize there, register for your ORCID step-by-step in two minutes. It will literally only take you two minutes to set up your ORCID. We really emphasize that minimal time commitment. This is the email invitation that they would see, so that would let them know to expect something that looks like this, arriving in their inbox. Note there that it says create or connect to cater for those who already have an ORCID ID. We tried to ensure the email was not too wordy. That's a personal problem I tend to write far too much, but we did want to tie the ORCID initiative to things like the university's strategic plan. So we put the button up top, but below that there's quite a lot of additional information that we were trying to communicate. This shows the permission state. So once they click on that button, click on that link, they're taken to the ORCID system and it's clear from there, from the ORCID system that it's Hong Kong Baptist University that's asking for permissions for their ORCID account. And that option there, allow this permission until I revoke it, that's that long life token that I mentioned earlier. You can't see it too well in this screenshot. You can sort of see the first box there. In fact, this form will be prefilled with all of their information because the link, remember, was customized for them. So when they click on that link, they're taken to this form, but their name and their email address will already be filled in. And literally all they have to do is choose a password and agree to the terms and conditions and click, you know, create my ORCID account. So this is the final screen here. So the creation process takes 10 or 15 seconds. They're then given a link where they can go and view their freshly minted ORCID account to check the publications that have been uploaded. It includes only the last five years, but if they want to add more, they can ask us to. So I've given you the link there if you want to view that screencast yourself. It didn't take too long to create. We did do a voiceover for it, but in total it took me about two or three hours to put that together. And we got some good feedback on it. So I mentioned there that we would add on demand, you know, publications older than five years, you know, faculty can send us their entire CVs if they want to because our guiding principle for this project was to really save faculty time. So apart from the benefits of having an ORCID account itself, we really emphasized throughout all of our publicity efforts that this project is intended to reduce the administrative burden on faculty, not to increase it. So for those faculty that wanted to include past publications, we told them, just send us your list. Our team will add it to their ORCID using the existing infrastructure that we've set up with the RVP database. So our rollout began in January of this year and by the end of April, all of our faculty had received invitations to set up their ORCID accounts. So how did we do? Pretty well, if I do say so myself. So you can see here, the faculty uptake of ORCID at HKBU, just over 70% of our faculty have connected their ORCID accounts to our system. So it's a much higher rate than any other case that I'm aware of. So I mentioned before, Hong Kong U reported 30% uptake of their ORCID initiative. I've also read that Texas A&M reported 25%. Both of those institutions used the create-on-behave approach. And I should emphasize that these institutions were sort of disadvantaged by being early adopters. At that time, the create-on-demand workflow, I think I'm saying was not supported yet. So they were kind of stuck with that approach. But we found that the create-on-demand workflow was much more successful. I should also say that this result was only arrived at after three or more email reminders to every single faculty and in many cases, follow-up telephone calls and office visits. So we really put a lot of effort into at least getting a response from our faculty. We really did a hardcore press to get them to sign up for ORCID. And we're still working on those 27% that haven't responded yet. So we're still sending out emails, still calling people, it's summertime now, so some of those faculty will not be here. But we will keep trying until we get a response. And we do really mean that. You can see there the third category, refusals, like some faculty, maybe they were so tired of being pestered by us that they just said, no, stop talking to me about this. That's not true in every case. Many of the sort of 3% refusals were retiring professors. So they sort of came and they responded to us and said, well, look, I'm retiring this year. I don't think it's really worth your time to do this for me, which we were fine with, absolutely. A couple of faculty had privacy concerns. They didn't want their list of publications to be out there in the wild, which I was a little bit confused by because one of our goals is to make their research more visible so that they get more recognition and more citations and wonderful things like that. But there were a few like that. And some were simply not receptive to our message. So maybe in those particular individual cases, we didn't do a good enough job convincing them of the benefits of ORCID and what it could do for them. But I think everyone is gonna run into it. When you have a sizable group of people of academics, you're gonna get 10 or 20 out of 400 that refuse. And of course, this is voluntary. They're not being forced to sign up for ORCID if they really don't want one. We will mark that in the database and we will stop sending messages to them. After this relatively successful launch of our initiative, our thoughts are turning more towards ensuring that this is sustained going forward. So one thing we're particularly thinking about now during the summer is how to standardize, routinize some of the workflows that we've established. For example, how do we handle faculty that retire or leave the university? Do we need to remind them to revoke the permission or our permission to edit their accounts? How do we handle new arrivals? So we're working with personnel office on how are we notified of new arrivals? And even things, how are we notified of promotions? Cause we want to update people's position titles in their ORCID accounts as well. Another important workflow is receiving new publication information. So for our initial setup, we sort of in one big batch from the graduate school, we got the past five years of publications, but we want this system to be responsive. We don't want to just be updating publication information annually or every year we want it to be in real time as far as possible. So we're working with the graduate school and with our Office of Information Technology to enhance our internal systems so that we will receive new publication information as soon as faculty submit it to the graduate school. So we will be simultaneously notified and within a few days, hopefully, we will be able to upload that to their ORCID accounts. So then, faculty can really see that there's, oh, they're doing something with all of this information that I'm giving to them. And this is important. We're gonna keep talking to our faculty about ORCID. I really feel that having an ORCID is just the first step because no matter, despite all of our efforts in communication, they might not fully be aware of what the ORCID initiative is about. They may have one, but they might not necessarily know what to do with it and they won't really benefit from ORCID if they don't start using their account. So our outreach will continue. Right now, we're identifying services that take advantage of ORCID integrations, things like impact story and kudos that have built-in ORCID integrations. So for those research services, if you have an ORCID account, you can just plug that in and it will draw all the publications in. So for something like kudos, you can quickly import your list of publications using ORCID. And we wanna promote that to faculty to show them that this ORCID account that we really pushed you to sign up for, there are really many use cases for it. And of course, the normal things like keeping them aware of when publishers adopt ORCID. So when you submit your manuscript to this publisher, you should use your ORCID so that you're properly credited and affiliated with that output. And of course, we wanna keep talking to our faculty. We wanna hear their feedback as well, what we're doing right, where we can improve and how they're using ORCID. I've spoken to a few faculty who've started using it and that's very exciting from our point of view. So thank you very much. That's pretty much all I wanted to present. Thanks for listening and thanks again to Australia National Data Service and call for the invitation.