Duke ITAC - February 28, 2019 Minutes
ITAC meeting Thursday, February 28, 2019
4:00 PM, Allen Building Board Room
Note: ITAC meetings are digitally recorded for meeting minute generation; audio files for each topic are posted to an authentication-protected repository available on request to any ITAC member. Presenters are welcome to request to have their audio program excluded from the repository.
All times below include presentation and discussion time.
The meeting agenda is below.
Agenda – February 28, 2019
4:00 – 4:30 – Special Guest: Provost Sally Kornbluth (30 minutes)
What it is: Each year the Provost attends an ITAC meeting to answer questions and share her perspective as it relates to technology and other topics; ITAC members are invited to pose questions of the Provost.
Why it’s relevant: As the university's chief academic officer, Sally Kornbluth has broad responsibility for leading Duke's schools and institutes, as well as admissions, financial aid, libraries, and all other facets of the university's academic life.
We are initiating the science and technology initiative. There is a search committee operating with cross-departmental membership and we are accepting faculty nominations. A primary focus is with artificial intelligence, particularly AI in health where we have a lot of strength and a great juxtaposition with the medical school. We are also thinking about handling large data sets and using them for population health services. One of the things to consider is that we see "physician burnout" due to of all the paperwork required for medical records. This also affects patient interaction because of all the data entry. Imagine if we had an "Alexa" equivalent that could listen to the appointment and enter the data. This is aspirational, but at the junction between healthcare and machine learning, there is a lot on the horizon.
The Data+ program has over 200 applicants and has a healthy representation of women in STEM. There may be some participants who find this program to not be as intimidating as taking a class where they lack experience in the topic. That said, Data+ is proven to be a gateway from summer programs into the standard curriculum. There is also the +Data Science program which has interest from the different schools including the Law and Medical Schools and which can be applied in any area that needs machine learning. The summer and winter sessions have been very popular and overenrolled. This is part of a larger discussion around making all Duke's students computationally literate. Every single student can't be in an introduction to computer science course and every student doesn't need to know how to code, but we do want students to understand the digital world around them. We need to explore how we could integrate and sustain shorter instruction within our curriculum. Some ongoing efforts include Introduction to Mobile App Development as well as portions of the Summer Doctoral Academy. There has also been web development and app development with both undergraduate and graduate interest. This will create continued demand on our infrastructure in addition to our educational needs.
There is also an increasing pressure from alumni to make the Duke education a life-long learning experience. How can we exploit the continuing education efforts both with alumni and with students who cannot come to Duke? This includes high school students who have Duke as an aspiration but may not be able to come to our expensive summer program. Part of our problem is that we need an overall effort for computational literacy and data literacy where we can aggregate activities, brand, etc. The other issue is that we do not have unlimited resources. Philanthropy pays for some of them, but we are not retiring services as we should. We would love to accept all applicants for Data+ and we have 400 students in Duke Engage. They would love to have more and there is demand but we don't have the resources to support that. What I hope will happen in the next year or so is that we will make more of these synergistic and develop a coherent program around them.
Q: Regarding unlimited resources, you have talked a lot about how these initiatives in information technology can empower our students and the community at large but this is also facilitated by the faculty. Should there be some kind of definition of the standard information technology services available for every assistant professor? Is there a way to make available certain information technology services that are limited so that you can facilitate Learning?
A: The question is always the resources. We are embarking on a serious fund-raising effort focused on science and technology because we recognize that these can be expensive. We have to be careful not to neglect the underpinning infrastructure that is necessary in a faculty-centric way. It would be ideal to have some latitude to tell donors specific needs for the success of faculty members.
In the past, new faculty members received an email address and a laptop or desktop. You had to spend money on labs or facilities. Now they need a computational storage environment and more.
We are getting foundation money for science and technology. We ask for funds for "startup" as everything the faculty members need for the lab, but we could reference technology needs into those requirements.
Some of the candidate startup aspirations are significant. Duke's traditional way of meeting those has been to give the faculty their slice to do with as they please. Instead, perhaps we could have a centralized facility that can be available for other research faculty.
We are competing with those who are offering candidates significant dollars. You might give them a little more for their core needs with a portion of the shared resources to stay competitive. Or you can address other factors like "control". And sometimes, we can't compete.
Q: I enjoyed your point about a life-long learning philosophy in engaging alumni. In the School of Nursing, the field changes quickly. Is anyone doing anything creative such as improving the students' presentation skills so they can reach out to alumni?
A: This is a great idea. We often think about student-alumni engagement in terms of alumni meeting with students (ex. from the business world). But in many areas, things have moved beyond the alumni experience when they attended Duke. The students could convey what they are learning to alumni. I like this idea.
We are somewhat internship-focused and are largely graduate programs. We don't use alumni as we could. Even in day-to-day coursework, I am seeing that students come in and have different experiences. I didn't know if there was a way to do some partnering either via personal contact or with online interaction.
Perhaps with the faculty being the linchpin for both.
There are good examples of interaction from alumni to students such as the Duke Reader Program where alumni read what students have written and help them hone their writing skills. There something interesting in flipping that dynamic.
Q: Is there a coordination among the alumni relations at all the different schools?
A: Yes. The Duke Alumni Affairs office touches all the schools but it is true that the professional schools tend to have their own robust alumni operations. These are early discussions when it comes to an "arc of learning". But there are two things concerns: one, it might lower the bar for all the schools in terms of what they are doing and two, there are resource issues. But these could be monetized. We are not going to charge our alumni for programming we offer but you could imagine a model where you consider all the schools together where you have this incredible suite of opportunities. There could be a subscription fee, the materials could be repurposed for other markets, etc. We are just starting to think creatively about this, but I think there is a huge opportunity. Other places are seizing them, and we should do it soon.
I think this is a tremendous opportunity and there is a task force focused on alumni engagement across the world. How do you engage alumni where ever they are? We have thousands of alumni in China: how are we going to reach them? I think the way to reach the millennial is through digital media. It's not going to be event and sports in the traditional sense. And then when you retire, there is learning for you. But professional learning is going to be the way that we support our alumni and as an educational institution, we have more to offer.
Q: Duke is a great place for inter-disciplinary research. Can we also do something to make courses and programs more interdisciplinary? Perhaps we could offer an introduction to computer science from a perspective of programming, bringing faculty from other units to co-teach. It is very challenging to set up those kinds of arrangements because of how the teaching load is counted.
A: The issue is figuring out a financial model for cooperative teaching and properly crediting instructors while meeting the requirements of the programs. Sometimes there is funding for interdisciplinary study but not often. We either have to go with a model where the school handles teaching and staffing differently or we will need more funding to distribute. There are instances where I would love to have Fuqua faculty teaching classes in the sciences and psychology where there is possibility for a great synergy, but funding would be an issue. This model is currently dependent on faculty volunteerism and faculty are stretched, taking on more and more teaching. We have faculty from medicine who are teaching on the undergraduate campus out of the goodness of their hearts because there is no model to compensate them. We need to think about our financial models. For example, allocations that go to the schools would need to be set aside for interdisciplinary efforts. Everyone likes these courses, but they are very hard to staff.
Q: Something we have heard from other first-year students is a lot of engineers at the undergraduate level would like to see more opportunities that traditionally has been skewed towards Trinity students.
A: Short instructional courses or "pop-ups" that don't take an entire semester are one option. The number of courses required for a credited engineering degree are so high that there is not a desire to keep other students out. Some schools have more flexibility in the curriculum, but we aren't sure how the engineering faculty would view this kind of approach. There are also courses that are of general interest versus those that meet requirements. Perhaps there could be common classes that are shared or other avenues such as online learning. We've started to move the equivalents for some of the premed requirements to Beaufort so students can go to the marine lab and satisfy requirements. We might be able to satisfy some of the engineering requirements in other locales to give students some flexibility outside of their narrow curricular requirements.
One of the other projects coming out of the Academic Affairs Committee is that a first-year senator in Pratt is in the process of creating a house course for Trinity students based off of the first-year design course to cover some of the basic technical skills.
4:30 – 4:35 – Announcements (5 minutes)
The Nicholas School representative was recognized.
The July 12th minutes were approved.
4:35 – 5:00 – Learning Innovation: What We’re Learning in 2019, Matthew Rascoff (15 minutes presentation, 10 minutes discussion)
What it is: Duke Learning Innovation aims to help Duke students learn more, and to enable more people to learn from Duke. Matthew, associate vice provost for digital education and innovation, will share topics and trends that are top of mind in 2019 and how they are changing the conversation about teaching and learning at Duke.
Why it’s relevant: Technology continues to play a starring role in educational innovation, both inside the classroom and beyond. This presentation will show where teaching and learning are headed at Duke, as well as outline what we are doing to stay at the leading edge.
Learning Innovation, created after the Center for Instructional Technology and the Online Duke departments were combined, has existed for about a year and a half. The mandate of Learning Innovation was a dual mission of enhancing learning for our own community and sharing what we learned with the wider world.
We were pleased to hear the provost mention the "arc of learning" that allows for in-time learning opportunities that are targeted to educational needs throughout your life. The cycle starts before you come to Duke through programs such as TIP and continues to the job market where your major may not be translating into employment. Our graduates are facing transformations in the labor market, especially related to skills and we need to equip them to be flexible and resilient learners who are able to respond to those challenges. The idea of enrollment and admission continuing to graduation becomes somewhat less important and the boundaries are a little bit blurred as we treat you as a lifelong learner. This means your needs and wants change, but you don't stop learning.
The Learning Innovation team offers faculty fellowship programs in the spring to assist in redesigning courses around how people learn. We have a course design institute for faculty who are building new courses that connects them with a community of others doing the same. This community supports them in incorporating practices that improve learning including instructional design and faculty development teams. We also have a special Duke Kunshan fellowship starting its second year. We inaugurated this for the first class of faculty at Duke Kunshan which was among the only institutions in the world providing faculty training through a high-quality professional development year, run as both an online and residency program. Over time, we plan to hand this over to the new center for teaching and learning at Duke Kunshan. The leadership is staffed by a team member at Duke Kunshan with a plan to transition this leadership.
The Learning Research & Development team was a direct result of the merger of the CIT and Online Duke departments. We want to advance the science of learning and at a research university, we have this unique opportunity to partner with faculty to study how people learn across disciplines. By partnering with faculty, we can help them build research profiles for teaching and learning education and help them publish. This partnership in research is essential to advancing that science.
Our arc of learning starts with "Blue Bridge”, our vision of learning about Duke for incoming students. They can learn what Duke could mean for them, especially those high-achieving low-income students who might not otherwise apply. The other components of the arc of learning are Digital Learning at Duke, Online Professional programs, and Forever Duke online (alumni learning). Blue Bridge focuses on the high school student. Digital Learning is about our on-campus offerings for the Duke community (students, staff, and faculty). Online professional programs are primarily oriented toward those with a master’s degree. And Forever Duke online means everyone in our learning community can stay in our community.
We offered the first "Level Up", a short mini-course created by faculty as a prerequisite for a residential course. An example was a Level Up on US civics and US politics introduced by Pres. Price and assigned as the Level Up for the US politics course required for all Sanford School of Public Policy graduates. The faculty found a lot of diversity in the preparation levels. Some were international students unfamiliar with the three branches of the US government while others had literally co-authored articles in the New York Times. The challenge was to design a course that engaged both. To address this, the faculty created a Level Up which also became an introduction to faculty and their areas of expertise. We measured the outcomes and were able to shift the curve of performance and the student’s perception of understanding as well as the faculty perception of their understanding.
Coursera for Duke is our platform for projects like the US civics and US politics Level Up course. We build high-quality content for Coursera but with Coursera for Duke, we bring this back to our community. For example, the computer science major is not the best option for making our students computer literate. Through Coursera for Duke, we may be able to offer technical content that addresses this need and allows a student to continue a liberal arts education. Students will have marketable skills for the first job and be successful in the long term. Coursera for Duke Alumni is the service we do for our alumni community and was launched about a month ago. Coursera for Duke has been up for a year and we passed 5000 in enrollment. The total estimated hours spent learning is almost 28,000 hours, twice the number for lynda.com. The ratings on Coursera for Duke are 4.9 out of five stars with a net promoter score of 62 which is excellent (net promoter score is a satisfaction measure that ranges between -100 and 100). Most of the participants have heard about Coursera for Duke through word of mouth.
This opens a window into the opportunity of stackable credentials using our open courses. This allows a learner to sample before signing up and would potentially allow a program to see who the candidates are based on their actual course performance. This idea was innovated by Mastertrack, a feature of Coursera. This is a way of unbundling from traditional paywall programs allowing greater transparency and exposure. This is a very exciting development and was a response to the decline of "Moocs" (Massive Open Online Courses). The model might be a non-credit course leading into a credit course or a boot camp to a certificate program or a degree. The cost of the acquisition of an online learner is in the thousands of dollars but you can reduce that cost through acquisition and pass savings to your students in terms of reduced tuition.
The alumni learning space provides connections into the broader professional market with more targeted learning experiences, both on and off Coursera with a particular focus on learning needs for alumni. This goes beyond our existing catalog with custom learning that is relevant and necessary. We have a course on applying entrepreneurial practices to your life. We had over 200 alumni participants with a mix of synchronous and asynchronous engagement including a Linkedin group.
Another targeted learning program is focused on physician burnout and physician wellness. 54% of doctors from a Mayo Clinic survey reported burn out but there has been a limited response in training and support. As an alumni organization, we have this unique relationship with our alumni. This program is online and if it is successful, there could be versions of this in other professions. The idea here is to focus on the "pivots" in life. We can help normalize the things that are challenging such as returning to the workforce, burnout, and career choices. We can design targeted learning strategies around these. These are more personal and group oriented then what is on Coursera and allow us to increase engagement. We can reach alumni in sensitive situations where this content can provide better assistance than local resources.
The Bass Digital Education Fellowship program will offer graduate students a year of funding in partnership with faculty members, training them in fundamentals of instructional design so they can become professional learning experience designers. This is mutually beneficial as the graduate student can include this on their resume and the faculty members have the support of someone who knows their field. We received 30 applications for the 7 open spots. We are currently running the course on digital design pedagogy and in the fall will begin the apprenticeship and assign these graduate students to projects.
The strategy for our learning research and development is to accelerate the process of informed consent and make it easier for faculty to partner with us to do learning science research specifically where an IRB approval may be required. We want faculty across the university to be able to participate in learning science research in order to get a discipline driven transformation. Peer instruction has taken off in physics because that community felt current methods were not working and students were not getting the promised outcome. We want to do this across the university. The "WALTer" or "we are learning, too" project will build a digital workflow to enable the acceleration of the IRB process for faculty who want to do a particular kind of preapproved learning research. We have accelerated this process from taking months to completing in days. This is only needed if you wish to publish. We have now published this procedure and "open sourced" it for other institutions to adopt. We believe we are increasing the standards of privacy for double-blind studies where we hold the hash. The instructor never knows who is in the experimental and control groups. The rate at which students are willing to participate increases because of the privacy standards.
"Kits" is the result of a deep partnership between Learning Innovation and OIT. There are thousands of educational technology companies which can be overwhelming. This could be managed by constraining choice. The Learning Management System (Sakai) is too constrained but we don't want 6000 applications in our ecosystem. Kits is the balance that we want to strike between modern tools and the best educational technology, where we have control and are not duplicating efforts.
We are trying to publish and share what we are doing with our own community and with the wider world (we recently had a piece published on academic innovation in a historical perspective). All of our projects are on our website and we update our projects each month and compare this against our annual goals. We have a newsletter as well. Sign up at at http://li.duke.edu.
Q: Regarding the level-up course at Sanford: what are the faculty incentives to do this? Do you rely on volunteers?
A: Sanford school has a complex system of faculty incentives. Everything you do counts towards some portion of the course buyout. The standalone Coursera course has been assigned and required for all students of the Duke course. There appears to be steady engagement online. The organizers of this level-up have a vision and we are working with community colleges and offering this to partner institutions.
Q: For Coursera for Duke, these are Duke courses that were created by Duke? And these are non-credit?
A: Correct. These are all taught by Duke faculty and there is no credit, but Duke users can take the courses for free.
Q: Are there any courses that dominate?
A: This is a moving target. The Stats course showed a spike because one professor required his students to take a class before they could volunteer in the lab. There are strong semester-driven cycles which suggests that faculty are assigning courses. Students are building complementary skill sets. This is also a low risk way of sampling some technical content.
Q: How do you incentivize content creation for faculty?
A: The same way we incentive content creation in the Coursera world. Step one is recognize the content we already have. Step two is create targeted learning experiences more like level-ups that address a specific learning gap in the student or alumni community. This may require a separate financial model aimed at the professional market t. This is unproven so far, but it is a reasonable hypothesis.
Q: Say you have identified a need and want a department to develop content: is there a way where extra compensation is involved, and the department will see this as the fulfillment of regular teaching obligations?
A: These obligations vary by department. Our goal is to partner with department leaders and do less targeting of individual faculty members and do more with chairs and alumni leadership. However, there are different incentive programs for the schools. This means we can't design a system that works across the university, but compensation has to be considered.
Q: For the alumni content, there is no cost to the alumni? Do they have to provide alumni credentials in order to access the content?
A: Correct. Content in the current program approach is free and is accessed using OneLink single sign-on.
Q: When grad students apply to the digital education fellowship, there is not a targeted faculty member assigned at that time?
A: This is correct. We want this to be mutual. We don't have them apply with someone mind although some may. We are going to design residency-matching so there is a good balance between a faculty member’s needs and a student’s wants. We don't want people working with their advisers. We want it to be interdisciplinary and a way to learn about the resources across the university.
Q: How do you find the faculty?
A: The faculty come from our funnel with projects in mind. We have around 200 projects in our funnel and estimate 1% follow through. The missing piece is the person or resource. This is also a valuable professional experience for the students that will make them more marketable.
5:00 – 5:20 – SAP Migration, Todd Orr (15 minutes presentation, 5 minutes discussion)
What it is: SAP is business application software that runs many of Duke's core administrative processes. Todd will explain the planned upgrade to SAP S/4HANA and walk us through the timeline of the update and its effects.
Why it’s relevant: SAP, the technology behind Duke@Work, has wide-ranging reach across the Duke community. Changes to the system will influence most of the university’s and health system’s core administrative processes, including accounting, accounts payable, accounts receivable, human resources, inventory management, payroll, post-award research administration, procurement, and travel reimbursement.
SAP is the financial and payroll system for both Duke University and Duke Health. We process all financial records, pay all 40,000 employees, and buy all goods and services and are involved in sponsored research administration and fund-raising support. SAP has been "live" at Duke for 20 years. Buy@Duke, iForms, and Duke@work are associated with SAP. We have included "bolt on" applications to meet Duke’s needs. SAP is run on-premises through administrative systems management by OIT.
When SAP was first introduced at Duke, it was known as "R/4". The "S/4" release is SAP's attempt to modernize and keep up with cloud vendors. S/4 uses an in-memory database called Hana (Duke has already switched) and presents a modern user experience including mobile support, support for on-premises or cloud, and optimized applications for the Hana database. In certain cases, SAP has said future components are not going to be available on-premises. For example, the HR payroll future path is in the cloud. This is also true of the travel and expense features using "Concur". As we migrate, this means a mixed hybrid environment. S/4 was introduced four years ago with a deadline for ending support of the application which we run at Duke on December 31, 2025 and we have developed a plan to address this. As for the current product, if there are bugs or if compliance is required, SAP addresses it but there is no investment in the current platform.
There are improvements in the travel and expense area with Concur. There is an opportunity to revisit implementation decisions which were made years ago using a new group of accountants which consider the way we do business today. Most of our current SAP infrastructure is on-premises. When this is completed, we expect to run S/4 finance, logistics, and organizational management on-premises. SuccessFactors for HR payroll and Concur will be cloud-based services. The fund-raising components are to be determined.
Planning started in the fall of 2017 and the project was approved in December. There was some consideration of enabling everything at once but because of the number of users, we decided a phased approach was better. Phase 1 focuses on the release of Concur for travel and expenses and phase 2 addresses on-premises finance and S/4 procurement running in compatibility mode which should be transparent to users. During the last phase, we are going to address HR payroll in the cloud with SuccessFactors as we have a mandate to leave payroll untouched until it must be addressed. Our timeline takes us through calendar year 2023 but we hope to complete the project sooner as our timeline is conservative.
Almost 8000 users will be impacted through the user experience. We will move away from pushing data into a data warehouse and doing reporting there. There will be a change in the reporting approach that is more real-time with embedded analytics. The travel and expense model will be new and give us an opportunity to minimize the amount of receipts that travelers have to manage. Travel services companies (hotels, airlines, etc) will be able to add receipts directly into Concur that will show up in an expense report automatically. There are financial management improvements in the HR space particularly around talent management. The user experience will also be changed to a tile-based web approach similar to some of the modules in the DukeMobile app where we have introduced a few of these features in selected areas.
Q: For the impacted users, are you defining these as administrative or back office users?
A: Yes. Anybody who has a security role beyond employee self-service. These include HR payroll, procurement, or multiple roles.
Q: Do the 8000 impacted users know that this is coming?
A: We have begun the communication process. We have talked about change management and made sure we do not overwhelm users.
This is one of the reasons why we are staying with SAP rather than switching to another product. Universities that implement entirely new administrative systems are spending hundreds of millions of dollars. That is the cost to do new implementations and that high cost is probably not even for a full suite.
Q: Regarding Concur, the corporate world has been using this more as a travel front-end instead of going to a travel agency. Is this planned for Duke?
A: There will be an option where you can use a travel agency, or you can use Concur to book tickets and make reservation.
Arranging travel takes a lot of time for our faculty and support staff so this is a large value-add feature.
5:20 – 5:30 – DukeMobile Update, Hugh Thomas (5 minutes presentation, 5 minutes discussion)
What it is: The DukeMobile application, available for Apple and Android, puts important information about Duke at your fingertips. Hugh will discuss the latest updates and integrations, as well as what they mean for users.
Why it’s relevant: Mobile access to important information about and services of Duke was identified as a high priority in the recent IT strategic plan; the continuous evolution of Duke Mobile advances a part of this agenda.
The "First Year Checklist" is a new feature for the DukeMobile app and requires the student to login. In order to support this, we implemented a sophisticated back-end OAuth technology. The user proves who they are the first time they login. This information is preserved, and the user shouldn't have to provide Shibboleth credentials again. They will access the checklist and the updates will come through. The menu is now customizable where a user can grab an icon, hold it, and move it further up the list.
The OAuth architecture is truly enterprise-class software and few universities in the world are using this level of sophistication for mobile apps. The mobile app talks to an authorization server on the back-end which talks to our OAuth system which connects to shibboleth. Once the user has logged in, a persistent token is preserved in the mobile app. Kong is what we are calling an API manager gateway that monitors all requests to our server. Kong is handling rate limiting and introspection from every request. If somebody leaves, no more requests or protected information will be passed back.
The First Year Checklist only applies to first year or incoming students. This is a move toward personalization and the new class is serving as testers. Those accessing this feature who are not in the group will see a message that this app is not available to them.
Other updates include Places information such as Duke Dining which will be visible to everyone, with other items such as conference rooms possibly coming in the future. Lactation rooms are already available in the next build. In the future, the DukeMobile app will display items relevant to the role of the user (student, staff, or faculty). For example, staff may see forms they need to complete.
Q: After the first Shibboleth authentication, the token stays in residence allowing the user access until they go away?
A: OAuth does have a timeout or refresh cycle that lasts for weeks. However, if they leave the community, the next time OAuth makes a request, protected information won't be passed back.
Q: With the future automatic login, is this only for first-year students or how does this work?
A: Every time the app opens up it checks the credentials of the person to see if they match the group.
Q: Will this persistent login in the future be available?
A: As we come out with new applications or if you would like greater privileges in DukeMobile, we will be providing a "login" button which will grant access. This is currently not exposed because there aren't very many applications, but we will over time.
Feedback: We sent out a standardized survey to the student body and heard from 250 to 300 respondents. One of the questions was "How can the DukeMobile app be improved?" The most popular items of feedback were having to login in each time for protected content and that people don't know about the DukeMobile app. This update addresses one of these issues.