Leveraging the Cloud for Analytics Instruction at Scale: Challenges and Opportunities
Data science and programming languages like R and Python are some of the most in demand skills in the world. Students interested in analytics and professors facilitating curriculums deserve to use industry-leading tools to acquire these skills. However, it’s challenging to enable this experience in an educational setting, especially at scale. The traditional tools to facilitate learning analytics simply aren’t great. Students and professors often spend way too much time troubleshooting systems and software, things that are a complete waste of time and detract from the learning experience. Additionally, there are seemingly endless IT hurdles and requirements. That’s part of the reason we created RStudio Cloud, a brilliantly simple but powerful solution for teaching and learning analytics, especially at scale. RStudio Cloud solves many of the technical and financial challenges associated with teaching analytics. It’s also a joy to use for professors, students, and IT administrators. In this presentation, Dr. Brian Anderson will discuss the challenges and opportunities associated with leveraging the cloud to deliver analytics instructions at the undergraduate and graduate levels at scale. Our hope is that you walk away inspired to think about ways you can leverage RStudio and the Cloud to enhance your students' experiences with learning analytics. Read more in the follow-up blog post: https://www.rstudio.com/blog/teaching-data-science-in-the-cloud/
image: thumbnail.jpg
Transcript#
This transcript was generated automatically and may contain errors.
me in the past, I have the very fun job of running growth for RStudio Cloud. In terms of housekeeping for today's event, we're going to have a guest presenter run through a presentation for the first half of today's session, and then we're going to jump into a Q&A for the second half.
Now, I have a colleague on the back end who's going to be hopefully juggling some questions as they trickle in via chat, and if you need to leave early, please don't worry, we will be recording this. And please keep in mind that the focus of these events are to address the questions most important to all of you. Now, with that being said, I'm super excited to introduce today's guest, Dr. Brian Anderson, Associate Dean and Professor at University of Missouri-Kansas City. Pete, thanks so much, appreciate it.
Thank you, everybody, and welcome. I'm so pleased that you all take the time to join us. And Pete, thank you and the team at RStudio and RStudio Cloud for giving me the opportunity to chat about a topic that I'm very passionate about, but also one that as an institution, we are very much still working through and still trying to figure out how we're going to be able to working through and still addressing, you know, our own path forward. I think it kind of reflects the very rapid pace and development, you know, in this field. And so I'm looking forward to sharing some thoughts about the challenges that we have identified as we look to move towards a kind of a cloud-centric solution for our undergraduate and graduate programs, and also talk about some of the opportunities ahead. And then, you know, certainly the chance to hear from you all, answer some questions, and have some dialogue.
About the Block School and UMKC
So in my role as Executive Associate Dean at the Block School, I oversee all of our degree programs. And so this notion of how do we leverage new and emerging technologies to teach topics, you know, certainly analytics at scale, is very much in my area of responsibility. So let me give a little bit of context about the Block School and the University of Missouri-Kansas City.
So the University of Missouri-Kansas City is part of the University of Missouri system. There are four campuses in our system. You may be familiar with our larger sister campus in Columbia, Mizzou. We serve the Kansas City area in the Kansas City region. So by Carnegie classification, we are a research-to school. So we are a research-centric school, but we also are very much a public urban research university school.
So as part of that mission, our student population very much reflects the diversity of the Kansas City region. So for example, we have a very large percentage of our students that are first generation college students. We have a large percentage of our students that come from historically underrepresented communities. Our student population reflects, you know, significant heterogeneity in socioeconomic background. And so when we think about leveraging technologies, it's important to kind of keep that student body population in mind.
Drilling down a little bit more, the Block School is what we would consider to be kind of a medium sized business school. So we have about 2000 total students. Our single largest degree program is our professional MBA, which has about 600 students in that program.
We focus on business analytics. I think that context is important here. Most of what you would think of as kind of core knowledge, core technologies would be very similar in kind of a standard computer science intro course or a kind of a basic undergraduate statistics course or a core graduate statistics course. But it's really taught from an analytics frame or excuse me, a business frame. So it's grounded in a business conversation and addressing business problems. And I think that's important to keep in mind that everything really kind of needs to tie to our core business context. I think an analogy might be say a health analytics space where the statistics, the analytics coursework behind it would reflect very common elements, but it would be taught in a health frame.
So we offer analytics specializations, what would effectively be like a major. We also have a smaller interest area and then we have a graduate certificate. And then the last contextual piece that I think is very important is the employment base within Kansas City. So most of our students come from the Kansas City region and they will work in the Kansas City region. And Kansas City is very fortunate that we have a very strong and diversified economic base. Our largest single sector is healthcare, but that's still only about a fifth of the economy here in Kansas City.
And I think that's important for us because as we think about preparing students with analytic skill sets, it's important that they are likely to serve in a variety of different verticals. It would be a disservice to our students, for example, if we focused on one particular context. And so it's important that the skill sets that we provide and the training that we provide really reflect the heterogeneity of employment that our students are likely to have here in the Kansas City region.
Problems we're trying to solve
So that's just a little bit of context, but let's jump in and talk about the problems that we're trying to solve. Because this really frames why we are looking towards the cloud for teaching analytics at scale. And I would put all of these questions with a pretty similar ranking. And so as we think about leveraging technology to teach at scale, we need to address all of these challenges.
So first and foremost, relevant and responsive coursework. R, Python, SQL, the common technologies that you find, say, in more of a software engineering or computer science type career field, we're going to see those same kind of technologies among our employers. So we need to ensure that we are providing students with coursework that aligns to employer expectations.
The second bullet point for us is very important. And this was a strategic decision that we made. And I think it's important from an academic standpoint that schools make this very purposeful choice. In our case, it was we want to train business professionals in data and analytics literacy. That is, in contrast to we're going to train data scientists in business literacy. So our content and our coursework is really meant to provide a baseline level of knowledge and understanding and exposure to analytics, skill sets and technologies, with the assumption that the vast majority of our graduates are not going to work in analytics specific fields.
In our case, it was we want to train business professionals in data and analytics literacy. That is, in contrast to we're going to train data scientists in business literacy.
So for example, you might say an accounting student, the accounting field is very much driven today with an analytics background. But first and foremost, they are accountants and they need to be CPAs. And so it's giving a CPA training in analytics to help that individual be a better CPA.
All right. For us, you know, cost of technology is always salient. And that's certainly true for the higher education industry in general. But for us, being a public urban research university, we are very cognizant of the heterogeneity in technology skill sets that our students come to us with. So we will have some students that have very strong technology skills. They've had exposure with programming in high school, for example. They have strong mathematics backgrounds. And we will have other students that did not have that opportunity at high school. And so they are coming to us with a more modest technology background.
That would also be true in our graduate programs as well. We might have individuals that are working as software developers. And so they are very comfortable with technology and learning new technology. We'll also have individuals that may be coming from a nonprofit administration background and have had very little exposure to technology. So lowering the cost and the barrier to learning new technology is a very important consideration for our context.
From a school perspective, then keeping costs reasonable for technology adoption is certainly also important. The criteria are the rapid increase in technology budgets, particularly in the last couple of years as so many universities shifted to a greater percentage of online teaching or other digitalization tools has dramatically increased the share of the technology budget from a university standpoint. And so areas in which we can keep cost growth reasonable and lowering technology complexity is an important consideration.
So the last two pieces, I think, tie very well together in that looking for solutions that address core curriculum and that core curriculum might be, say, a course that has, you know, 400 students per year taught across potentially eight sections with two or three faculty members. And so having that within course experience taught at a larger scale while also being able to provide opportunities for students that want to take deeper dives into more technically advanced concepts to have a similar consistent experience.
Current solutions
So just real quick, let me talk about the current solutions because I think this really reflects where most universities and most business schools are. First and foremost, there's a not just a requirement but an assumption for students and employers that Microsoft Excel and competency with Excel will be an expectation for a business school curriculum. And so every business student needs to have a baseline competency in Microsoft Excel.
For finance and accounting students in particular, these areas where Microsoft Excel is a heavy part of employer expectations, these students have to have, you know, a fairly sophisticated training and exposure to Microsoft Excel. And that's not going away. And I think that's important when you keep in mind there's always going to be this technology that will be a required element of the student experience.
But then outside of that, you know, SAS, R, some legacy, SPSS, a little bit of Stata. We have, and we'll talk about this a little bit more, we have started using RStudio Cloud for some of our graduate courses and we're continuing to expand those. And we've had quite a bit of success with that. Python, you know, certainly in the data science community, Python has, you know, been, you know, a very, you know, not just emerging but, you know, very prominent technology for some time. We're just now kind of seeing interest among employers to have employees having a kind of at least a little bit of exposure and understanding of Python.
And then from a data visualization standpoint, Tableau and Power BI are by far the dominant platforms that we find in industry for our employer partners. And, you know, having students having exposure to those software tools is an important consideration.
So I think this is important to kind of note that even as we look at leveraging cloud solutions to teach at scale, it's part of a suite of technology tools. It's not to the point, nor do we really imagine it becoming the one solution that allows us to accommodate all of our student needs, right? But by the same token, if we can minimize the number of software tools that students are exposed to and are expected to learn while still addressing that relevant and responsive coursework, then that's a very powerful value proposition for the school.
Where we'd like to go
And so where would we like to go? As we think about, you know, analytics instruction at scale. And so this would be at scale would mean, you know, we've got 2,000 students that are going to be taking a variety of analytics courses every year. What we'd really like to avoid is direct installations on student and school computers. And that really leads to a cloud-centric solution.
That addresses a number of student issues. One, things like tech support and just the hurdle rate, if you will, of spending class time to, you know, install software, troubleshoot software that's directly installed. It also reflects changes in, particularly among graduate students, where they are using, whether that's, you know, a laptop, whether it's a tablet or something that is being provided by their employer. And so because of that, they're allowed to use that employer-provided laptop for schoolwork, but they can't install software on it. And so this is where for a good portion of those students, you know, the ability to offer a solution in the cloud really is an important part of the value proposition.
It needs to be laptop and tablet agnostic. So it cannot be tied to a particular hardware requirement or a particular operating system requirement. So whether that's a Chromebook, whether that's, you know, Mac OSX, or it's a Windows environment, it needs to be agnostic. It needs to allow for that consistent student experience and allow for incorporation of multiple languages. And again, it's part of that workflow situation where students are still going to be expected to learn multiple technologies.
So for example, you might have a course project that is, you know, drawing from Microsoft Excel spreadsheets, where a student is going to manipulate, wrangle, and do modeling in RStudio Cloud, and then a product of that would be exported for visualizations through Microsoft or through Tableau, for example. Those kinds of projects and workflows and making it easy for faculty to develop those kinds of, you know, very immersive, very experiential projects are very important.
Challenges: financial
So this is where we'd like to go, but we're not there yet. And we're not there yet because of a variety of challenges. And so let me kind of shift a little bit and talk about some of these challenges. And I've split these off in financial, technical, and then faculty considerations.
So the first off, it is very important in this context to address the financial considerations. And here there's kind of two different approaches. One, if the school is going to be buying the technology, and that could be that the school is absorbing the cost of the technology purchases just out of an operating budget, for example, or if they're going to pass on that cost in the form of a fee to students. But ultimately, the school becomes the purchaser.
So here are some of the challenges if the school is going to be the one buying the licensing. You know, the first is flexibility in the licensing agreement. Things that lock down or lock up into exclusive relationships, those can be more of a challenge for universities to buy into. And there's a variety of reasons for that. But the ability to have some flexibility in how they structure the licensing agreement and how it might conform within the context of other university licensing and licensing rules is fairly important.
I think this is the flexibility on the part more of vendors than it is on the part of the universities, because in many cases, particularly for public universities in the United States in particular, there's a number of public regulations and state laws that governs what a university can actually do. And so the vendor might desire to, you know, have a licensing agreement that looks a particular way, but the university simply can't because they're prohibited under a state law. And so having that flexibility is going to be is pretty important.
The scalability issue, I think, is also you know, a really important question. And that also relates to that per seat licensing question. So from the standpoint of if you're going to offer, you know, say in our case, we're going to have 2000 students, well, we might have situations where one student might have multiple courses. So it's, you know, it's not just per student, but it's per student seat in a variety of courses. But we may not know what that demand actually is until we're very close to the start of the term. And so the ability to scale and adjust very quickly in response to not just enrollment, but student demand is an important consideration.
It's also important to note that most university driven licensing agreements, you know, reflect more of an enterprise type approach to software licensing. And so that would be the more you buy, the less you pay per seat. And so when the school is the purchaser, having that understanding that we're really kind of approaching this from an enterprise mindset is, you know, is an important consideration. And that can be a challenge because we're talking about potentially very large software purchases. You know, inevitably require additional time, effort, and energy to navigate through the purchasing process.
Another, you know, way to look at a financial challenge is if the student is going to be the one purchasing the software. And by the student as the purchaser, I mean, they're going to be the ones who make the buy with the vendor itself. So whether that's a credit card or electronic check or something, the university is not involved in any way in the actual transaction. And I use this term Sesame Street simple. And Sesame Street simple, it does not necessarily mean, you know, a dumbed down experience. It means an exceptionally clear and easy way for students to purchase the license that they need.
With a caveat that if you're going to do this successfully and roll this out, you are likely to have multiple semesters and multiple courses that a student might be involved with. So let me give you an example. If you have a student who in a given term is taking two different analytics courses, that student may understandably say, I just want to buy, you know, buy use for one semester and be able to use it across all of my courses, as opposed to paying for both courses to use the same software product. Now, that could be, you know, certainly addressable, but it is an important consideration. And then if the student is going to use that software across multiple semesters, then making it very easy to either renew a license or being able to buy a license for a duration of a program, for example. But regardless of what a purchasing approach is used, it has to be simple.
Challenges: technical
So those are some of the financial challenges. And then it's important to think about some technology challenges too. You know, student privacy and data network security, you know, are not only very salient, but in so many cases are really driving purchase and adoption decisions. So any kind of technology that is going to touch student data in some way, there is an expectation for a very high degree of security and data integrity.
Now, I think in a lot of ways, this actually supports cloud-based solutions that are, you know, that avoid having, you know, data and student records that may be stored on laptops that can be lost, that can be stolen. So having cloud-based solutions that provide a high degree of security and integrity, that is an important technical challenge to keep in mind.
Many universities are moving to, you know, what we would think about as a more centralized administration of technology resources, not dissimilar from what we would see, you know, in other businesses or large organizations. The net result being though is a college or school or department may not have organic resources to provide technology support, right? So from the standpoint of needing what we would think of as kind of like help desk type of support resources, many universities and many schools within universities simply may not have those additional resources.
And then this last piece is also such an important technical challenge. As schools have gone to, you know, the adoption of learning management systems, online course management systems, and leveraging extensive technology resources within what we would even consider to be a regular classroom-based course, the requirement that these technology tools really must work and play well with others is just absolutely paramount. In fact, if it doesn't, that is a substantial barrier to adoption. And so this could be, you know, things like working with Zoom or Panopto or Canvas, you know, some of these other platforms, but also the ability to pass student records, you know, student names, student email addresses, single sign-on type solutions. All of these things are additional requirements that have now become part of the student experience. And so we want to look for ways that minimize friction points for students in adopting new technologies.
Challenges: faculty considerations
All right, so those are some financial, those are some technical. And then let me talk about this not as a challenge, although for some of my fellow faculty members that may be listening, I think you can appreciate that sometimes faculty, you know, have strong opinions. You know, while I'm a dean, first and foremost, I feel that I'm a professor first. And so I can appreciate the importance of faculty considerations. And these ones that we're going to talk about, I would all put in a similar importance category, right? Any kind of technology solutions and a cloud solution, for example, must meet all of these requirements.
So the first one, and this ties with, you know, what we just talked about, the importance of integrating with an online classroom management or learning management solution, right? And that could be as simple as ensuring that links get populated into, but also that assignments, that deliverables, we'll talk about grading in just a minute, all of those things can be put in one spot and easily accessed and shared across multiple required platforms.
Pedagogical flexibility, I think, is another critical piece. And this would be, you know, for example, let's take an analytics course where you might have a faculty member who prefers live coding as, you know, a, you know, how they prefer to teach their course. And so they're going to be interacting with students. They're all going to be coding at one time, for example. That is how the instructor would like to interact with the technology and deliver the course. You might have another faculty member who prefers project-based work that, you know, classes are kind of meant for discussion, identifying some issues, and then students are going to go on their own time, work on projects, develop, you know, deliverables, and then bring them back into, you know, a course for discussion. Ultimately, the platform must be able to have sufficient pedagogical flexibility to really meet faculty where they're at.
And then today, when you think about instructional materials, it is very common for textbooks, for other kind of packaged curriculum products to have a very kind of integrated approach. So quizzes, projects, deliverables, outside readings, everything kind of comes in a package that the faculty member can certainly, you know, adopt, not adopt, you know, kind of pick and choose. But there is kind of an expectation that these things will all kind of work together. So when you think about analytics textbooks, for example, you know, many of them today are either tied to a software product or, you know, so it could be Microsoft Excel, for example, or they have examples across a wide variety of technologies. So you might see, okay, here's how we fit a regression line using Microsoft Excel. And then in the appendix, they would have something in R and then potentially, you know, Python and, you know, Stata and so forth. So having instructional materials that are kind of built for the solution is a very important consideration to encourage faculty adoption.
Modality flexibility, you know, there's certainly as campuses reopen, you know, coming out of the pandemic, we're certainly seeing more students in class, which is a wonderful thing. But I think we're also in an environment now where there is significant expectations from students that they might engage with a wide variety of classroom modalities through a degree program. So that could be online, asynchronous, where it's recorded lectures, it's, they're really not interacting in any kind of live way with an instructor, but they're, they're going to have, you know, a heavy online experience, all the way to, you know, a blended classroom where you might have students in a class and online at the same time, and you're engaging in a fully synchronous discussion and classroom activity. So any technology solution needs to be usable in a wide variety of course modalities that reflects where we are today as, as an industry.
You know, student management, and really, this is kind of as simple as who's in my class. And students are already registering. And in most cases, learning management systems get pre-populated with students, they reflect students that add, drop, you know, at various points in the semester, you know, faculty members might go and create groups, you know, in these environments. The ability to ensure that faculty do not have to necessarily duplicate processes to ensure that students are added as appropriately to a course, that is an important consideration.
As would be, for example, the ability to stack courses. So let me give you an example. As a scale question, you might have 200 students in a semester taking an introduction to business analytics course, and they are in four different sections that might be taught in a variety of different modalities. Maybe two sections are on campus, one section is online asynchronous, and another section is online synchronous, right? So we have, we have multiple modalities. We'll keep it, we'll keep it simple. There's one instructor, but we have multiple modalities. We have 200 students. And some of those students are going to add, they're going to drop, they're going to be in groups. As a faculty member, it's the same course. And so I might want to manage my RStudio Cloud course portion that is agnostic to which section the student is in. So it's kind of one group of, say, 200, but I need to be able to easily manage the lists and manage grading according to that individual section structure.
And now this would be, you know, very similar to how you might see an experience in Canvas, for example, where you're able to, to combine multiple independent sections and manage them in one, in one area. Then the grading and student feedback. When we think about the, kind of the barriers to grading courses at scale. So we've got 200, you know, for example, students are expecting, you know, a reasonable degree of feedback and interaction with an instructor. And if you have multiple deliverables, you know, so let's say you have five projects over the course of a semester, well, that's a thousand instances that must be graded and returned to students. Now, could be the case that artificial intelligence, machine learning solutions, kind of automatic graders, those are becoming, you know, a little bit more common. That could be, you know, one possible solution, but ultimately the ability to grade easily and to grade quickly and to have those grades move into course management systems, you know, that is a very important faculty consideration. So ensuring that the faculty member does not have to duplicate a lot of effort, because anytime that that's the case, faculty adoption is likely to go down substantially.
Driving change and adoption
Okay. So let's then, you know, talk about what, what we have been doing to, to drive change and drive adoption. So, and I want to, I want to be clear on a couple of things here. We, as a school, we have a vision for where we would like to go, but we are not yet at the point where we are ready to simply turn on a cloud-centric solution and make that our de facto standard. Some of that is, you know, reflects that cloud-based analytics software, including RStudio Cloud, which is a wonderful product, and we've had a lot of success with it, is itself still developing. You know, it doesn't have some of those, you know, technical capabilities that would really allow us to, to think about using that package to teach 2000 students per year, you know, with a very, you know, faculty-friendly and student-friendly way. That doesn't mean that the product doesn't continue to advance and get stronger. It absolutely is, but it's not quite ready yet, but neither is really anybody else.
So, it's, it's also kind of important to keep that in mind. The technology itself is still developing, but this first bullet point, when we think about driving change and using cloud-based solutions for, for analytics instruction, it has to be faculty-centric. Faculty have to be driving it, and there needs to be a critical mass of faculty that are saying, not just we want to use this product, but we feel that this product is better than alternatives, and it's better because we can do more in our course. We can lower student friction and increase student experience if we use this product, right?
Faculty have to be driving it, and there needs to be a critical mass of faculty that are saying, not just we want to use this product, but we feel that this product is better than alternatives, and it's better because we can do more in our course. We can lower student friction and increase student experience if we use this product, right?
And I think that, that message, that critical mass from faculty is such a critical piece to this, you know, from administration kind of pushing down to faculty, you know, you will, you will now start using this to teach your courses is not likely to be a, a productive approach. Instead, it's more of a, a productive approach. Instead, it's more of a bottom-up, you know, the necessity of a bottom-up, which might require some additional faculty education.
And, you know, if you have faculty who have, you know, been, you know, very used and very comfortable with teaching with a particular technology in a particular way, you know, they might, they might be willing to make the shift and adopt, you know, you know, in this case, a cloud-based solution, but that's going to require rethinking their course, potentially rethinking deliverables, and learning the new, you know, software product themselves. And so, the importance of having faculty education to help them see how they can use this technology to make their course better, that is a critical part.
It's also helpful, you know, for those faculty members who are, you know, say tenure-track or, you know, they have a strong research component, there's certainly value in having a connection to their research technology. And I think this is an advantage for R and RStudio because of the adoption in academia of both R and RStudio. An advantage of RStudio Cloud is the interface of the cloud product is effectively a mirror image of the interface of the desktop product, and so likely to already have some faculty with some comfort in R and RStudio, but that's certainly not universal.
If I have a faculty member, for example, in finance who is very used to Stata, and I'm trying to encourage, you know, him or her to think about RStudio Cloud as an alternative, you know, that faculty member may very reasonably say, you know, I know how to use this product. I'm very comfortable with it. I, you know, I'd rather just not teach the course that you want me to teach if I can't use the technology that I'm most comfortable with because that's my research is all tied into that existing, you know, into an existing platform. So, you know, important to kind of keep in mind that, you know, a research connection, that would be ideal, but certainly not a requirement.
It's also, you know, in any kind of organizational change initiative, you know, having senior leadership support is important, but I would add a little bit of a caveat to this. You know, certainly having, you know, a CIO or chief data officer or e-learning officer that is supportive of the technology, that is very valuable, but I would say it's more the instructional leadership, so deans, provosts, chief academic officers, you know, those who have the direct connection to faculty that can help address faculty concerns, but also provide, you know, incentives and support to encourage faculty to use this technology to improve their courses, right?
And then, you know, certainly, you know, creative financial solutions. By creative, you know, universities are generally under budget and cost pressures from a variety of different ways. That is not likely to mitigate in any meaningful way over the next decade. You know, certainly varies by, you know, geographic area and around the world, but I think it's probably safe to say that on balance, cost pressures are going to be, you know, a meaningful consideration for not just for a university that might be looking at adopting this technology, but certainly also for students with the increased cost of higher education. So, you know, always kind of keeping in mind, you know, sometimes it's a lot easier to solve the technology barriers than it is to solve the financial barriers, but solving the financial barriers, you know, is certainly a key requirement.
Q&A
So, that is it for the conversation that I had. I know there's been some chat going, but I certainly would like to, you know, extend the opportunity for any questions that anybody in the audience might have, or Pete, I see you coming back on the screen. Great to see you again. If there are some other questions or opportunities to expand. Thanks, Brian. That was super insightful. Just like all the times leading up to this that we spoke, I took a ton of notes.
So, I really appreciate that. Let's see if we have any questions coming in live. If not, then I will throw some of my own in to get the conversation started further. Let me just double check. And would certainly echo the observation in the chat. It's wonderful to see so many folks from around the world joining us. Absolutely. Love seeing that. So, I'm having a little tech issue getting the questions to show up on the screen, because apparently my questions are too long. But not to worry. I'll just read through them.
So, the first one I had was, you know, the idea that you are appointed as dean, but you're also a professor. Do you think that the fact that you do have that dean appointment kind of changes your experience adopting a solution like cloud compared to someone that might be a professor only? You know, that's a great question, Peter. And the short answer is certainly yes. You know, I think from an individual standpoint, you know, because you occupy a leadership role, the approval process, if you will, is a little bit streamlined, because you're the approver. So, from the standpoint of being able to pilot and try things out, I think in our case, we were able to do more experimentation. And that experimentation started in my own class. And using RStudio Cloud in my class to kind of teach it in a different way. And then I was able to use my own experience in working with other faculty about how, you know, successes that I had, challenges that I observed, student friction points, but then also being able to, you know, speed up approval processes, because I was then an advocate for other faculty wanting to use the product.
Now, when you mentioned the current stack that you had used, I think you used the term employer need. And that kind of helped drive which tools were going to be used. Curious kind of how you got that feedback. I don't know if it was through alumni or through various employers in industry that you might keep in contact with, but what does that process look like? No, that's another good question. I think, and I think business schools, you know, are able to have that dialogue with employers in a more organic way, for example, than other units that might exist on a university campus, because we are so connected with the business community. I mean, you know, that is our mission. And so we generally have our own career services teams. We have, you know, various advisory boards. We, you know, are interacting with alums on a very regular basis. And so there's kind of a continuous dialogue that happens as a result of simply being, you know, existing to serve the employer community.
But I think one of the, one tool that we use quite extensively is through our career services, you know, team, but also you can do this with other available sources, is job postings. And so within our region, so within our region and for positions and employers who are not necessarily hiring, you know, a substantial number of what we would think about data scientists, but do require new hires to have broad analytics experience, what are some of the technologies that they are asking for? I think a good example here is a bank, right? So certainly a bank will be hiring, most banks today will be hiring a certain number of data scientists for, you know, very heavy technical jobs, but, you know, a commercial lender, right? An investment manager, you know, these individuals are now expected to have, you know, a, at least some level of competency in a variety of technologies. And this is where, you know, our Python sequel have kind of risen to the top. And then from a data visualization standpoint, Tableau and Power BI.
Interesting. I just noticed that some folks are actually asking about your slides. I don't know if you're open to sharing that, and if so, there's a good place for people to find them? Yes, absolutely. And so perhaps, you know, Pete, what might be helpful is I will send these to you and then you can put them on the website for the event. Would that be helpful? Absolutely, happy to work offline on that. So folks are listening, but we'll definitely get Brian's slides out there. And Brian, thanks for being open to sharing those. Absolutely.
So I believe earlier you mentioned that there's this perspective of having a cohesive tool set. So when you think of professors that might have a use case that's targeted purely for undergrads and others that might be looking at a graduate education program, or maybe, you know, there's those out there that kind of have to think about both. Do you think it makes, you know, a lot of sense to ensure that what you're using in undergrad is going to be similar to what they're going to experience at the graduate level? Any thoughts on that? No, I think this touches a little bit on that student experience and consistency of student experience. You know, some technologies are better equipped than others for the different student populations.
You know, and particularly if you think about, you know, undergraduate programs that might be targeted to, you know, particular student population, and then graduate programs that, you know, are targeted to a very different, you know, student population. You know, some products, you know, some analytics products are better equipped than others that to kind of address differences in those student populations. So, you know, for example, you know, the SAS, you know, is an example of a product that, you know, very robust, very powerful analytics product. But the learning curve, if you will, is probably a little bit steeper than what you would, what would be ideally suited for an undergraduate student population that might have substantial variation in technology preparation, you know, and literacy. Whereas a graduate program could be perfectly appropriate to use in that context.
So I think as we've been thinking about it, are there some products that are flexible enough such that you can design, you know, course experiences that align with the different student populations? And so, and this is where I think we've had some success with the RStudio Cloud product, for example, where we're using it in an MBA class that is all about data wrangling, what would you consider to be a very analytics heavy course with a high degree of technology sophistication. But then we are also using it in an executive MBA course that, you know, these are individuals who absolutely are not from strong technology backgrounds, they are senior leaders, they are not interested in getting into the nuts and bolts of learning programming, but we're still able to use the technology to, you know, create an accessible, you know, student experience.
So I think that's the overarching criteria, Pete, is student experience and ensuring that the technology solution doesn't create friction in that experience. And then if you can allow for consistency across, then even better.
That makes a ton of sense. So good news, we have some questions from the audience trickling in. So thanks for bearing with me there. The first one, it looks like it says, can you speak more to teaching business students analytics? We are developing similar programs and need to differentiate from data science. You know, this is such an important criteria, and we struggled with this as well. And it gets to that notion that we talked about, what population do we want to serve? Do we want to serve data scientists and give them, you know, business skills and business toolkits? Or do we want to train business professionals with analytics skill sets? And I think that difference was how we helped communicate, not just to students, but also external stakeholders and senior university leadership, that we need technology solutions that help business professionals understand analytics. So they need to be accessible. They need to be grounded in a business context. And by business context, this is what businesses are using. You know, this is the toolkit that we are finding in business. And business students need to have experience with this toolkit.
So let me give you kind of a contrast. You might find in a statistics department that the department makes heavy use of MATLAB software. You know, for example, very robust, excellent statistics and mathematics software. You are not going to go find that in use in industry. And so I think part of it is ensuring that the technology aligns to what employers are expecting and then grounded in that business context. So to the question about, you know, teaching, you know, business students analytics, gearing it towards business problems. So, you know, we might find, you know, a textbook, you know, statistics textbook might have a toy problem about, you know, maybe it's a healthcare issue or something like that. We might change that and say, okay, we're going to do an analysis of employee compensation. Or we're going to do an analysis of customer satisfaction. And we're going to build a predictive model of customer churn. You know, the technology and the curriculum is centered around a business problem in a very applied way.
Thank you, Brian. The next question here, looks like we're able to put it on the screen. I wonder about the partnership between administration, IT and faculty. Our IT has limited ability to implement remote and cloud resources. Yeah, yeah, absolutely. I think this is where vendors can play a role. You know, the nice thing about cloud-based products is you're already not talking about adding existing, you know, hardware or software requirements to the campus IT environment. So you've already kind of removed that constraint, which is important.
I think the other piece of it is vendors providing robust technology support that is geared towards faculty and making it very simple for IT staff to deal with this new tool that faculty want to use in the classroom. So, you know, for example, you know, APIs that vendors have developed to create plug-ins, and we've talked about this in the past, Pete, to, you know, plug into Canvas, that the, you know, that plug into various ERP systems that are used for student records and grading. I would say those are things that vendors need to take the lead on providing because schools just are not going to have the resources or the desire to take on that product. If it's, well, listen, you're going to have to figure out, we'll export the student records, but you're going to have to figure it out. That's, you know, that is not going to be, you know, that's going to create additional barriers to adoption.
And so this notion of partnership, I think, is very critical. I would put faculty as need to lead, administration needs to be champions, and IT needs to not be a barrier. And I don't mean that in the context that, you know, IT teams are barriers, right? They have a tough job to do, you know, they're dealing with a lot of things, and, you know, increasingly budgets are very constrained. There are, you know, lots of competition for talent. It's making it very easy for them to support this technology. And in some cases that could be, they don't need to support it really at all.
I would put faculty as need to lead, administration needs to be champions, and IT needs to not be a barrier.
Interesting. You know what, that's a question and topic that comes up often when I'm in various conversations with institutions. So definitely resonates well with me. Whoever asked that question, thank you. So next up, thanks for the great talk. Where can one go to start experimenting with RStudio in the cloud in our own classroom? So there's a few options, probably the best ones at the top there. You can go to that link. There's a free version that's not, you know, feature gated or anything. You can also feel free to email me directly, and we can go into, you know, particular use cases. Looks like we're coming to the end of our time, but we do have...
Yeah, if I can just add on that. One of the very helpful resources that you guys have put together in the help section of RStudio Cloud are examples of using cloud in the classroom. So the ability to create assignments, for example, is, you know, that's a great resource. I'd also say that all of the cheat sheets for RStudio, you know, that are also kind of very easy. I found that students very much appreciate having access to those cheat sheets and help resources all within the cloud. So I think you'll find it, you know, for those faculty who want to just kind of experiment with it, I think you'll find it fairly accessible. Absolutely. Thanks for pointing that out. The cheat sheets are super popular.
Down to the last minute, it looks like we have one last question here. On the financial constraints of adopting solutions, is it possible to teach how to do financial analysis, for example, with R, and what about employment opportunities with this route? So, I mean, in terms of the content, you know, teaching financial analysis or ratio analysis or some of these, you know, teaching an accounting class, for example, in R and everything that goes with it, yes, you could absolutely do that. I would kind of balance that with the expectation for Excel skills, you know, just from an industry standpoint. But it is possible to do a lot of things like that in R. And I would say opportunities to find complementary uses of technologies within some of these more, you know, what it would kind of consider to be more quantitatively sophisticated courses, you know, are useful. So you might have, you know, modules that have students working in an Excel product. You know, they're doing, you know, different kind of analyses, different kind of works. But then, okay, you're going to move it into R, and you're going to do, you know, more sophisticated predictive modeling or things like that. And you're going to use these tools complementary. You know, that's a pretty powerful combination. And that's something that we're going to be doing in some of our accounting classes starting next fall.
Exciting. Well, it looks like we're coming to a close for today. So thanks again to Brian and all of you for joining. If you have the time, I'm going to throw this up on the screen right now. There is a event tomorrow, a virtual event, where our colleague Mine will actually be running through some RStudio Cloud Best Practices. And this will be during a keynote. And then in the new year, we're going to be having our next event for this live online series. So till then, happy holidays, and bye for now. Thanks, everybody. Thank you.
