Presentation Descriptions

Some content from the Higher Education Data Warehousing Forum is restricted to members only.

Become a Member

or existing members please login

This is an alphabetical listing of the member presentations due to be given at the 2019 HEDW Conference in Ann Arbor, Michigan. The formal Conference Agenda, listing specific times and locations for each presentation, will be available as the conference draws nearer.

Member Presentation Descriptions

“Super User” Communities: An Extension of the IR Office
Shelby Albers, University of Kentucky
Over the past 5 years, the University of Kentucky has transformed its data environment to support broader user access and interaction with data. From these efforts emerged a community of “Super Users” that serve as data liaisons within their College/business unit, generating their own self-service reports for staff and administrators and providing necessary figures for in-house decision making. This presentation will discuss how this “Super User” community came to be and how the Institutional Research and Advanced Analytics team continuously engages with these users to drive innovation in data modeling and reporting.

28 Months Later – A Data Modeling Journey Through Knowns and Unknowns
Matthew Fields, University of North Texas
In January 2017, the University of North Texas launched a new enterprise data warehouse to expand the university’s analytical capability, and to answer long-standing questions unaddressed by previous solutions. To correct existing deficiencies, new development is focusing on previously untouched source tables. On this journey, the data modeling team is finding new paths of understanding on source system metadata, business rules, and reporting practices that enlightens knowledge of targeted subject areas and brings opportunities of reflection on the decision-making process initiated at the beginning of the project. This presentation focuses on outcomes, decisions made, and approaches to future developmental journeys.

A Student Data Framework
Mick Haney, Nevada State College
This presentation will define a framework for integrating student data from admissions to graduation and everything in between.  Once the framework levels are defined, we will technically discuss how to populate the framework. We will then technically discuss how the framework drives the population and enhancement of an integrated student database.  The presentation will close with a technical discussion (and SQL Samples) on how the framework design is utilized to support retention and graduation rate analysis for any identifiable COHORT.  This presentation will be of interest to Data Architects, Persons interested in creating a Data Driven Environment, and database developers.  The framework is currently supporting 5 institutions in Nevada, and is also being successfully utilized at California State University Channel Islands. Your IR office would love to have this capability.

Advance data-informed decision making on U-M Campus
Mandie Chapman & Julie Martinez, University of Michigan
Many units on our campus do not appear to have consistent, reliable solutions for data-driven decision making. At worst, this results in ad-hoc solutions that require significant overhead and lack key business data. At best, it results in units approaching the U-M BI team for assistance to derive a solution. As an integral player in the access to valuable business data and a vital resource for business intelligence expertise, our BI Team is the ideal group to act as a solution provider for consulting, design & development and delivery of BI solutions that support accurate and reliable data-driven decision making for campus.

Adventures in Data Modeling Part Deux – Retention Model Expanded
Nadya Balabanova, Loyola Marymount University
This presentation will build upon my 2017 presentation at the University of Arizona. I will review the retention model I presented then, and I will explain how we’ve expanded the model since then. I will also share additional lessons learned and best practices I’ve come to rely on when it comes to modeling.

Architecting (Logically) for Student Success
Jennifer Wilken & Donna Burbank, Arizona State University
This presentation follows Arizona State University’s engagement to create a student success logical architecture: a holistic view of the student journey and an actionable roadmap for data-centric architecture as a backbone to continually improving student outcomes. Presentation will include salient highlights from the resulting student-centric logical data model that defines the entities involved in the student journey and their relationship to each other. We will discuss the ways having such a model might influence decision makers, technologists and analysts; the hurdles, victories and surprises found along the way; and actively solicit ideas from peer attendees for continuous improvement.

Automating Your Data Warehouse Lifecycle Using Azure DevOps
Pete Benbow, Davidson College
Deploying changes to your data warehouse can be an elaborate, hands-on affair. Everything from work item tracking to code review, merges, builds, deployments, and documentation can take days to orchestrate across a plethora of systems. In this session, we’ll show you how Davidson College uses Azure DevOps as its one-stop shop for all these functions, creating a streamlined and automated lifecycle for our Enterprise Data Warehouse that saves valuable hours of productivity and allows us to iterate very rapidly across the database, ETL, and semantic layers.

Avoid Surprises on Census Day—Track Retention Daily!
Sarah Yoshikawa & Nadya Balabanova, Loyola Marymount University
Retaining students to graduation is universally a focus of, and often a struggle faced by colleges and universities. In the 2016-7 academic year, Loyola Marymount University (LMU) experienced a dip in the official first-year fall freshmen retention rate. In response, the Office of Institutional Research and Decision Support (IRDS) began tracking retention on a daily basis throughout the registration period. A collaborative effort between the Institutional Research (IR) and Decision Support (DS) teams, IR conceptualized the report, and DS operationalized it. The report—available to administrators as well as deans and associate deans—provides users with the retention rate and information on the students that have not yet registered.

Big Ten Data Governance’s Journey – A Panel Discussion on Successful Collaboration
Laura Gast, The Ohio State University
Many institutions are realizing the importance of Data Governance (DG) and have started building their own DG programs to improve their data. On their DG journey, each organization ventures into different areas, encounters challenges and tries to overcome them to make progress. Learning from similar organizations helps accelerate the progress of individual programs. With the help of BTAA, Big Ten universities decided to do just that! Forming a Big Ten DG group to share their data governance experiences has proven to be of tremendous help over the past year.

Bridging the Gap: Connecting the Path Between Data Warehouse and Analytics
Dan Reichl, Yurong Hu & Jennifer Watson, University of Washington
Human Resources Information Services (HRIS) at the University of Michigan faced a challenge as it started adopting Tableau as its primary analytics platform over the last several years. What is an efficient and maintainable way to bring data into Tableau from the existing data warehouse, so the group could spend more time analyzing data and less time bringing data into Tableau? This presentation will focus on how HRIS went about tackling this challenge and explain the benefits/issues that were encountered along the way.

But They Told Us that would Never Happen: Automated Data Quality Strategies and Tips
John Mobley, University of Michigan
The University of Washington’s Data Integration team developed a set of tools and techniques for dealing with low quality (bad) data. This session will cover: how the UW differentiates Data Quality (DQ) and Quality Assurance (QA) issues, how the UW manages low quality sources, how the UW deals with Workday data when users create silent errors in Workday, how we engaged our campus and source data partners to improve the quality of data, and where we are going next. This presentation will include mistakes, opportunities, and eureka moments we learned along the way.

Choosing the “Right Fit” Business Intelligence
Brenda Ulin & Dawn Moore, University of Iowa
In Business Intelligence shape and size matter and determining the “Right Fit” BI is critical. With objectives varying from strategic insight, operational efficiency and data exploration, fit can be difficult to gauge.  We developed a framework by coupling BI categorization with development strategies and technical architecture. Through requirements gathering we determine the best solution categorization: dashboard, report, infographic, exploratory tool, or self-service with direct access to data mart or tabular model.  Wrapping BI solutions in a service layer creates a seamless user experience.  We will share our framework, technical landscape and BI solution demonstrations, leading to “Right Fit” BI solutions.

Collabarating with Academic Library to Integrate Their Resources with Boston College’s Enterprise Data Warehouse
Ravindra R Harve & Lance Tucker, Boston College
BC – ITS has collaborated with Academic Library to help them develop a roadmap to integrate Library data with BC’s enterprise data warehouse. As part of the presentation, we will discuss BC’s – EDW team’s journey to understand the data and how they overcame some of the challenges thrown at them to get this data into the staging area and how they plan to develop the Dimensional Model for the reporting and analytics

Communicating with Data at the University of Exeter
Dan Isaac, University of Exeter, UK
In order to communicate the story of our data we need to understand the principles of effective data visualisation. Taking examples from the business intelligence processes and products developed at the University of Exeter, this presentation will discuss some of the core principles of data visualisation and the theory that underpins them. Our overall aim is to challenge and inspire you to communicate with data more effectively at your institution.

Creating a Successful Data Governance System and Data Dictionary in the Cloud
Yuko Mulugetta, Mary Jo Watts & Rob Snyder, Ithaca College
This interactive session presents how Ithaca College in NY took lessons from an unsuccessful top-down initiative, created a new approach, and empowered a diverse group of data professionals to design and implement an innovative data governance process. The Ithaca team also presents the current state of the new two-tiered governance system and discusses the recent endeavor to create their own data dictionary with cloud tools in the Azure. The audience will be encouraged to share their own experiences and participate in brainstorming factors that may significantly influence an institution’s ability to build a successful, inclusive data governance system and tool.

Cybersecurity in Higher Education Campus Environments
Irving Bruckstein, LCCC
Data is among our most valuable assets. Higher education campus environments pose a unique Cybersecurity landscape. Having relatively open and ubiquitously accessible systems and networks that facilitate the free exchange of information poses unique challenges and can be inviting to cybercriminals. This talk will outline the typical components employed on higher education campuses for Cybersecurity, the threat landscape observed within campus environments, tools and methods for attaining the optimal blend of security for technology environments and ethical aspects of cybersecurity.

Data Governance: The Journey through the Wild West, Tropical Jungle and Every Other Terrain
Data governance (DG) has a pivotal role to play in higher education’s analytical journey. Organizational dynamics, strategic focus, and technology contribute to the success and inertia of DG initiatives. This presentation covers five elemental steps to guide your DG journey in a pragmatic manner. You will see how top tier universities have used our framework to jump start their DG programs and draw insights for your data governance initiative.

Data Integration in the Cloud Era: Case Studies, Tips, and Tricks on How to Succeed
Performance Architects
We have all experienced a major shift to cloud solutions and to hybrid on-premises/cloud environments. As a result, we’re struggling to integrate data and metadata between these solutions. This presentation presents an array of options and best practices to manage data integration and data movement in this “brave new world.” Learn from real-life case study examples, as well as tips and tricks for what to avoid in order to be successful.

Data Modeling for Beginners
Rochelle Smits-Seemann & Abby Kaplan, Salt Lake Community College
Have you ever built a warehouse data mart to incorporate a new line of business (or department), only do discover that what you built did not match the business process and needed to be significantly modified? Data modeling is a framework for understanding and communicating data definitions, relationships, and process rules. When done properly modeling ensures understanding between the development team and the business team regarding business processes. Besides clearer communication, data modeling is also used to plan and document the physical structure of the data. This presentation will overview the different types of data modeling used in data warehouse development. Attendees will learn the importance of data modeling in the data warehouse development lifecycle, how to develop business, logical, and physical data models, and how SLCC has employed data modeling in their data warehouse development.

Data Quality Strategies – Purdue University
Sarah Bauer, Purdue University – West Lafayette
Panelists from the offices of the chief data officer and the comptroller, as well as an academic college, will share their strategies regarding improving data quality. From the early days of DSS (decision support) to the creation of the BICC (BI Competency Center) and a focused data governance position, data quality has been an overarching goal at Purdue. Perspectives from previous experiences in both the private sector and higher education, along with recent initiatives: the implementation of Data Cookbook, an official seal on key reports, and the completion of a major BPR project, highlight Purdue’s focus on quality data.

Data Visualization: How to Use It For Good or Evil
Matthew Pickus, University of Michigan
Now that we have the tools to provide data visualizations, how best do we do so? Bad data visualization techniques can lead your viewers astray, causing management to make poor decisions, or users to simply not understand the takeaways. Worse, it is possible to use visualizations to purposely mislead. Can you tell when you are being misled simply by looking at the charts? This presentation will guide you through the basics of good visualization design practices, and warn you of many of the pitfalls.

Data-Driven Decision-Making: Empowering Business Through Data Visualization and Self-Service
Andrew Maliszewski & Manish Devjani, New York University
This presentation focuses on how NYU IT is enabling the University community’s reporting and analytics offerings by leveraging a self-service approach in order for our business to understand, connect to and visualize their data to make better business decisions with speed, accuracy and confidence.

DevOps Approaches to Data Warehousing and Business Intelligence
Max Michel, UC Berkeley
Come to this session to learn how DevOps tools and culture can facilitate cross-team collaboration, support agile software development initiatives, improve the quality of releases, and increase customer satisfaction. Max Michel will present the DevOps approach and share his experiences leading a DevOps team for UC Berkeley’s Data Warehouse group. At the end of this session, participants will be able to: – Explain the goals and benefits of DevOps. – Identify the challenges their own organizations might face in moving towards a DevOps approach. – Understand the tactics, options and tools needed to move to a DevOps environment.

Dimensional Modeling Design Patterns – A Review
Neil Belcher, Cornell University
This technical presentation will review the characteristics of a well-designed dimensional model at the table and column level. I use the term “Design Patterns” to represent easily recognized conventions that might vary slightly from one institution to another. By internalizing good “Design Patterns”, a data modeler can superimpose those patterns on top of a new design early in the process and guide it to completion without having to rework the design to meet local conventions. We will walk through various design patterns like table naming, column naming, column hierarchies and support columns for both dimensions and facts.

Don’t Re-Invent the Wheel – Simplify Tableau Sources
Jeff Wixon & Sally Luoma, University of Michigan
Tired and confused joining enterprise data for your Tableau Viz? Our new solution uses an existing application that is a trusted source for enterprise data. Data can be refreshed using the Tableau scheduler. The new tool is available through our API Directory for secure access to this service.

Driving Meaningful Change Using a Comprehensive Data Platform
ASR Analytics
County College of Morris (CCM) worked with ASR Analytics to develop a comprehensive longitudinal data platform. This platform will allow for the implementation of institutional-level metrics to support strategic decisions as well as operational reporting. Decision makers will now have data available to make better decisions, and support strategic plan analysis. This will free up staff to perform high-value analysis and exploration using the same platform, which will lead to insightful discoveries and meaningful change initiatives.

Join us to learn how this two-pronged data approach makes these discoveries and efficiencies possible, and the techniques used to develop the platform and address the institution’s unique data structures.

Early Adventures of a Data Warehousing Team in the Land of Workday
Rodney Pacis, Georgia Institute of Technology
Georgia Tech’s Enterprise Data Management group (EDM) was a year and a half in developing a new Enterprise Data Warehouse (EDW) when Workday was introduced as the new source for Financial data, moving away from the relational database world. Workday proved to be a disruptor to a very traditionally architected and developed data warehouse. In this presentation, we wanted to share how we pulled data down from Workday into our EDW developed in Oracle, using IBM DataStage, as well as our early data modeling process, the challenges we have faced and a look into the future.

Enhancing Data Transformation with Analytic Functions
Jonathan Havey, University at Buffalo
Analytic Functions are a powerful tool for transforming data. They have three major advantages over other methods. First, they are typically much faster than self-joins if you need to display detailed data together with aggregate data on the same row—for example a student’s current credits attempted with their cumulative or rolling credits attempted. Second, the use of custom partitions allows the inclusion of multiple levels of detail on the same row. A student’s credits can be displayed beside others in their major or cohort. Third, the windowing functionality allows very precise control over the rows used in an aggregation.

Expanding SIS Data with Engagement and Learning Management
Robert Snyder & Joseph Betram, Ithaca College
Ever struggle with where and how to start on new data integration projects, including blending your new data with existing data marts? Challenged with data linkages between student engagement, LMS and SIS data sets? Join us as we discuss best practices on breaking data integration projects into manageable sizes and using the art of the possible to expand existing student data with engagement and LMS data. We will walk you through lessons learned and provide key tips on how to engage with your data in a meaningful way.

For a Giant Leap Forward, First Take Two Steps Back
Ryan Schlagheck, Yale University
In the 90s,Yale built a monolithic data warehouse to address operational reporting needs.This warehouse was deprecated during our adoption of Workday.Since then,University leaders’ needs have evolved such that an integrated repository responsive to strategic decision-making is essential.However, rather than re-instantiate a monolithic data warehouse, Yale are building a platform and growing integrated reporting and analytics “organically”, through specific use cases which answer questions important for leadership and ONLY where the use of integrated data adds value.The platform interweaves sustainable data management, data quality and data governance practices and cloud technologies to deliver an innovative and scalable environment for the future.

Formalizing Data Governance in One Semester: A Nimble Approach at Bowdoin College
Peter Wiley, Bowdoin College
Instituting data governance doesn’t have to be an onerous effort. With an efficient task force that met for just four months, Bowdoin College created and adopted a set of policies and practices that formalized a data governance program. A practical approach will be shared, including concrete ways to organize productive committee meetings, accelerate the work of drafting policy, and keep the entire committee engaged. To ensure a successful start, Bowdoin began foundational work months earlier to build the College’s appetite and capacity for data governance. Expect to come away with actionable resources for starting data governance, or inspiring new approaches for an existing program.


Gaining a Holistic View of Data for Analytics Insights
The challenges faced by centrally provided IT services (aging ERP systems, rapidly changing business needs, cloud systems / data fragmentation) are often replicated at a smaller scale within schools that house their own IT departments. One of the top business schools in the country, the University of Michigan’s Ross School of Business, has had its own ecosystem of applications and local data for twenty five years, and is working to better integrate with campus-wide systems and to enable its internal customers and administration to strategically stay ahead of the curve. Federated data management is key to success across the U.

Join this session to hear Brian Greminger, Director of Applications at Ross, MasTech and Informatica discuss how modernizing data infrastructure can gain efficiencies as well as valuable insights.

How to Support a Data Driven Institution
Kristin Kennedy, Arizona State University
This presentation will talk about the difference between supporting a traditional institution versus a data driven institution. Wherever you are in your journey to a data driven institution we will talk about how to support an institution now, or how to best prepare for when you get there. Whether you are called Analytics, Business Intelligence, Reporting, Decision Support, Institutional Analysis or any other name, we will talk about challenges we all face and some ideas on how to overcome them. Data is becoming more and more important to everyone so let us prepare for the future together.

Integrating Canvas into Your Data Warehouse, Part 1
Canvas data offers a rich and fine-grained view of student engagement and performance, but extracting data from Canvas in bulk is not easy. We will download data using the Canvas CLI tool, review Salt Lake Community College’s process for getting nightly updates, and discuss the structure of the available tables. Bring your Canvas API key and secret to start accessing your own Canvas data during the session. Sample files with dummy data will be provided for participants without access to API credentials.Abby Kaplan & Rochelle Smits-Seemann, Salt Lake Community College

Integrating Canvas into Your Data Warehouse, Part 2
Abby Kaplan & Rochelle Smits-Seemann, Salt Lake Community College
Canvas data offers a rich and fine-grained view of student engagement and performance, but it’s most useful when integrated with SIS data. We will discuss how to relate Canvas data to your SIS and wrangle it into a useful star schema. Attendance at part 1 is helpful but not required; sample files with dummy data will be provided for participants without access to their own data. Come prepared to upload data files into your own database or sandbox; SQL that illustrates some of the basic necessary ETL will be provided.

Is this a Center of Excellence for Ants?!: Starting Small and Thinking Big for Enterprise Business Intelligence
Nick Chaviano & Rodney Pacis, Georgia Institute of Technology
Over the past four years, Enterprise Data Management has created a paradigm shift concerning data, business intelligence, and governance at Georgia Tech. After the successful launch of the university’s first enterprise data warehouse and the Leading Insight Through Empowerment (LITE) BI platform, campus awareness and interest exploded concerning data and reporting. While LITE provided up-to-date and accurate reports, enterprise users had the appetite for more. This presentation will discuss our BI, data warehouse, and governance strategies, the steps in creating an enterprise center of excellence from strategic planning to launch, and explore our future.

Just in Time Data Management
Successful data governance needs momentum. How can you keep staff mobilized? iData recommends a “just in time” approach to achieve immediate results and consistent progress for all governance maturity levels. Using our Data Cookbook product, you can focus participation in your governance activities without a lot of bureaucracy. With each successive problem solved, you maintain momentum and gradually achieve success. In this session we will show how to manage a business glossary, report catalog, and quality rule catalog; automate assessment of data against quality rules; present highly technical metadata to technical and business users; and control changes to valid values and schemas. The Data Cookbook is a collaborative data dictionary and data management solution. Whatever tool you use, this session offers an approach to governance that gets everyone on the same page quickly, so you can get on with making better informed data-driven decisions.

Lessons Learned as an Early Adopter of Cognos 11 Dashboards
Roxane Meachem, University of Regina
The University of Regina is an early adopter of the web-based IBM Cognos Analytics 11 reporting and ad-hoc dashboarding. This dashboarding is a completely new functionality in Release 11, competing with products like Tableau. For those considering migration from Cognos 10, or new installs (like us), we will provide a quick overview of Cognos Dashboarding and our experiences with it to date, including: general impressions of the functionality vs Tableau; enterprise implementation; and issues in sharing with the user base, especially security, distribution, and licensing. We hope to promote a locus for sharing Cognos expertise and experiences among HEDW members.

Machine Learning and Decision Intelligence: Predicting Student Retention and Drivers
Data warehouse and business intelligence practitioners in higher education are now managing more data from more areas of the university than ever before. Going beyond BI and applying both predictive and prescriptive analytic techniques is the next logical progression in effectively leveraging this data, as has been seen across a variety of other industries. We discuss multiple case studies of applying machine learning and decision intelligence approaches to various points in student’s educational journey. One particular focus will be on student retention: predicting retention likelihood for individual students as well as the main drivers of potential non-retention. This information can be used to accurately identify at-risk students and better address their individual needs.

Mythbusters: Confronting Narratives with Data
Melissa Hartz, Colby College
Every campus has them: stubborn narratives based on gut feelings and squeaky wheels instead of data. Hear how the Office of Institutional Research at Colby College is using institutional data to confront these narratives, and the unexpected “ripple effects” of these analyses.

Netflix for Dashboards and Reports: The UR Data Analytics Portal
Mike Salisbury, University of Rochester
Enabling decision-makers at all levels to easily and efficiently access and understand content is a challenge in gaining adoption and usage of analytics. We now live in an “app world” where accessing information is friendly and intuitive, which has raised user expectations for ease-of-use for accessing dashboards, reports. The UR Data Analytics Portal helps guide the users through personalized galleries of Cognos and Tableau dashboards and reports to the information they need. These galleries are based on the user’s organizational role, categories of interest, personal favorites, as well as organization data domains. This presentation will included a demonstration.

Practical Governance: Using Governance and Warehouse Metadata to Automate Security and Technical Documentation
Joel Dosmann, Augie Freda & Pieter Visser, University of Notre Dame
The term data governance is often thrown around within institution, but rarely with a common definition that triggers action. In this presentation, we’ll take you through our journey of working with Campus Data Stewards on defining a common language, how we manage our metadata in the dataND website, and how the metadata is used to automatically create secured views.

Princeton’s IDEAS for Higher Ed– A Unique Model that Seeks to Provide Business Intelligence Value to Administrators and Contribute Towards Research in Higher Education
Drew Allen & Valerie C. M. Ching, Princeton University
Princeton’s Initiative for Data Exploration and Analytics (IDEAS) for Higher Education was established in July 2017 as an academic research unit by a faculty director and serves as an internal analytics consulting service to the university. In this presentation, we detail the development of this unique admin-research hybrid model and provide examples of our work, including how we connect faculty and graduate students to institutional data and how we combine administrative institutional data from various sources (e.g, enterprise data systems, information warehouse, administrators flat data files, and publicly-available data) and transform it into formats specifically designed for academic research.

Self-Adapting Tables and Integrations: How Ohio State Is Driving Automation into Its AWS DW
Nate Polek & Charles Venci, The Ohio State University
Ohio State is taking on Workday, and moving it’s data warehouse to AWS. Python has drastically changed our architecture, and it’s allowing us to scale much faster. We’ll discuss these technologies: Workday, Airflow, S3, RDS, Redshift, Glue, Tableau, DMS, and more. We’ll also discuss how we got here, and our goals and plans for the next couple of years.

Spend Your Time Working with Data, Not Looking For It!
Dawn Moore & Brenda Ulin, University of Iowa
It’s a cliché: a closet full of clothes and nothing to wear. Apply that to institutional data … a plethora of BI artifacts and yet no one can seem to find the data they need. The University of Iowa tackled this classic information management problem using information management best practices. Discussion will include the impetus for a project to address the problem, challenges encountered, FY19 solution roadmap, and an overview / demo of the solution – called Campus Data – the University of Iowa believes allows their campus to spend their time working with data, not looking for it!

Taking Data Governance to the Next Level – Data Quality Measurement and Monitoring at The Ohio State University
Laura Gast & Meenal Kharabe, The Ohio State University
Ohio State University’s priorities in building and implementing a Data Governance program are defining data terms and improving data quality. Improving data quality to increase confidence in the data is a vital part of data governance at OSU. Data cleansing efforts have been initiated across the university to improve data quality. To provide visibility to key data cleansing efforts, they are actively measured and monitored. A picture is better than a thousand words! Trend and data on these data cleansing efforts is captured, analyzed and graphically presented using Tableau, to increase confidence in the data and promote data cleansing.

The Data Discovery Journey – Where Is Our Data?
Amber MacKenzie & Vasudha Ramani, University of Michigan
At University of Michigan, our CIO asked a simple question.  Where is our data? The answer was not as simple. ITS ventured out on a journey to assess and visualize our institutional data.  We started with the enterprise data supported in our central IT department, and are now moving on to campus. Our first effort assessed the data from a system level – capturing audit logs from integration platforms, databases, web services, flat files and APIs.  The next step of our journey was to create high level “blueprint” diagrams of the data landscape across enterprise systems to demonstrate the complexity and illustrate risks . Our data discovery is our predecessor to our ERP modernization effort.

Third Time’s the Charm: How One University Finally Launched Enterprise Information Management
Beth Prince-Bradbury & Dave Pecora, Rochester Institute of Technology
Google the term Enterprise Information Management and you’ll get a host of definitions. They include phrases like “optimizing the use of information”, “managing data as an asset”, and “enabling business insight.” These are great goals, but figuring out where to start can be challenging. This presentation focuses on RIT’s EIM journey and how – after three tries – we finally gained traction. We’ll discuss our EIM organizational structure, our data stewardship model, and the domain model upon which it is based. Finally, we will share our vision and plan for an enterprise information catalog.

Understanding Complex Higher Education Data Through Visualization
Christopher Gardner, University of Michigan
The ever changing environment of a higher education institution creates complicated data. This presentation illustrates the use of Tableau to visualize some of these complicated data sets in unique ways. The ideas presented will be underscored with examples from U of M such as: 1. Using a Sankey chart to visualize student movement from enrollment to graduation or staff movement from year to year. 2. Comparing faculty salaries against peers using a box and whiskers or highlighted line charts. 3. Examining admissions trends with nested bars. 4. Comparing faculty and staff counting methods with an interactive cross tab. 5. Providing more granular detail in HR using combination parameters.

User Intelligence Powering Effective Business Intelligence
Sandeep Bedadala & Chase McCoy, Indiana University Bloomington
Beginning in 2016, Indiana University-Bloomington implemented a series of decision support dashboards called Academic Metrics 360 (AM360) as a part of the larger Decision Support Initiative (DSI). As AM360 and DSI have evolved, so has our need to more actively engage with our users: institutional decision makers. This presentation will present the steps we have taken and what we have learned in the process of increasing the reach and usage of AM360 across the Indiana University system through user assessment and by using the assessment results to design a holistic user experience by leveraging the power of web communications.

Using Tableau with WebI: From Novice to Knowledge
Barbara Hoover, University System of New Hampshire
With more than a thousand hands-on consumers of 32,000+ reports, WebIntelligence is our go-to reporting solution at the University of New Hampshire. Then came Tableau, with its bubble charts, dashboards and data stories. Not to worry! Tableau experts were hired, so we WebI developers could stay happily ensconced in our reporting world. That is until this year, when our team was tasked with creating Tableau solutions using WebI data. This case study will walk through our learning process, some pitfalls and best practices, and demonstrate the creation of a call center dashboard using WebI data and Tableau visualizations.

What Makes a Difference? Cloud Automation of Near Real Time Student Retention Prediction
Robert Snyder & Andrew Siefert, Ithaca College
In forecasting and addressing year-to-year student retention, effecting proactive measures to improve retention requires not only the prediction of the specific retention probabilities for each individual student, but also that those predictions be produced early and often in response to changing student behaviors. The analytical modelling and machine learning behind just-in-time retention prediction has historically been complex to implement. In this presentation, members of the Analytics and Institutional Research team at Ithaca College will present their solution to a fully automated, cloud-based, just-in-time student retention framework – a framework largely built on freely available open-source tools and technologies.

What to Consider when Starting Predictive Analytics at Your Institution
Heather Chapman, Weber State University; Craig Rudick, University of Kentucky; Drew Allen, Princeton University; Yuko Mulugetta, Ithaca College
Do leaders at your institution ask you for more and better data? Are you being asked to predict into the future how your students will behave, or what characteristics they will have? Have you considered using predictive analytics at your institution but don’t know where to start? If any of these fit your situation, this session may be for you. This open panel of four HEDW peers from both large and small campuses is intended to answer questions of those new to or considering using predictive analytics on their campuses. Any and all questions are welcome.

When Harry Met Sally, Again, After 3 Years of Courtship: A Retrospective on How Princeton Has Actively Supported Multiple BI Tools and Is Taking a Look into Their Futures
Ted Bross & Amy Such, Princeton University
This is a 3 year retrospective on the successes, failures and lessons learned at Princeton in actively supporting multiple BI platforms at the enterprise level. The original presentation, given at the 2016 HEDW Conference, mapped what and how this would be accomplished. This presentation compares those expectations with reality and also gives a glimpse into what happens next.

Where Did They Go: Using BI to Build an Institutional Repository for National Student Clearinghouse Data
Steven Lonn, Carson Phillips, Julie Martinez, Jeffrey Jenkins & Michael Filipiak, University of Michigan
The National Student Clearinghouse (NSC) has student enrollment information for over 98% of institutions in the United States – most institutions utilize this resource for ad-hoc queries. In this presentation, a team of institutional researchers and business intelligence professionals will describe how the University of Michigan has built one of the first known institutional data warehouses for NSC data. The presenters will explain how this project has automated manual tasks and normalized data into student cohorts that can be used to answer key questions about students’ academic pathways that will hopefully lead to improved student success outcomes.

Workday Financial Data at Scale: The Good, the Bad and the Ugly
Jason Green & Esmeralda Angel, Arizona State University
Arizona State University’s journey to extracting data from Workday has been a long and winding one. What seemed doable at first, became difficult once live data was being loaded and soon became out of control. This presentation will discuss the patterns and tools we adopted to get this Workday data tamed—what worked well and what we had to partition into six thousand RaaS calls to work. Once we had the data in our Redshift Data Warehouse, there was still work to be done to do to keep our users happy.

Your Institution’s Data Superheroes – Use Their Power for Good
August Freda, University of Notre Dame
Explore the roles and attributes of typical Data Superheroes at your institution and how to leverage their expertise and influence for the good of the institution as a whole. These superheroes hold titles like: “Data Aggregator”, “Data Collector”, “Data Protector” and “Data Whisperer”. Each of these roles has their own particular set of data superpowers. Always with the best of intentions, over time their superpowers mask bad or missing data, “integrate-by-Excel” and spread data outside of security and access controls. But, this domain knowledge and their reputation within the institution can make a significant difference in improved data-driven decision making.