The Director's Blog

Editor’s Note: CSFE helped send Ally Bevers, a senior math and econ major, to the Claremont Graduate University Empirical Workshop held June 10th through the 13th, 2019. Ally is planning to enter a doctoral program in finance or economics next fall, so she is already putting to work the research skills that she learned at the workshop. Hope you enjoy her short reflection on the experience. 

The Claremont Graduate University Empirical Workshop was an amazing experience and I am beyond thankful for the opportunity to attend. It was located in beautiful Claremont, California and the attendees were from all over the country. Well over 200 students applied and I was lucky enough to be one of only three undergraduates who attended. Along with learning about empirical theory and methods, I got to make friends and hear about their graduate school experiences in a crucial time where I am preparing to make decisions about graduate school myself.

Ally Bevers (back center, in gray) poses with other attendees from the Claremont Workshop

The workshop covered a wide variety of topics during the week. We heard from many different professors on the topics of workflow, randomization and randomization inference, causal inference, regression discontinuity, instrumental variables, difference-in-differences, synthetic controls, hidden curriculum, machine learning, and web scraping. If it seems like a lot to fit into a week, trust me, it was! We worked from 9 am – 12 pm, got an hour break for lunch, and worked from 1 pm – 5 pm. Each day was very content-heavy, and we moved quickly through the material. The workshop was designed for 2^nd– and 3^rd-year graduate students, so there were times I found myself lost because I was not familiar with the content. However, this challenged me to make the best of my situation and take a moment to breathe and reorient myself so I didn’t get frustrated. This experience was important for my personal development, as generally in my undergraduate courses I am on top of the content and don’t tend to get lost. Challenging myself in this way helped me grow as a student and person.

The content I found most useful was workflow, hidden curriculum, and difference-in-differences. I am currently doing the Summer Undergraduate Research Program at Western Carolina University and my research project uses difference-in-differences. Going back through the history, mathematical theory, and code in Stata was extremely beneficial and I took notes to help me with my research. The hidden curriculum was everything we needed to know in order to run productive regressions that we wouldn’t have learned in econometrics. We learned how to avoid mistakes in workflow so we didn’t make major errors in our programming. This can be done by having organized subdirectories, automating the creation of charts and graphs, having a common naming convention, controlling versions of your work, and annotating your code so you will remember what you did and why when you go back and look at them. I was so inspired by the workflow talk we heard that I went through my laptop folders and reorganized and renamed my files in the airport before I came home.

The most interesting content was the coding. I loved learning about machine learning and web scraping in R and I took notes on different resources to look into so that I could learn how to code better. It was so intriguing to see what was possible with just a few lines of code. I got excited about applying the code to real-world examples, and along with the advice I had received from people throughout the week, it helped me become closer to making a decision about my future plans.

Overall, the Claremont Graduate University Empirical Workshop was integral in shaping my decision about the future. It provided me with the insight I needed about what economics PhD programs are like through hearing real experiences of others and seeing the level of content taught in an economics PhD program. I got to receive unbiased third-party opinions from professors whom I hadn’t met before but plan on keeping in contact with. It was truly a wonderful experience and I am so thankful for the opportunity to attend.

This entry comes in the form of a letter from Sean Duffy to Angela Dills, who is the project director on the North Carolina Data Dashboard. Sean has contributed as one of four student workers on the Dashboard. He has a great story, only a small slice of which is captured in his endearing letter below. Enjoy.

Dear Dr. Dills,

I wanted to reach out and pass along a thank you letter. I recently moved to Madison, WI, following a dream job for Trek Bicycles Corporation as a full-time Business Intelligence(BI) Analyst.

I want to thank the College of Business at WCU and more importantly, the Center for the Study of Free Enterprise for helping to pave the way to a stellar career. Looking back I realize just how critical my experience from the NC Data Dashboard was in making me an eligible candidate for a mid-senior level position at the world’s largest bicycle company. The minimum requirement for this position was three years in an IT Business Analysis Role with at least two years of hands-on experience working with SQL Server BI tools. I feel confident to say that if it weren’t for the NC Data Dashboard that I wouldn’t of even come close to landing the initial interview. Upon three promising phone screenings, I was scheduled to fly up for an interview and had received a salary offer within 24 hours!

Throughout my interview, I was able to speak clearly and passionately about my work on the NC Data Dashboard and the CIS curriculum. The experiential learning opportunity I gained from my work on the NC Data Dashboard is invaluable and ultimately was the ticket to landing a job of this caliber. Feedback from the interviewers suggested that it was unique to see a recent college graduate with real business level experience using the same softwares/tools that their company uses. I was able to speak about real business experiences and troubleshooting technical issues rather than a topic or theory learned in a classroom. The CIS curriculum built the necessary foundation for me, but it was the NC Data Dashboard and knowledge gained in the early days of the project from Brad Bergh that shaped me into the perfect candidate.

Lastly, I wanted to thank WCU and CSFE for allowing Daniel Hartness and myself to travel to the World Wide Data Vault Conference to deliver a presentation this past May. This conference brought everything that I have learned and applied over the past four years into a complete circle. I wanted to give a truly heartfelt thank you to everything you all do to make these projects and opportunities possible for students. It’s an honor to of been a part of such a great program and hope to hear of many more success stories from students alike.

I hope you all had a great summer and are ready for another semester!

Thanks,

Sean Duffy

Computer Information Systems

Class of 2019

Think of the last book you read. Was there a dedication? Most likely it was to the author’s mother, spouse, child, friend, mentor, or colleague. At the very least it is very likely that the book was dedicated to someone known to the author personally, right? The same thing goes for dissertations, with many being dedicated to parents. Mine, however, was different. My 6-year-in-the-making tome was dedicated to three men that I have never met, but without whom my research would have been impossible. Those men are John Young, Ken Mattingly, and Charles “Charlie” Duke Jr.—the Apollo 16 astronauts who returned 211lbs of rocks from the lunar surface in 1972, just shy of 11 years after President John F Kennedy challenged the US to send an American to the Moon and return him safely before the end of the 1960’s.

This was an amazing accomplishment and remains one of the most incredible feathers in the American hat.

(L-R) Charlie Duke, John Young, and Ken Mattingly in the recovery raft after splashdown back on Earth at the end of the Apollo 16 mission (NASA image S72-3950).

Twelve in All of Civilization

Apollo 11 landed on the Moon in 1969 and was followed by 5 more successful manned landings between 1969 and 1972 before the Apollo program was cancelled due to funding. Three men were on each mission, but only 2 walked on the surface while one stayed in the orbiting spacecraft (in the case of Apollo 16, this was Ken Mattingly). That means that in the entire history of humanity, only 12 people have set foot on another planetary surface. Twelve people in thousands of years of human history! And one of them is coming to our little slice of heaven in Cullowhee, NC—Apollo 16 Lunar Module Pilot and the 10^th man on the Moon, Charlie Duke.

Charlie Duke on the surface of the Moon at station 1 during the first extra vehicular activity (EVA). Duke is standing next to Plum Crater (131 ft wide, 33 ft deep) with the lunar roving vehicle behind him (NASA image AS16-114-18423HR).

We’ve Only Scratched the Surface. Literally.

Although we have sent humans to the Moon several times, we have hardly begun to explore it. All of the Apollo missions landed on the nearside of the Moon (i.e., the side that is always facing Earth due to tidal forces) relatively close to the equator.

Prior to the Apollo program, the NASA robotic Surveyor missions landed in similar locations on the nearside—Apollo 12 astronauts landed so close to Surveyor 3 that they walked over, took a piece of the lander, and returned it to Earth to study the effects of space weathering and micro-meteorite impacts. The Soviets also sent robotic landers (and rovers) to the Moon with Luna and Lunokhod, respectively, and had several successful sample return missions (1970-1976) using these instruments; these also landed on the nearside, but a little farther spread than the US missions. Fast forward to 2013 and the Chinese became the third nation to touch down on the lunar surface with a robotic lander. Want to guess where it landed? Yep. The nearside.

Why so many to the nearside? Communication. Because the Moon is tidally locked with Earth such that one side is perpetually facing us, we cannot see the farside (NOT the “darkside”) of the Moon from Earth, which makes communicating with humans or a spacecraft difficult. It wasn’t until December 2018 that a lander was successfully sent to the lunar farside by the Chinese—a tremendous feat to explore an area no lander or human had visited before (although we do have high-resolution imagery).

Even if we look at just the Apollo 16 mission, where Duke and Young drove a rover approximately 7 miles, they still did not cover much of the lunar surface. For comparison, take a look at the image below of Cullowhee overlaid by the paths taken by the Apollo 16 astronauts (this image is at the same scale!).

Apollo 16 traverse (to scale) overlaying a Google Maps basemap of the Cullowhee, Sylva, Dillsboro area showing the extent of exploration of Duke and Mattingly on the surface of the Moon.

Imagine that you are an alien species visiting Earth and this is your species’ 5^th landing, but first time to land in this location. Follow that path and let me know what you learned about Cullowhee, WCU, North Carolina, the USA, or Earth It might just be easier to ask it the other way…what did you miss? Deserts, glaciers, volcanoes, oceans, lions, tigers, bears (oh my! Actually you might see a bear). The point is that even with 6 manned missions to the Moon, we have not explored much of the surface and therefore the rocks that have been returned are not entirely representative of the Moon.

Just a Big Gray Rock?

Big deal…the Moon is all just a grey rock anyways, right? Wrong! We know that there are many different types of rocks on the Moon composed of the same minerals that we have here on Earth. (That is not surprising when you learn that the Moon coalesced from impact debris of a Mars-sized object hitting Earth over 4 billion years ago.) When you look up at the night sky, you will see light grey (the highlands where Apollo 16 landed) composed of a rock called anorthosite and black composed of a volcanic rock called basalt. These dominate the surface, but there are certainly more types, each that give geologists more information of how the Moon formed, cooled, and evolved over the past 4.5 billion years. In fact, the materials on the lunar surface are not all grey or black either—Apollo missions returned glass beads that were green and orange and formed from processes associated with large asteroid impact events and volcanic eruptions, respectively. We know all of this (and more) because the Apollo missions brought back samples.

Old Rocks, New Discoveries

Despite humans barely scratching the surface of the Moon, geologists have made many discoveries about the Moon that have completely changed our understanding of the Moon’s origin, the presence of water in and on the Moon, and the timing and composition of asteroids hitting the Moon (and Earth) over billions of years. The Apollo astronauts returned rocks that geologists (myself included) continue to study with better instrumentation that give us higher resolution information each year. These rocks keep providing clues to questions we didn’t even know we should have back in the 1960s.

Astronaut Charlie Duke noting the collection location of a sample (67935) at Outhouse rock (~ 30 ft across; NASA image AS16-116-18649)

This is why it is spectacular to have Charlie Duke, one of only 12 men to have walked on the Moon, visit WCU. The samples returned by the Apollo missions are the gift that keeps on giving to science and humanity. The astronauts risked everything to reach another planetary surface, and for millennia to come we will treasure these samples as they continue to open our eyes and minds to the secrets of the solar system. Thank you to those men for the greatest gift.

About the Author

Dr. Amy L. Fagan is the Operations Chair of the Lunar Exploration Analysis Group, an Assistant Professor in Geosciences and Natural Resources at Western Carolina University with an expertise in lunar geology, and a complete Lunatic.

First of all, it’s great to have the opportunity to contribute to the Center’s blog. I wanted to tell the story of me being one of four student workers on the North Carolina Data Dashboard, and how that has led to amazing opportunities in data science and CIS curriculum.

One of my student co-workers on the Dashboard, Sean Duffy, and I recently were the first and only students ever to make a full presentation at the World-Wide Data Vault Consortium (WWDVC). The WWDVC is an international data science conference aimed at helping improve data processes and other business intelligence functions. Our goal at the conference was to explain how we used the Data Vault 2.0 methodology to build and operate the North Carolina Data Dashboard Project (NCDD). We could not have anticipated the enormous response we received from conference attendees.

After giving our presentation, we became the biggest hit of the conference. A crowd of people came up afterward to ask us questions and encourage our innovative application of the Data Vault 2.0 method. We ended up interacting and talking at length with the founders and pioneers in the field. Everyone wanted to talk to us and were curious about the great things that we are doing at our university. It was an incredibly valuable experience that was only possible due to the support of Professor Dills and the Center.

Sean Duffy (Left) and Daniel Hartness (Right) present at the WWDVC Conference.

Data Vault and the NCDD:

In case you didn’t know, Data Vault 2.0 is the most efficient way to structure, build, maintain, and work with data and data warehouses. In short, as more and more organizations reach the point of having to manage big data, just as we did with the NC Data Dashboard, it becomes harder for them to scale effectively.

One temporary, but expensive solution that is commonly used is restructuring and buying more storage. However, this solution only works for so long, as it eventually becomes infeasible. Data Vault 2.0 allows for natural scalability, no matter the data size, without having to restructure each time data needs to be added. It also incorporates Agile and Six Sigma methodologies.

These factors, along with many others solidify it as the future of the data warehousing industry. This annual conference acts as a gathering of the most brilliant minds in the industry to reflect and showcase new innovations with Data Vault 2.0 and what the industry looks like going forward.

Bringing Home Ideas

There are several reasons why this conference was so beneficial, the first of which was the numerous vendors that attended. Among these were Snowflake and WhereScape. These specific ones demonstrated some of the most cutting-edge pieces of software available. Before attending, I didn’t know technology like this was even available. The conference setting served as an easy way to speak to each of the individuals representing these companies on a personal level.

Each of the groups were eager to learn about our position as students relative to our project and to discuss the possibility of using these tools to enhance the NCDD Project. That opportunity would be incredible. WhereScape is a metadata-driven automation and code generation tool used to model and build the data warehouse and all ingest processes. Using it as a data-modeling tool would shorten our development processes immensely, increasing our efficiency by a factor of 3 – 5. This means the NCDD could add between 3 and 5 new data series in the time it currently takes to add one data series. On the same token, Snowflake would decrease the time it takes to structure our data due to their more efficient innovation on SQL: SnowSQL. Furthermore, Snowflake is a cloud-based service. That means both storage and the updating of that storage wouldn’t be a concern for the NCDD team. That would cut costs, freeing up funds for further expansion of the project. Even more importantly, this technology could be used by WCU for all of the same reasons, just on a more massive scale.

The idea of not having to continuously purchase expensive storage space as well as moving important systems to the cloud could be revolutionary for the university. As far as expenses are concerned, Snowflake charges per second per node (computer) of use, giving it a competitive edge on many other technologies. Combining this with Data Vault 2.0 could put WCU in an even better position compared to most of the other colleges in the country and exponentially increase efficiency overall.

Connecting with Founders and Pioneers:

Another key piece of the conference was the relationships Sean and I built with other attendees. The first is Dan Linstedt, the creator of Data Vault. He gave Sean and I the chance to be the first and only student presenters to attend the conference and share our story. He created Data Vault when he was working for Lockheed-Martin in the 90s. Lockheed-Martin was in a unique position. It was dealing with some of the largest datasets in the world at that time. A rough estimate of the size of the data was around 15 terabytes, which was unheard of. 15 terabytes is small potatoes in today’s world, but it was expensive to buy more storage and nearly impossible to restructure in the event of introducing new, large amounts of data. Dan was tasked with finding a way to fix this problem and Data Vault was the answer. More and more organizations are now reaching this point and turning to Data Vault. Dan shared this history with us and helped us understand why Data Vault is the future. Dan found a solution to the problem so many companies are facing today, twenty years ago. What was even more amazing was that he vowed to help us going forward in sharing the importance of this.

Another individual we met was Bill Inmon, who is known as the “Father of Data Warehousing” and was a keynote speaker for the conference. Where Dan created the superior way to house data, Bill created the entire industry. It was daunting when we realized this, and he wanted to spend some time getting to know us. He talked to us in the hour before our presentation about everything other than data. He just wanted to sit and get to know us. He offered us advice for our paths going forward in data and encouraged us to just go and enjoy our presentation, not to be nervous. Bill also explained to us that in early August of last year, he was in a fairly rough accident, where his lungs were damaged. A few problems resulted from this including sporadic coughing and chronic pain. Because of the pain, Bill would typically leave after his presentations, (his keynote being presented the morning of ours), greeting a few folks, but leaving the same day to go home and rest. After meeting us however, he not only waited to see our presentation, but extended his stay into the next afternoon in order bid us farewell. This was a true honor and was incredibly meaningful. His actions, along with the slough of praise given by the other attendees, let Sean and I know we were truly in a unique position.

In conglomeration with meeting Dan, we had the privilege of having dinner with a number of other genius individuals such as Cindi Meyersohn, President of Data Rebels, Kent Graziano, Chief Technical Evangelist at Snowflake, Data Vault Master Sanjay Pande and others. Talking with these folks at length brought me true inspiration for the future of Data Vault and the potential future WCU could have using it and getting involved more in Data Analytics directly through the CIS Program.

Pictured from left to right: Dan Linstedt, Sean Duffy, Bill Inmon, and Daniel Hartness at the WWDVC Conference

Lessons for the Curriculum:

This is what I believe to be something with incredible potential for WCU. Currently, our CIS program doesn’t specialize in any specific aspect, but includes crash courses in many subjects like networking, SQL, and Web Development. I’d like to propose that we move to specialize in Data Analytics as well as business modeling. The benefits that would result from this are objectively huge. The first of which, is the preparation of our students for the real world and business. Part of being a successful CIS major is acquiring and tempering the ability to be a communicative bridge between the Information Technology and the administrative sides of a business. This is of course the reason we take several business core classes, but we are missing more technical skills. Currently, we learn a bit of what data is and how to query it, but the meaning behind our queries and manipulation of the data is lost. In order to analyze data correctly, we have to be able to understand several things about the business/client that is requesting it: What is the information flow in the business, what do they care about most, what are their goals, and what are the questions the business needs to answer? This is where Business Modeling would come into play very naturally, because of our situation in the College of Business. We could incorporate some of the latest tools like Snowflake and WhereScape for experience in both modeling and warehousing. Along with this, we could create more opportunities for students to have Experiential Learning by using our upper level classes to maybe both model a business and its data as well as use analytics to help it grow. This would pair well with MGT 404 and widen the scope of the CIS capstone. Additionally, more project opportunities like the NCDD project could result from this focus, giving students more business experience as well. This would help to continually fulfill WCU’s mission for student Experiential Learning. If we teach the core principals of Data Management/Analytics as the key enablers of business strategy execution, our students will be capable of making immediate significant business contributions upon graduation.

To this end, Dan Linstedt and Sanjay Pande have offered to advise the CoB in establishing a Data Vault 2.0 focused Experiential Learning curriculum in a Data Management/Analytic context. Chris Stewart, WhereScape GM and VP for U.S. Operations, has offered similar support and licensing to WhereScape as part of a Data Vault 2.0 focused Data Management/Analytics curriculum. Kent Graziano has offered advice and access to SnowFlake for students and faculty also in conjunction with a Data Vault 2.0 focused Data Management/Analytics curriculum. Brad Bergh, a 30+ year veteran of Data Warehousing and Decision Analytics, has offered to work with CoB faculty to define a Data Management/Analytics Experiential Learning curriculum incorporating Data Vault 2.0, WhereScape, and SnowFlake.

Just Wow:

This conference was completely eye-opening for the future of data. Sean and I received a degree’s worth of education from our week there and I’m eager to help bring it back for the benefit of WCU and it students. We would prepare our students to be shining stars in the industry upon graduation if we specialize our CIS program in Data Management/Analytics and give them many opportunities for Experiential Learning. So many people are ready and willing to help us succeed and propel our college forward. All we have to do is ask.