Preparing Analytics-Enabled Professionals in Finance Using a Simultaneous Team-Teaching Approach: A Case Study

Abstract

To meet the demands of industry, undergraduate business curricula must evolve to prepare analytics-enabled professionals in fields such as finance, accounting, human resource management, and marketing. In this article, we provide a case study of developing a rigorous, integrated finance and data analytics course that was delivered using a simultaneous team-teaching approach within a regional, teaching university. This case study describes developing the learning outcomes; defining the integrated teaching approach; identifying and developing course content and materials; and planning the course delivery. Next, we describe the course delivery, including student achievement of course learning outcomes, and a summary of student feedback from end-of-course evaluations is provided. Finally, we present lessons learned from delivering the course and provide considerations for future work such as scaling this course and replicating it in other business domains.

KEYWORD:

• Analytics
• Business
• Data science
• Finance
• Interdisciplinary
• Team teaching
• Undergraduate

1 Introduction

The world has been transformed by the evolving field of data science, which builds on an array of technological developments and the massive amounts of data being generated by business, science, and society (National Academies of Sciences, Engineering, and Medicine 2018). As a result, the needed skill sets for graduates and employees are becoming increasingly focused on data and computation (Donoghue, Voytek, and Ellis2021).

In higher education, and specifically in the business disciplines, the transformation of curriculum is not keeping pace with the needs of industry partners. A gap exists in the preparation of analytics-enabled professionals who have deep knowledge in their domain (or field) while also having fundamental data science skills such as the ability to effectively interpret data to inform decision making (Gundlach and Ward 2021; Peng et al. 2021; Vance 2021). In 2020, 67% of the 2.72 million estimated job openings in data science were for these analytics-enabled positions, creating an imperative for higher education to further evolve their data science and analytics offerings (Business-Higher Education Forum and PricewaterhouseCoopers (PwC) 2017).

To address this imperative, a course was designed that directly integrated analytics and modeling into the finance curriculum. In addition, this course was team taught with both finance and analytics faculty members being present for each class session (known as “simultaneous team teaching”) to reflect the importance of integrating these fields. The remaining sections of this article will describe the need for analytics-enabled professionals within finance; how we shaped a simultaneous team-teaching approach to address this need; the delivery plan for the resulting course and its evolution; the student achievement of learning outcomes; the faculty lessons learned from this experience; and future curricular development work that should be undertaken to address the growing need for analytics-enabled professionals.

2 Need for Analytics-Enabled Professionals in Finance

Organizations are under constant pressure to adapt to changing conditions and innovate their operating processes. As a result, organizations must be agile and able to make quick decisions related to their strategies, tactics, and operations. Processing the data, information, and knowledge to support these decisions usually requires some technical and computational support, which is where analytics becomes a core competency for leaders, managers, and professionals.

A key challenge for meeting this need is a sufficient workforce with the necessary knowledge and skills. As business schools evolve their curricula, they must decide how best to integrate analytics offerings into their existing program domains such as accounting, finance, marketing, operations, information systems, and human resource management. Prioritization of this work has often been domain dependent and based on industry need. For example, a 2015 study found that finance is one of the main areas in which analytics is applied (Lismont et al. 2017). Therefore, undergraduate finance programs must evolve to include analytics within their core curriculum to ensure their graduates can compete now and in the future.

Finding qualified faculty is a key challenge in effectively modifying undergraduate finance curricula. Existing faculty may have been trained before current analytics models and technologies were widely available and integrated into finance graduate programs. One approach to addressing this mismatch is rigorous development and simultaneous team teaching of a course integrating finance and analytics, which is described in the remainder of this case study.

3 Shaping the Simultaneous Team-Teaching Approach

We (the authors) are business faculty within a regional, teaching university: one faculty member within finance and one faculty member within analytics. After identifying the potential value of an integrated finance and analytics course within our undergraduate finance program, we discussed multiple approaches for design and delivery. This discussion included review of Management Education at Risk, a 2002 Report from the Management Education Task Force of the Association to Advance Collegiate Schools of Business (AACSB). This report identifies that business programs often have functional silos that do not align with actual business problems or solutions (especially in technology-related areas). In addition, the report encourages boundary-spanning teaching approaches (AACSB Management Education Task Force 2002). This course provided the perfect opportunity to span the boundaries that exist between finance and analytics.

As discussions continued, we quickly realized the skill sets of both faculty members would be required to sufficiently support student learning. For example, although the finance faculty member could learn programming and modeling methods from the analytics faculty member, this approach would likely result in deep coverage of financial concepts and shallow coverage of analytical methods and models in course delivery. In addition, collaboration of faculty members can be a strong predictor of student gains; model professional interactions for learners; improve instructional effectiveness of faculty; and support ongoing learning for and development of faculty (Wenger and Hornyak 1999; Shibley 2006; Sandholtz 2020). Because of these factors, we decided to use a team-teaching approach for designing and delivering this course.

At the most basic level, “team teaching involves a group of instructors working purposefully, regularly, and cooperatively to help a group of students learn” (Buckley 2000). This approach, though, can take a variety of forms from allocation of responsibilities to joint planning, instruction, and evaluation (Perry and Stewart 2005; Sandholtz 2020). Based on the integrated nature of the content we planned to deliver, significant collaboration and communication would be required. Therefore, we decided to use a simultaneous team-teaching approach to work together in designing and delivering this course. Following best practices, we met regularly over several months to jointly design and plan the course (Helms, Alvis, and Willis 2005; Backarach, Washut Heck, and Dahlberg 2008; Bajada and Trayler 2013; Pope-Ruark, Motley, and Moner 2019; Sandholtz 2020).

Based on this approach, the university’s institutional review board (IRB) determined this work did not require IRB submission or review due to not meeting the definition of a systematic investigation.

3.1 Developing Learning Outcomes

The first step in course design and development was the identification of learning outcomes for student achievement (Wenger and Hornyak 1999; Shibley 2006; Sandholtz 2020). We conducted a review of research to identify the core skills for analytics and data science positions. The most common skills identified included domain knowledge and context; data cleaning and wrangling; data visualization; computation techniques and tools; statistical and machine learning methods; and inquiry and communication skills (Nolan and Temple Lang 2010; American Statistical Association Undergraduate Guidelines Workgroup 2014; Business-Higher Education Forum and PricewaterhouseCoopers (PwC) 2017; De Veaux et al. 2017; Donoghue, Voytek, and Ellis 2021; Gundlach and Ward 2021; Schwab-McCoy, Baker, and Gasper 2021).

Based on this review and considering the focus of this course (integration of finance and analytics), we determined that student learning would focus on the following outcomes:

• Identify and describe business decision making scenarios appropriate for financial modeling.

• Compare and contrast common financial models such as forecasting, capital asset pricing, credit risk analysis, portfolio analysis, and Monte Carlo simulation.

• Demonstrate data wrangling techniques required to prepare data for use in financial modeling.

• Create and interpret appropriate financial models.

• Professionally communicate the process and results of financial modeling to a variety of stakeholders.

3.2 Identifying Student Population

With the learning outcomes identified, the next step was to identify the target student population for this course.

The university has undergraduate degree programs in finance and actuarial science, which focus heavily on critical thinking and quantitative techniques. Students in these programs are required to take core courses in the finance area. As the finance faculty member would meet with alumni from these programs, recent alumni would indicate the importance of analytics skills in their professional positions. In addition, older alumni and employers would express a desire to hire interns and graduates with analytics skills. Therefore, finance and actuarial science students were identified as the target population for this course during its initial delivery.

Within these programs, the typical class size is 15–24 students. To have the necessary enrollment to make running the course fiscally sound and obtain feedback on the course, administration agreed that this course would substitute for an existing course requirement within the finance major and minor. Leadership and faculty within the actuarial science program were individually contacted to share information about the course and its contents. With this approach, we expected 10–20 students to enroll.

To be successful in this new integrated course, students would need a background in computation and basic finance concepts. Within our university, all students pursuing a Bachelor of Science (B.S.) degree are required to take two 4-credit courses in computational thinking that integrate learning of statistics and computer science skills. In these courses, topics include data organization, analysis, visualization, and interpretation, as well as techniques within descriptive statistics, correlation, regression, ANOVA, and basic experimental design. Spreadsheet software (such as Excel) or other statistical software are used for the analysis, which may include some scripting. For the first required finance course, the prerequisites are accounting, computational thinking, and economics. Basic finance concepts are covered using the Fundamentals of Financial Management textbook (Brigham and Houston).

The new integrated finance course applies computational thinking concepts within a finance context, which provides a powerful foundation for the higher order application that would be required for the new course. Therefore, the computational thinking courses and this basic finance course were identified as prerequisites for the new course.

From a scheduling perspective, the course would be offered immediately following the first required finance course. This placement would allow us to recruit students for the new course from those students taking the required finance course. In addition, students would have recent coursework that would support their learning of the more advanced financial concepts and techniques planned within the new course content.

3.3 Defining the Integrated Teaching Approach

Knowing that most students taking the course would be finance or actuarial science majors, the next step was to determine the teaching approach that would be used to integrate the content areas and provide an inclusive environment for all students.

We agreed that hands-on, real-world applications of the course concepts would be critical for student learning, as supported by research in teaching data science and analytics (Nolan and Temple Lang 2010; American Statistical Association Undergraduate Guidelines Workgroup 2014; Soesmanto and Bonner 2019; Charles A. Dana Center at The University of Texas at Austin 2021; Donoghue, Voytek, and Ellis 2021; Gundlach and Ward 2021; Peng et al. 2021). Therefore, we quickly identified that a significant amount of the in-class portion of the course would involve working on “labs”—problems that required use of analytics techniques and models to support financial decision-making scenarios.

To support these labs, students would need a solid background in financial concepts such as capital asset pricing and portfolio management. We decided that the finance professor would review core financial concepts prior to introducing the lab to “level the playing field” for all students. In addition, most of the finance curriculum relied on the use of Microsoft Excel for computational work, so the finance professor would walk through any related spreadsheet models and computations in Excel to solidify student understanding.

Once the financial concepts and computations were reviewed, then the analytics professor would introduce the analytics models and methods that could be used to support the selected financial decision-making scenario. For students to be successful in this work, foundational concepts in data quality, structures, and wrangling would be covered in the first several weeks of the course. In addition, basic programming concepts (such as variables and control structures) would be reviewed. With this structure, we would be able to demonstrate the integration of knowledge that is critical for successful careers in analytics and data science (De Veaux et al. 2017).

We determined that we would both be present during in-class meetings and would hold one combined office hour each week. By using a simultaneous team-teaching structure, we could enhance our communication with students by providing in-depth and integrated answers to questions posed and being consistent in our instructions. In addition, our approach aligned with best practices for inclusive data science education (Charles A. Dana Center at The University of Texas at Austin 2021). First, by reviewing previous financial concepts and techniques, students could experience the transfer of previous knowledge and skills to a new context. Second, the in-class lab sessions created a safe environment for students to explore and struggle with techniques while having immediate support from two instructors and their peers. Third, the review and lab sessions included low-floor, high-ceiling activities accessible to all students, which built students’ confidence while identifying areas where further instruction and assistance was needed. Finally, students observed two instructors, from different fields, learning from each other and collaborating to solve problems, modeling behavior necessary for success in their future careers.

We also selected this structure to embed communication opportunities throughout the course and support student achievement of the final learning outcome (“Professionally communicate the process and results of financial modeling to a variety of stakeholders”). First, students would have the opportunity to discuss their work with each other during the lab sessions, which would demonstrate different ways to describe and present results (Parke 2008). Second, as students asked questions and explained their work to us, we would quickly identify any misconceptions or confusion, provide feedback, and support students in improving their oral communication skills (Parke 2008; Khachatryan and Karst 2017). Finally, each of these labs would build the analytical and communication skills needed to complete the larger cases within the course. These cases would require a written report for students to demonstrate communication of their approach, analysis, interpretation, and overall results (Khachatryan and Karst 2017).

3.4 Selecting an Analytics Programming Language

With the simultaneous team-teaching approach determined, the next step was to select an analytics programming language to be used within the course.

As Microsoft Excel is a common tool in finance departments, some have relied on Visual Basic for Applications (VBA) macros to automate tasks. Because the finance professor planned to review core financial concepts using Excel, VBA was the first language we evaluated. Although macros can be powerful for some tasks, they do not have built-in analytics support and can be an entry point for malicious software; therefore, use of VBA macros was not considered for this course. Based on this decision, when reviewing concepts, the finance professor would focus on spreadsheet models (variables, computations, interpretations), and these models would not include macros created using VBA.

Although many analytics programming languages and platforms exist, R and Python are generally listed in the top three (Babu 2019; Kumar 2019; Piatetsky 2019; Some 2019; Schwab-McCoy, Baker, and Gasper 2021). Python is cited as the best multi-purpose language that is relatively easy to learn and provides a good “point of entry” for those learning analytics and data science (Theuwissen 2015; Kervizic 2020; Paruchuri 2020; Saha 2020). In addition, the computational thinking courses required for the target student population often included an introduction to Python. Therefore, Python was chosen as the analytics programming language.

Although Python can be a good starting point for learning coding and modeling, code editors and integrated development environments may provide minimal support forin-depth graphical and textual explanations of concepts and code. Therefore, we chose Jupyter Notebooks to present Python coding to students. Within these notebooks, we would document background information, concepts and code involved, modeling steps, and evaluation and interpretation of the model. For example, a flowchart graphic could be included to show the steps involved in creating a specific model, followed by the code that represented these steps. These elements would be documented similarly within each notebook to provide a consistent interface and flow for student learning. In addition, students would have an example and format to follow when writing their lab reports (Spurrier 2001).

3.5 Identifying and Developing Course Content

A final step in the design and development of the integrated finance and analytics course was detailed identification and development of the course content.

The course was structured to follow a standard 15-week undergraduate semester schedule with three 70 min in-class meetings each week (Monday, Wednesday, and Friday). Classes would be held in a computer lab on campus that provided access to all the necessary software for the course.

For the technology and software needed for the course, we collaborated with the university’s information technology (IT) staff to ensure students would have access to the necessary resources (on and off campus). Students were provided with instructions for using their personal devices or any open campus computer lab to study and complete homework. If students had any concerns about the technology or software requirements for the course, they were encouraged to speak with the instructors to determine alternative arrangements as needed. In addition, IT staff participated in early in-class sessions to troubleshootissues.

Like others who have attempted to create boundary-spanning courses, we could not find an existing textbook that adequately addressed the content planned for this course (Helms, Alvis, and Willis 2005; McWilliams and Peters 2012; Bajada and Trayler 2013). Therefore, we decided to develop or collate the necessary resources and provide them via the university’s learning management system (Canvas).

We divided the course into three units with each unit taking approximately five weeks.

• In the first unit, the focus would be core programming and data wrangling skills needed to create financial models. The analytics faculty member led the development of this unit.

• In the second unit, the focus would be financial models and case studies. Specifically, this unit would reinforce financial models that students had learned in previous courses while showing them how to create and interpret these models using Python. We were equally responsible for developing this unit.

• In the third and final unit, the focus was simulation requiring more complex programming and integration of concepts. The finance faculty member led the development of this unit.

For more information about the week-by-week plan for this course, refer to Appendix: Course Schedule and Content.

Following best practices, our next step was developing our delivery dynamic as co-instructors, including our turn-taking approach and teaching motifs (Wenger and Hornyak 1999; Shibley 2006). Based on the content we planned to deliver, we decided to have one instructor designated as the primary instructor for each class session (the sequential motif). The goal of using this approach was to present information as efficiently and effectively as possible (Wenger and Hornyak 1999). Although this motif would continue during lab sessions, we agreed that each of us would clarify differences and identify nuances in our disciplines when applying concepts and techniques to labs, an approach known as the distinctions motif (Wenger and Hornyak 1999). Our goal in using this motif was to model professional, interdisciplinary interactions for students (De Veaux et al. 2017).

Based on this dynamic, for each of the content areas identified, brief lectures would be used to review knowledge from previous courses and introduce new concepts. Following these lectures, students would be provided short labs using Excel or Python (with Jupyter notebooks). We would circulate throughout the computer lab to answer questions, provide clarification, and generally oversee student work. At the completion of each topic, students would complete a case that integrated their knowledge.

For example, Weeks 6 and 7 focused on forecasting. As part of this topic, students would review concepts they learned in the prerequisite courses, such as descriptive statistics, simple regression, accounting statements, and attributes of stock returns. Then, students would learn additional knowledge and skills through a variety of activities, such as daily discussions about newly released economic data, the impact on stock and bond prices, and the forecasting techniques that could be applied as businesses attempt to make decisions using this new information.

In Week 6, the finance faculty member would review the financial concepts and calculations involved in forecasting. The demonstration would use data sources that were covered in the first unit (for data wrangling) to provide connection between the first and second units in the course. This week would begin with a discussion of time series data, including examples of this type of data and how these data can impact business operations. Next, students would work to understand trends, smoothing, and regression within forecasting by completing examples in Excel and using publicly available datasets from the Federal Reserve Economic Data (FRED) source. They would use a variety of forecasting functions within Excel such as LINEAR, SEASONALITY, and CONFINT. Finally, the finance faculty member would discuss the importance of forecasting and its appropriate interpretation to support business decision making. At the completion of this week, students would use FRED data to create all the types of models covered during the week, from naïve forecasting to multiple regression. Throughout the week, materials such as Microsoft PowerPoint presentations, Microsoft Excel spreadsheets, and comma-separated value (CSV) data files would be provided to students.

In Week 7, the analytics faculty member would introduce working with time series data within the Python programming language. With this foundation, the faculty member then would demonstrate the creation of appropriate forecasting models using Python, including naïve, moving average (MA), weighted MA, exponential MA, and regression-based forecasts. While the forecasting models from Week 6 would be used, the datasets would be different to illustrate how differences in data can affect models. At the end of the week, students would complete a case that required them to integrate their learning from Weeks 6 and 7 by creating and interpreting forecasting models using Python. Throughout the week, materials such as Microsoft PowerPoint presentations, Jupyter Notebook Python code samples, and CSV data files would be provided to students. For more information and for examples of the materials for Weeks 6 and 7, refer to the GitHub repository (https://github.com/mccart82/Financial_Analytics) associated to this work.

As weekly materials were developed, we intentionally chose datasets from publicly available sources, problems related to current events, and cases mirroring real-world situations to support students in connecting what they were learning to their lived experiences (De Veaux et al. 2017; American Statistical Association Undergraduate Guidelines Workgroup 2014; Charles A. Dana Center at The University of Texas at Austin 2021). In addition, this approach supported ongoing discussions about the evaluation of data sources (veracity, bias, etc.); the need to clearly define problems and ask questions; and the appropriate interpretation and decision making related to their analysis outcomes. As instructors, we could model the storytelling principles of show, explain, and narrate and provide real-time feedback to students as they worked on the communication skills necessary to achieve the final learning outcome of the course: “Professionally communicate the process and results of financial modeling to a variety of stakeholders” (Donoghue, Voytek, and Ellis 2021).

As the final project for this course, students would be required to complete a comprehensive financial analytics case involving stock valuation analysis for a theoretical firm using the corporate valuation model and free cash flows. Students would be provided with key financial accounting and background information for the firm, as well as other variable factors such as the standard deviation of loan rates, revenue growth rates, and the cost of goods sold as a percentage of revenues. Using this information and the course materials, students would answer some conceptual questions related to financial modeling, use of software, and application of simulation techniques to demonstrate mastery of foundational knowledge within the course. Students would then complete four analyses (two in Excel and two in Python) related to tax rates and stock valuation. As the final portion of the case, students would interpret and summarize their analysis results (taking into consideration all the analysis performed), and then provide recommendations based on this summary. We would grade these comprehensive cases individually and then discuss our findings to assign a final grade for each student.

4 Evolving Course Delivery

The course began as scheduled in the spring semester of the 2019–2020 academic year with 18 students enrolled. Of these students, seven were majoring in actuarial sciences, nine in finance, and two in business administration.

During the first three weeks of the course, we quickly determined that students required more direction and structure for learning how to “study” for the course based on the results of initial assessments and questions being asked during in-class labs. Therefore, for each of the labs in the course, a set of study questions was developed to help direct students during their work. Answers to these questions could then be covered as students completed their labs or in the class sessions following. With the introduction of these study questions, students could learn to troubleshoot any problems in real-time with faculty support, allowing students who were struggling to be quickly identified and supported. As a result, students’ confidence in their skills and knowledge grew.

As student confidence grew, they requested additional labs to enhance their learning. In response, we created a series of “challenges” that were either ungraded or provided extra credit for students interested in further practice. These challenges motivated students to do their own research to create analytics models, which provided valuable experience assessing knowledge sources and websites related to analytics and finance. In addition, while these challenges were not required for students, students who struggled with the course content expressed frustration at not having time to complete the challenges. This frustration motivated them to meet with us for additional help and support within the course so that they could work on the challenges as their skills improved.

As students entered the second unit of the course, review of the assessments also indicated that several students had not completely retained the necessary foundational knowledge in Python to work effectively with the financial models. Therefore, the Jupyter Notebooks provided were updated to include more explanation of the code and additional hands-on practice exercises to assist students in mastering these foundational skills.

Spring break occurred at the end of the seventh week of this course. During that break, the university determined that classes would need to move fully online due to the COVID-19 pandemic. One week was spent transitioning content online, meaning one less week on simulation within the course. After this transition week, the entire course was conducted online with recorded lectures and tutorials; study questions that were turned in three times a week to track student progress; and online office hours to answer questions and provide clarifications as needed.

In addition, we continued our collaboration with IT staff. They developed and provided a virtual computer lab environment that students could access using their personal devices. This virtual computer lab provided all the necessary software for the course with a consistent interface, which allowed the instructors to have the same setup as students. Because of the coursework involved, students were encouraged to use laptops or desktop computers to access the virtual computer lab. For students without access to this type of equipment, arrangements were made on a case-by-case basis with the university’s IT department.

5 Assessing Student Outcomes and Feedback

By the end of the course, 16 students remained (with two dropping the course due to challenges with the course content and other conflicts related to the pandemic).

To assess student learning, faculty members first reviewed the results from the formative assessments, which included an option for students to resubmit their assessment after receiving initial grading feedback. Because many students chose to incorporate feedback and resubmit, the faculty members used the last submissions of each assessment to determine student achievement of learning outcomes. Based on this review of formative assessments:

• 16 out of 16 students (100.0%) successfully achieved the “Identify and describe business decision making scenarios appropriate for financial modeling” learning outcome.

• 14 out of 16 students (87.5%) successfully achieved the “Compare and contrast common financial models such as forecasting, capital asset pricing, credit risk analysis, portfolio analysis, and Monte Carlo simulation” learning outcome.

• 14 out of 16 students (87.5%) successfully achieved the “Demonstrate data wrangling techniques required to prepare data for use in financial modeling” learning outcome.

• 14 out of 16 students (87.5%) successfully achieved the “Create and interpret appropriate financial models” learning outcome.

• 12 out of 16 students (75.0%) successfully achieved the “Professionally communicate the process and results of financial modeling to a variety of stakeholders” learning outcome.

Next, the faculty members reviewed student achievement based on the final project for the course (which was designed to assess all the learning outcomes). This project was structured as a summative assessment without the option of resubmitting the project. Based on these projects, 13 out of 16 students (81.3%) successfully demonstrated their achievement of all learning outcomes for the course.

Two out of 16 students (12.5%) struggled with the learning outcome: “Create and interpret appropriate financial models.” For this learning outcome, students were required to create multiple financial models in Excel and Python. In general, students who failed to achieve this outcome missed a significant step in the creation of a model or used incorrect factors when creating a model.

Three out of 16 students (18.7%) struggled with the learning outcome: “Professionally communicate the process and results of financial modeling to a variety of stakeholders.” For this learning outcome, students were required to explain the steps in their analysis, state their conclusions, and provide recommendations based on the business decision making scenario provided. In general, students who failed to achieve this outcome provided minimal explanation of their steps; stated incorrect or illogical conclusions based on their modeling and analysis; or provided recommendations not supported by their modeling and analysis.

While these findings are consistent with faculty expectations given previous experiences teaching statistics, programming, and finance courses, the faculty members are considering course modifications that could further support student learning related to these two learning outcomes, such as recording videos as references for each financial model to assist students with completing the steps involved and interpreting the models correctly.

In the end-of-course evaluations, students felt the course work was challenging and addressed the course outcomes. They appreciated having two instructors, which supports continued use of this simultaneous team-teaching approach for future offerings of this course. From a technology perspective, students appreciated the use of Jupyter Notebooks for learning Python and having recorded lectures and tutorials (once the course moved fully online) for reference.

Based on this feedback, the simultaneous team-teaching approach will be maintained for future offerings of this course. In addition, Jupyter Notebooks will continue to be used, and portions of the course may move online to allow more time for working through labs and cases during in-person class meetings.

6 Faculty Lessons Learned

During the first iteration of this integrated course in finance and analytics, we learned several key lessons that can be applied in future iterations.

First, we strongly recommend ongoing joint planning, as this was a critical factor in the success of this course. Commit to meeting regularly to adjust the course content and schedule based on student needs. Having this regular time commitment means that issues can be addressed quickly while ensuring the faculty members remain aligned with each other and when working with students. For example, these meetings proved to be crucial when the course needed to transition quickly to a fully online format.

As part of this planning process, we encourage faculty to create a course outline document that describes the plans for each course week including learning objectives, materials, and faculty member responsible. During joint planning meetings, these plans can be referenced and adjusted quickly. In addition, the outline documents which tasks are required each week, reducing the confusion that could naturally occur when multiple faculty members are involved in running the same course. In our case, when the course needed to change in response to the COVID-19 pandemic, we could quickly assess the tasks remaining, determine which needed to be modified, and assign an appropriate faculty member to complete each task.

For faculty attempting this approach, building a level of trust, empathy, and humility is critical. These characteristics allow open discussion of what works well within the course and what needs to be improved. Having honest conversations evaluating the course progress allow faculty to share and integrate their skills, which creates a more seamless integration for students. In addition, when course changes are needed, faculty can quickly assess their skills and create a transition plan that aligns tasks with the faculty member best able to complete each task. For example, when the course moved online during the COVID-19 pandemic, one faculty member was skilled at creating and maintaining course sites using Canvas, so this faculty member took the lead in implementing the necessary changes within Canvas.

This level of trust, empathy, and humility should also extend to the students taking the course. Early in the course, we recommend sharing with students that the course is new and will likely require revision. Encourage students to share their questions and provide feedback throughout the course. As students provide feedback, discuss potential changes with students and make appropriate revisions. As this process continues, students begin to trust that their concerns and ideas are valued. For us, having this trust was critical during early course modifications to improve student learning and engagement and to create a safe and inclusive environment for all students.

Additionally, carefully consider the technology (software, devices, etc.) selected to support course delivery and involve IT staff early in the planning process. For example, what devices are students likely to be using and will they work well with the selected technology? Can students access the technology on- and off-campus? What costs may be involved to obtain the necessary technology? How much work (faculty, student, IT, etc.) will be required to maintain whatever technology is selected? Carefully considering these factors and collaborating with IT staff can support equitable access for all students.

Finally, faculty should be prepared for significant overall workload. For this first iteration of the course, the university assessed each faculty member a half course load (for a total resource of one faculty member for the course). Due to the integrated nature of the course delivery, though, the time commitment for each faculty member was that of a full course. While this faculty commitment was necessary for the success of the course, faculty considering this approach should be sensitive to the compensation structure of their institution and the support available for this type of simultaneous team-teaching approach.

7 An Integrated Future: Course Maturation and Curricular Considerations

Overall, the first iteration of this course was a significant learning experience for everyone involved.

From a scheduling perspective, we plan to make this a required course within the finance curriculum and schedule it to provide students the maximum opportunity to complete internships during their course of study.

From a technology perspective, we will continue to work with the university’s IT staff to ensure equitable access for all students. In addition, we will continue to modify the materials based on technology and software updates.

From a faculty workload perspective, any faculty member attempting to teach an integrated course needs to be aware of the time commitment involved and be willing to participate in honest evaluations throughout the course. As a result, future iterations of this course will need to ensure appropriate compensation for the time commitment.

From an assessment perspective, future iterations of this course will include structured learning and study activities from the beginning of the course, especially related to Python programming. Study questions and challenges will continue to be integrated into the assessment plan on an ongoing basis in addition to the homework cases and final comprehensivecase.

In addition, Jupyter Notebooks will continue to be used as students appreciated the in-depth explanations provided, and they often used these for reference when working on homework and cases for the course. These notebooks will continue to include significant explanation of concepts and code in addition to the actual Python code for each model/technique.

To address student achievement of learning outcomes, we plan to add videos that walk through the Jupyter notebooks to support learning the appropriate steps in creating and interpreting models. In addition, portions of the course may move online to allow more time during in-person class meetings to focus on activities that allow students to practice communicating more effectively about their models, interpretations, and recommendations.

Although individual work will likely continue to be the focus to allow us to continually assess student learning, future offerings of this course may incorporate group projects, especially for the final project within the course. We would prefer students have opportunities to work with industry partners for their final projects, and the typical size of these projects would likely require a group approach. In addition, working in a group does provide valuable experience dividing analytics work, which is common within business operations (National Academies of Sciences, Engineering, and Medicine 2018; Gundlach and Ward 2021). Therefore, as the course is reviewed and redesigned for future offerings, we will consider how projects with industry partners could be incorporated into the course.

Improvements to the structured learning activities and Jupyter notebooks as well as the addition of prerecorded videos will also create a scaleable and transferable course for increasing student enrollments and changing of assigned faculty. During this first iteration, a significant portion of faculty time was spent re-explaining concepts and techniques covered in lectures or during labs. Developing ongoing support resources that students can use to review concepts (such as the notebooks and prerecorded videos) should reduce this workload. By having structured learning activities earlier in the course, faculty can more quickly assess and address gaps in individual student learning. Having a set of planned materials, rubrics, and assessments allows teaching assistants to participate in course delivery, where appropriate, to reduce faculty workload.

Finally, although the focus of this case study was the integration of finance and analytics, similar integrations are needed within other business disciplines. Until higher education can produce sufficient analytics-enabled faculty, this integrated simultaneous team-teaching structure may be the most effective approach to ensure graduates have the skills needed to meet the demands of industry.

8 Conclusion

With increasing demand for analytics-enabled professionals and the need identified by AACSB to span the traditional boundaries that exist between business fields, business schools within higher education have an imperative to be intentional about covering data science and analytics skills in the context of specific domains such as finance. Yet most faculty have expertise within a single domain, making spanning historical silos challenging, especially in the context of emerging fields such as data science and analytics. This case study describes a course designed specifically to address this challenge and fulfill the need to prepare analytics-enabled professionals.

In designing and delivering this course, we discovered that simultaneous team teaching provided an important model for students of the multidisciplinary nature of data science and analytics, as well as the critical need for communication skills. Students had the opportunity to learn computational skills using real-world data that required significant data wrangling and troubleshooting. A comprehensive final project required students to interpret their analytical models and their outcomes in the context of a business decision-making scenario.

As next steps, we would recommend including this course as a required component of the undergraduate finance program with possible expansion to allow deeper learning of the content and incorporation of a team-based experiential learning project with corporate partners. In addition, to ensure all undergraduate business students have a basic understanding of data science and analytics, similar integrated courses should be developed within other programs and domains. Finally, faculty members involved in this development should be adequately compensated for the workload involved.

The demand for professionals with data science and analytics skills will continue to grow, requiring higher education to expand their coverage of data cleaning and wrangling, data visualization, computation techniques and tools, statistical and machine learning methods, and inquiry and communication skills within their curricula. As educators, we must continue to find innovative and engaging ways to integrate these skills so that students are prepared for their future careers.

Acknowledgments

The authors express their appreciation to Dr. Sarah Esveldt, Dr. Muztoba Kahn, Dr. Thomas St. George, and Dr. John Symms for their review of this article. The authors express their appreciation to the Carroll University Office of Information Technology (OIT) staff, especially Boyang (Romeo) Yang, for their support of this course. The authors express their appreciation to the JSDSE Editors and anonymous referees for their review and feedback.

Data Availability Statement

Information and example materials (data, spreadsheets, code, etc.) for Weeks 6 and 7 of this course are publicly available within a GitHub repository (https://github.com/mccart82/Financial_Analytics) associated to this work.

Disclosure Statement

The authors report there are no competing interests to declare.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Appendix:

Course Schedule and Content

Table

Week(s)	Content	Responsible Faculty
1	Python basics (installation, use of Jupyter notebooks, mathematical operators, logic operators)	Analytics faculty member
2	Python data and control structures (tuples, lists, dicts, arrays, vectors, matrices, DataFrames, if/elif/else, for while)	Analytics faculty member
3	Data sources (common sources of financial data, basic data loading and cleaning using Excel and Python, Python packages for financial data)	Finance faculty member and analytics faculty member
4-5	Data wrangling (sources of financial data; file types; reading and loading data files; reviewing, cleaning, filtering, grouping, and aggregating data)	Analytics faculty member
6–7	Forecasting (time series analysis, financial decisions involving forecasting, sources of time series data, naïve, moving average, weighted average, exponential smoothing, regression)	Finance faculty member and analytics faculty member
8	Capital asset pricing model (financial decisions involving CAPM, data sources, challenges with CAPM, interpretation and evaluation of CAPM)	Finance faculty member and analytics faculty member
9	Ratio and debt analysis (financial decisions using ratio and debt analysis, strengths and weaknesses of analysis, data sources, trend analysis and industry comparison, Altman’s z-score)	Finance faculty member and analytics faculty member
10–14	Simulation modeling (data tables, randomization, probability distributions, Monte Carlo simulation, sensitivity analysis)	Finance faculty member and analytics faculty member
15	Final exam (final project for the course)	Finance faculty member and analytics faculty member