Treffer: Data Visualization in Introductory Business Statistics to Strengthen Students' Practical Skills

The objective of this study is to present and discuss how data visualization can be incorporated into teaching approaches by business faculty in introductory business statistics to strengthen business students' practical skills. Data visualization lessens difficulties in learning statistics by providing opportunities to illustrate analytical findings in graphic form, which is essential for learners with different learning styles. Familiarizing students with Excel, Python, or other software in introductory business statistics is beneficial in helping them attain statistical literacy by analyzing real-world data such as COVID-19 statistics. Using such data equips students with knowledge of statistical implementation--a core skill in the business world.

As Provided

AN0154795839;d8y01mar.22;2022Jan24.04:16;v2.2.500

Data visualization in introductory business statistics to strengthen students' practical skills 

The objective of this study is to present and discuss how data visualization can be incorporated into teaching approaches by business faculty in introductory business statistics to strengthen business students' practical skills. Data visualization lessens difficulties in learning statistics by providing opportunities to illustrate analytical findings in graphic form, which is essential for learners with different learning styles. Familiarizing students with Excel, Python, or other software in introductory business statistics is beneficial in helping them attain statistical literacy by analyzing real‐world data such as COVID‐19 statistics. Using such data equips students with knowledge of statistical implementation—a core skill in the business world.

Keywords: teaching; data visualization; introductory business statistics; teaching statistics

INTRODUCTION

In business statistics, the gap between theory and practical application has appeared difficult to bridge; however, transforming data into information through visual presentation and analysis may present solutions. Business statistics is sometimes described as the art of measurement in decision making [7]; however, in the classroom, business educators often find it difficult to make students realize the practical role of statistics in implementing decisions. Gould [12] emphasizes the importance of using real data in introductory statistics to help students learn the applicability of statistics to real‐world problems [12]. For example, the relevance of COVID‐19 data to students' lives is clear, such as its effects on classroom vs online learning. Challenges in teaching statistics include equipping students with sufficient proficiency to understand and draw accurate meaning from analytical results in solving practical issues. Ograjenšek and Gal [22] state that qualitative methods driven by real‐world situations should be implemented in a statistics education curriculum or program to achieve educational goals from a pragmatic perspective [22].

Since students encounter constantly updated data on the internet, Gould [12,p. 309] recommends that students know how to recognize data and connect its usefulness to their work. Over the past two decades, the increasingly open nature of software development has made tools such as Gapminder available to students, who can use them to apply theories learned in the classroom to real‐world data. In statistics education, enrichment resides in content (more data analysis, less probability), pedagogy (fewer lectures, more active learning), and technology (for data analysis and simulations) [12]. However, in business statistics, there is still the belief that memorizing theorems and formulae appear to be necessary for students to acquire standard statistical knowledge. Although statistical training provides a solid foundation for business analytics, in teaching introductory business statistics, the complexity of cases and examples is reduced for students in the belief that they need to develop an abstract and basic view of primary principles. For students who are interested in analytics, introductory business statistics is still a major structured course in too many business faculties, in which the onus consists of learning and applying theorems and formulae in applications.

Typically, simplified datasets of 20 to 50 data points are included in examples or exercises in textbooks of introductory business statistics. The gap in scales of datasets and complexities between typical examples used in introductory business statistics and real data in the contemporary business environment is enormous. By using real‐world data, students can be connected to statistics about global social and economic development [17]. Moreover, considering the growing importance of big data, big data analysis and concepts should be introduced early to students in higher education [1].

To successfully connect students in the 21st century to the real world requires teaching strategies that are in line with technological development. Casas‐Rosal et al. [4] mention that there is a need to apply technical and strategic knowledge to increase students' motivation for statistics by working with real tasks and data [4]. Real‐world data can be used as an instrument for public policy development and for gaining insight into and proposing solutions to societal problems [15]. In particular, the advent of big data provides educators with the opportunity to demonstrate or teach students how statistical concepts apply to their everyday lives, and a growing body of research has begun paying attention to this field [1]. Examples, therefore, should be presented in the context of real‐world problems [18].

Szobonya et al. [24] mention that students today basically "grew up with electronic devices in their hands" and learn statistics after becoming familiar with such devices, which means that using computers is foundational for them [24,p. 23]. For statistics educators, the skill base of teachers and students needs to be extended to include an understanding of the analytic techniques suited to accessing, storing, and analyzing high‐volume unstructured data [23]. Gapminder was a pioneer in using real‐world data in teaching statistics, with highly innovative visualization that is still relevant today (https://www.gapminder.org/data/).

Using statistical software in introductory business statistics can improve students' knowledge about statistics and its usefulness in real life. Emphasizing real‐world applications with computers has become more prevalent in introductory business statistics at universities [3].

In application, business analytics relies on structured databanks. Whether students are capable of applying statistical principles and results at work depends on personal attainment of statistics and job requirements. At university, statistics is taught across several disciplines and departments [8]. However, for a large proportion of students who take only one course in statistics, introductory business statistics is the preferred option.

The objective of this study is to present and discuss how data visualization can be incorporated into teaching materials in introductory business statistics to strengthen students' practical skills, in particular regarding conceptualizing formulae with data applications. Examples are provided, and student feedback is listed to reveal the learning process with enhanced data visualization using real‐world data in introductory business statistics.

Data visualization

The National Forum on Education Statistics [19] provides a clear definition of data visualization: the transformation of data into information through visual presentation and analysis [19,p.1]. Computer‐assisted instruction, as well as the preliminary concepts of data visualization, have been used in statistical education for decades. A large number of instructors use some type of technology in courses, such as requiring students to learn a spreadsheet, a statistical software package, and web resources such as datasets and applets. With these, students are able to operate and see the results and figures of a complex dataset themselves, which helps them more easily understand statistics [6,10].

Hudingburgh and Garbinsky [11] and Nolan and Peret [21] have indicated the implementation of diverse approaches to incorporating data visualization in introductory courses that could be useful to statistics instructors and that students would find interesting and appealing in terms of expressing their statistical communication skills. They also provide advice on issues such as assessment and facilitating group work. As the need for data visualization continues to grow, advances in technology and software make tools and procedures easier to apply and more accessible; hence, using and interpreting analytical results with solid fundamental concepts is vital [2, 8]. Although teaching students advanced analytical methodology in introductory business statistics may exceed both the objective and focus of the course, data visualization is one area where students can quickly gain knowledge if applied.

Facts are embedded in data. Text, tables, and graphs are effective communication media that present and convey information. They aid readers in understanding the content of research, sustain their interest, and effectively present large quantities of complex information [13, 14]. When data are presented graphically rather than as a set of two‐dimensional measures in tables, attractiveness is enhanced. Proper data visualization facilitates the recognition of patterns and relationships to communicate a message in a compelling and interesting manner. Good data visualization helps the audience understand the quantitative relationships more explicitly [3].

Applying data visualization in an introductory business statistics course

In this introductory business statistics course, target students were undergraduates majoring in business studies such as in Business Administration, Marketing, Accounting, Finance, Management Information Systems, and Applied Economics. It is a required core course and is the first statistical training before students take advanced analytical courses. The core learning outcomes are independent analytical skills and the applicability of statistical knowledge.

To connect introductory business statistics to the world and to expand students' international perspectives, students were informed that assignments were accompanied by real‐world data related to global issues. Students were required to produce data visualization for multiple cases and were given the opportunity to select a case for interpretation. In the class, databases such as "COVID‐19" and the "Dow Jones Transportation Index" were selected. Students were assessed through attendance and in‐class assignments that had to be completed before the end of the class.

The course was designed in such a manner that visualization through basic or entry‐level Python programming was offered as supplementary assignments, similar to Hudiburgh and Garbinsky [9] and Nolan and Peret [21], alongside unabridged, regular introductory business statistics modules. For facilitating this structure, the class of 92 students was split into two and rotated between the regular classroom and the statistics laboratory, where each student had access to a computer. Moreover, the data visualization supplementary assignments that involved Python programming were arranged into three stages: namely, introducing the basic concepts of coding; then practicing using entry‐level coding; and finally, students writing their own Python code for data visualization. Last, students were given the option of participating in this supplementary instruction, although only a few opted out of it.

Microsoft Excel provides a convenient visual tool for business statistics that is especially useful at the introductory level and that focuses on descriptive statistics and graphs [16]. Although often overused and over relied on at the expense of other tools, Excel nonetheless represents a highly useful visual tool that is easily accessible to students at the introductory level. Excel was used to create graphs for Figures 1–2 in this study. In this introductory business statistics class, students learned to use real‐world data step by step. In contrast to the typical simplified examples provided in business textbooks, real‐world data are usually larger and more complicated. Students were assigned tasks so they could get used to working on these datasets. Before making use of a given database, students started by identifying and downloading required datasets. In one assignment, for example, data of COVID‐19 deaths in the United States from Centers for Disease Control and Prevention (CDC) was utilized.

Figure 1 shows weekly data of COVID‐19 deaths in the US from January 2020 to October 2021 with a forecast using exponential smoothing and three‐point moving averages. Students were provided with very basic training in diagnostics so that they understood how to distinguish the differences between two highly popular forecasting techniques in application. Using real data with illustrations to enhance visualization of the situation helps students understand what business statistics could do in a presentation.

The pandemic affects people's mobility, which is reflected in Dow Jones Transportation Index. Figure 2 shows the daily close of the Index from January 2020 to Early November 2021, with the lowest level in April 2020 when confirmed cases saw the first upsurge. Figures 3 to 6 are box plots and QQ plots of Dow Jones Transportation trading volume and daily close in 2020 and 2021, respectively, to illustrate whether the data trends are away from normality. With fluctuations of the Index during this period, the data is shifted away from a normal distribution with few outliners making the distribution skewed on one end, which could be counterintuitive to students that the assumption of normal distribution could be violated even with large datasets. This case was especially useful in demonstrating how real‐world data often do not satisfy theoretical or intuitive expectations.

In other assignments, students were required to work on more complicated problems with more advanced settings and computations. For instance, they may be asked to practice with various types of charts and pivot tables before being required to analyze and describe the connotation of the charts of their choice. Some introductory Python programming was also included in this course. Python was originally developed around 1990 and is characterized by a high‐level language, so the grammar is simple and clear to students. The most useful characteristic of Python is its wide array of large free libraries, which makes programming easy and enjoyable. The library used in this study is Seaborn version 0.9.0 (Michael Waskom, http://seaborn.pydata.org/) library, which is a Python data visualization library based on matplotlib. It provides a high‐level interface for drawing attractive and informative statistical graphics. Moreover, there was no need for any prerequisite courses in elementary programming techniques, and only the most basic or even zero familiarity with the Python program was necessary since only basic or entry‐level Python programs were used. However, students were provided with the option of extra teacher assistant (TA) hours in case they were overwhelmed or required further assistance, which only a few students took advantage of. In general, most students found the class interesting and demonstrated few signs of struggling with the training or the concepts taught.

Figure 7 illustrates an example used in class to display the scatter diagram of "life expectancy in five continents" and its Python code. With this friendly library for data visualization, charts or diagrams can be made conveniently within only a few lines of coding.

Student outcomes and feedback

Most students were grateful for the practice introduced in this course. Student comments and feedback from end‐of‐semester course evaluation primarily referred to the benefits of the training. In addition, students were provided the opportunity to self‐evaluate their progress in business statistics. The results of the self‐evaluation are illustrated in Table 1.

1 TABLESelf‐evaluated skill improvement based on statistical courses

<table><thead valign="bottom"><tr><th align="left" /><th align="left">Logical thinking</th><th align="left">Problem solving</th><th align="left">Analytical skills</th><th align="left">Teamwork</th></tr></thead><tbody valign="top"><tr><td>Statistics I and II</td><td>7.40</td><td>7.35</td><td>7.18</td><td>7.68</td></tr><tr><td>Regression analysis</td><td>7.44</td><td>7.39</td><td>7.33</td><td>7.72</td></tr><tr><td>Business forecasting</td><td>7.75</td><td>7.56</td><td>7.63</td><td>7.81</td></tr></tbody></table>

1 Note: A 1 to 10 Likert‐type scale was used, with 1 representing the lowest rating, and 10 the highest.

The results depicted in Table 1 reveal that the students rated themselves on average a score of 7.4, 7.35, 7.18, and 7.68 in the skill categories of logical thinking, problem‐solving, analytical skills, and teamwork for this introductory course in statistics (Statistics I and II), respectively. From these results, it can be seen that students did experience improvement in the listed skill categories, although this improvement was slightly lower than recorded in Regression Analysis and Business Forecasting. We cannot tell for sure what effect that data visualization had on this self‐rating, and classes and studies that focus more heavily or exclusively on data visualization are needed in the future.

Moreover, the feedback from students regarding the course appeared to emphasize the practical or pragmatic framing of the classes. Students appreciated very much the ability to link abstract statistical concepts with practical or everyday situations and experiences. The following represents the feedback from students.

On the other hand, some difficulties and challenges that students usually face in introductory business statistics remained. Some of the main issues commonly mentioned in students' feedback primarily related to the formulae and statistical concepts:

As a solution to these problems, we recommend in the future the promotion of fully‐fledged visualization courses that centralize visualization for pragmatic and real‐life statistical application. For example, Nestorov and Sippo [20] designed a data visualization course, which featured a real‐world project component for an undergraduate course. Such a course has the ability to demonstrate to students the practical applications of the statistics that they learn in class, which can often appear as too abstract and useless to students. Moreover, in the future, there should be more emphasis on big data, as this is an issue that students face in their everyday experiences, such as their use of social media and interactions with their mobile devices. This inclusion would make the class more accessible to students. Last, we plan in the future to include a more rigorous and direct assessment of students' attitudes and competencies in data visualization so that they can be presented as outcomes in future studies.

CONCLUSIONS

Data visualization lessens difficulties in learning statistics by providing opportunities to illustrate analytical findings in graphic form. Real‐world data are more complicated than the simplified examples provided in textbooks of introductory business statistics. The rationale of using real‐world data is not only to strengthen students' practical skills in statistics by using computer‐assisted instructions but also to connect students to reality.

For facilitating students' conceptualization of the fundamental principles of introductory business statistics, curricula generally include simplified cases and examples. For reducing the gap between introductory business statistics and pragmatic applications in the workplace, Excel, Python, or other software can be introduced to analyze real‐world data to attain statistical literacy. Suggestions for instructors teaching introductory business statistics are as follows:

Introductory business statistics is the first statistics course that students encounter where they can rationalize the focus on business analytical skills. The data visualization presented in this work can be generalized to more complex models, especially in advanced analytical courses. Once students become accustomed to working with real‐world data, the training provides extensive benefits in terms of developing business analytical skills. The novel aspects of the present study are aimed at business faculty and include our presentation of pedagogical approaches using open‐source software (ie, Python) and data pertaining to a contemporary issue (ie, COVID‐19) in an effort to illustrate the applicability of data visualization in introductory business statistics.

CONFLICT OF INTEREST

This study strictly follows the guidance, and there is no conflict of interest.

By Jane Lu Hsu; Abram Jones; Jia‐Huei Lin and You‐Ren Chen

Reported by Author; Author; Author; Author