. 2IntroductionThe demand for energy consumption in the world is growing at an annual rate of close to 2% peryear ( Welker, A., et al., 2022) ) and that translates to about 3,598 Twh in 2022. In the UnitedStates (US), the energy consumption growth rate is 2.6% which translates to about 106 Gwh in2022 (EIA, US Electricity Overview, 2023). The US energy generation sources in 2022 (USPrimary Energy Source, 2022) are shown in Figure 1. Figure 1 US Energy Source Distribution (Source EIA)Fossil fuels (Petroleum, Natural Gas, and coal) make up 78.8% of the total energy source in theUS. Petroleum is the largest source and solar is the lowest. Renewable sources add up to 13.1%(Solar at 1.9%). It has been a fact that these non-renewable sources
we complete our study, we believe our findings will sketch the early stages of thisemerging paradigm shift in the assessment of undergraduate engineering education, offering anovel perspective on the discourse surrounding evaluation strategies in the field. These insightsare vital for stakeholders such as policymakers, educational leaders, and instructors, as they havesignificant ramifications for policy development, curriculum planning, and the broader dialogueon integrating GAI into educational evaluation.1. IntroductionThe advent of generative artificial intelligence (GAI) has heralded a new era in higher education,prompting extensive research and discussions, particularly concerning its impact on traditionalassessment practices. Recent
industry,it is essential to explore their views on data science. This includes understanding its significanceand practicality in relation to their field of study, alongside gauging their willingness to integratedata science principles into their academic and future professional pursuits. Understandingstudents' attitudes towards data science is essential for several reasons: 1. Curriculum Development: Insight into students' perceptions allows educators and curriculum developers to tailor educational programs that not only meet the technical demands of the industry but also align with students' interests and aspirations. This ensures the development of engaging and relevant content that enhances learning outcomes
pattern calculated at 35 comparedto 25 for the revised pattern, as determined by Equation 1. This decrease in complexity correlateswith an improvement in graduation rates. Figures 8a and 8b provide MDP representations of theengineering curricular patterns from Figures 7b and 7c, denoted as M1 and M2 , respectively. Thepolicies π M1 (a/s) and π M2 (a/s), integral to these representations, are also included in Figure 8.Generally, these policies are inherently shaped by the curricular structure itself. For instance, inFigure 7b, a student completing the P recalc course is typically limited to enrolling in Calculus I,barring any dropout scenarios. Therefore, for the P recalc state in Figure 8a, the only viable actionis enrolling in Calc I (i.e., π
know, un-derstand and be able to demonstrate at the end of some learning experience. For instance, ABETstipulates a minimal set of student learning outcomes that describe what learners should knowand be able to by the time they graduate from an ABET-accredited engineering program.1 It isalso now common practice to articulate course-level learning outcomes for each of the coursesoffered by a college or university; these indicate what a learner is expected to know and be ableto do after successfully completing a course. A common approach used by curriculum design-ers, known as backwards design, involves designing a curriculum from the bottom up by startingfrom the program learning outcomes, and then creating course-level objectives that would
sentiment analysis Its value comes fromanalyzing large amounts of text data [2]. For example, its applications have been used to analyzesocial media posts to track public opinion and identify trends (e.g., O’Connor [8]). In the field ofeducation, it has been applied to the analysis of student essays to provide feedback, teamworkreview analysis, and students’ feedback loop [1], [3], [9]. Another application is in the generationof natural language text (e.g., machine translation systems use NLP to translate text from onelanguage to another) [10]. In addition, it has been used to generate feedback on student writing [11] and to createpersonalized study materials [12]. It also can facilitate more personalized and effectiveinstruction [13]. By
-triggering mechanism. This method involves classifying student sentences into pre-definedclasses [43, 44] that reflect specific dimensions of the engineering design generic problem-solving process [17, 18]. An example of the dimensions included in the testing done for theProblem Definition stage, presented in this contribution, is included in Figure 1.Figure 1 illustrates the models and dimensions adopted for the problem-definition stage of ageneric problem-solving process.The fine-tuning process is based on the “further in-domain pre-training” strategy described bySun, et al. [45], using sentences from 32 reports taken randomly from a pool of approximately140 reports each term. The selected reports are from two terms equivalent to two
Thermodynamics QuestionsIntroductionThis paper describes the results of a study where generative Artificial Intelligence (AI) was usedto analyze short-answer explanations to two conceptually challenging chemical engineeringthermodynamics problems. This work comes from a collaboration between machine learning andengineering education researchers utilizing machine learning to analyze student narratives ofunderstanding in short-answer explanations to conceptually challenging questions [1], [2].Concept questions, sometimes called ConcepTests [3], are multiple-choice questions involvingminimal calculations and give students experience applying conceptual knowledge. Whenutilized within active learning pedagogies, concept questions have been shown to
user experience survey. The survey results gave some constructivefeedback for the developers. Overall, the project can deliver a feasible solution for courseinstructors to handle many student project teams. In the future, a generative AI feature -CHATME will also be available on the front end to help the user check the status of each studentgroup, which is built using NLTK and TensorFlow. Moreover, if a team issue arises, theplatform will alert the users, and provide constructive suggestions on how to improve the groupperformance.IntroductionIn engineering education, fostering collaborative skills [1] among students is crucial, and team-based learning has become the primary approach. It is an approach particularly prevalent infoundational
engineering course. We characterized ChatGPT usage as either productive or unproductivefor learning and defined four general reasons why students engaged with AI in this course:ChatGPT as 1) A learning aid 2) A coding resource specifically 3) An inevitability 4) A personal perspectiveWe discuss some of the ways that students can use AI responsibly as an asset to their learning.Their responses also show student awareness and current understanding of the positives andnegatives of AI use for acquiring and applying foundational programming skills. The results alsoshow that the majority of students who chose to use AI did so to enhance their learning ratherthan replacing their original work.Introduction and
they might utilize itin their courses. 1. I would use LA to identify students that are struggling. 2. I would use LA to identify students that are high performers. 3. I would use LA to examine topics that students struggle with. 4. I would use analytics to identify content that struggles to engage students. 5. I would use analytics to make improvements to the course content.The second section of post-LA addressed any concerns or challenges the faculty had about LA intheir courses. 1. The amount of information can be overwhelming. 2. The amount of time required to generate insights can be overwhelming. 3. The insights from LA are already available from in-person teaching. 4. The learning analytics insights
15 H2 Art History 400 Level Electives 3 H3 Elective 43 H4 First-Year English Composition 3 1 H5 GE General Math Strand 3 H6 Upper Division Art Electives 9 H7 Art History Major Emphasis 21 H8 Core Courses 6 H9 Studio Art Course 4 H10 Second Language Fourth Semester 1
reliability, stability, and suitability. The final analysisindicates that 11.56% of students report low confidence, 11.54% record high confidence, andthe majority express moderate confidence. Lower confidence levels confidence were foundaround “model development” and “model evaluation,” which can be tied to “analysis andcalculation skills,” “optimization skills,” and “technical and computing skills.” To booststudents’ confidence using the remedial suggestions, individualized support sessions shouldbe used to discuss student concerns, address any questions or misunderstandings they mayhave, and offer personalized guidance and encouragement. Additionally, peer support groupscan show students that they are not alone and provide opportunities to
but also provides empirical evidence of its effectiveness, offering avaluable contribution to the field of educational research and the development of ASAC systems.IntroductionIn the dynamic landscape of education research, the evaluation of student experiences andperspectives is integral for fostering effective learning environments. Short answer questions insurveys and assessments provide valuable insights, capturing student perspectives on support,learning outcomes, and satisfaction. Traditional qualitative methods, while valuable for theirdepth and nuance, often struggle to efficiently handle the vast amount of textual data generatedby student surveys and assessments [1]. This data, typically collected through short answerquestions
regression modeling, neuralnetworks, and tree-based methods, are emphasized. Each topic is motivated by the analysis of acanonical, open-source data set such as those found within the UC Irvine Machine LearningRepository [20]. Data Storytelling is a method to communicate relevant information tostakeholders in an effective manner using data visualizations along with written and/or oralcommunication. The challenge for the analyst is to decide which information to include, andwhat to discard to focus attention appropriately to convey meaning with data [13], [21].This course emphasizes practicality and interpretations of each method presented and assesseslearner performance and understanding by evaluating learner-generated technical reports. In
Gen ED, Technical CompositionENGL 1033 HIST or PLSC or PSYC or Elective II SOC Intro to Object-Oriented Analytical Geometry andDASC 1204 MAT 2204 Programming for DASC (JAVA) Calculus I Role of Data Science in Today'sDASC 1223 DVSC 1013 Intermediate Data Science World UofA NorthArk Year 2: Semester 1
Paper ID #42410Credit-Hour Analysis of Undergraduate Students Using Sequence DataTushar Ojha, University of New Mexico Tushar Ojha is a graduate (PhD) student in the Department of Electrical and Computer Engineering at the University of New Mexico (UNM). His work is focused on researching and developing data driven methods that are tailored to analyzing/predicting outcomes in the higher education space. He works as a Data Scientist for the Institute of Design & Innovation (IDI), UNM.Don Hush, University of New Mexico Dr. Hush has worked as a technical staff member at Sandia National Laboratories, a tenure-track
green channel correlation method for versatile identification.Miah Abdullah Sahriar1†, Mohd. Rakibul Hasan Abed1†, Ratchanok Somphonsane2, Houk Jang3,Chang-Yong Nam3, Saquib Ahmed5,6*1 Department of Materials and Metallurgical Engineering (MME), Bangladesh University ofEngineering and Technology (BUET), East Campus, Dhaka-1000, Bangladesh2 Department of Physics, School of Science, King Mongkut’s Institute of TechnologyLadkrabang, Bangkok 10520, Thailand3 Center for Functional Nanomaterials, Brookhaven National Laboratory, Upton, New York11973, USA5 Department of Mechanical Engineering Technology, SUNY – Buffalo State University, 1300Elmwood Avenue, Buffalo, NY 14222, USACenter for Integrated Studies in Nanoscience and Nanotechnology
, business, and computer science [3]. Data Science professionals areexpected to master this diverse skill set [3]. However, the Data Science domain is constantly andrapidly changing as new technologies are incorporated into the field [3]. This ever-evolvinglandscape poses a difficult challenge to universities tasked with educating the next generation ofdata scientists. To adequately prepare students for the dynamic demands of the Data Sciencedomain, the data science competencies taught in university courses must align with the requiredskills demanded by industry. This study analyzes the alignment between Data Science competencies taught in 136undergraduate Data Science programs across the United States [4] and the skills required for full
next generation of STEM enthusiasts.B. Market InsightEAB conducted a feasibility study of a Bachelor of Science in Data Science at Wentworth. Their reportrevealed promising opportunities for program graduates based on strong employer demand trends andpositive employment projections. However, EAB also cautioned about the challenging competitivelandscape, suggesting potential difficulty for the proposed program to capture student demand. Toaddress this, the report recommended offering accelerated degree options, such as a 4 + 1 program, andincorporating experiential learning opportunities into the curriculum. Both of these suggestions wereincorporated into our curriculum.C. Program Enrollment StatisticsOur first class of Data Science
advancement in the global society, so it is crucial to understand theunderrepresentation of women in these fields.Data from the Integrated Postsecondary Education Data System (IPEDS) was used to constructFigure 1. The visualization shows the number of male and female students that completebachelor’s degrees in engineering for a given year between 2012 and 2021. The secondary axison the right shows how the male to female ratio is changing over this time span, visualized in thegray line at the top of the graph. From this graph, certain trends can be deduced. In general, overthe ten years, male and female degree completions have increased, thus the total number ofengineering degrees is increasing as well. Near the end of the time span, this number
computer science or computer engineering-specific formal education or degrees. Toassess varying perspectives, we conducted a study utilizing Reddit posts. Reddit is a platformwhere many engineering students and practitioners may talk openly about different topics. Wecollected data using web scraping and analyzed it using a couple of Natural Language Processing(NLP) techniques, including Latent Dirichlet Allocation (LDA). Using the top keywords, wethen took a manual approach, using whole posts for context to perform thematic analysis toderive the topics. Our findings suggest that non-computing engineers are generally positive aboutdata science and its potential applications. They see it as especially important for 1) CareerProspects and
3 2 1 0 None Q1 Q2a Q2b Q3a Q3b Q4 Q5a Q5bFigure 1: Distribution of quizzes selected by students to be attempted a second time. Fivestudents chose not to improve any of their quizzes.No significant difference was found in quiz performance between genders. This was basedon 95% confidence intervals for the difference of median performance between males andfemales, generated from 5,000 bootstrapped samples for each quiz shown in Figure 2. Difference in Median Score (M−F) 100
Diversified Workforce in Nuclear Energy and SecurityAbstractA workforce equipped with essential data analytics skills is crucial to maintaining the UnitedStates' economic growth and security, especially for nuclear energy industries and non-proliferation. Data analytics skills are in high demand in order to generate data-driven, robustsolutions to solve global challenges and support decision-making for stakeholders in nuclearenergy and security areas. This paper presents the technical approach that facilitates theintegration of fundamental data analytics skills into pipeline building toward a diversifiedworkforce through a suite of well-designed, comprehensive summer enrichment programs forhigh school, undergraduate and graduate students. The summer
., compared to 2.6 million in India and 4.7 million in China [1].STEM literacy is critical to human capital competency for the economy [2]. Therefore,encouraging more high school students to aspire to STEM careers can increase the likelihood ofapplying for jobs in STEM fields. Because many internal and external factors may influence highschool students’ aspirations for STEM careers, previous research on this topic often employs atheory-driven approach to identify predictors from large scale survey (LSS) data and formulatehypotheses for statistical tests. Existing LSS datasets, such as the Education Longitudinal Studyof 2002 (ELS:2002), promise a comprehensive investigation of the factors that contribute to highschool students' persistence in STEM
, Geometric Abstraction, and Mathematics as they relate toengineering and art. Woven into the theoretical content are hands-on projects where studentslearn basic sketching skills, hand build a ceramic still-life piece, visit local galleries andmuseums, and, using elements or art and principles of design, turn data into data visualizationsand data physicalizations: data-driven physical artefacts whose geometry or material propertiesencode data. Students use an adapted Jansen and Dragicevic [1] information visualizationpipeline to move from raw data to data wrangling to visual and physical presentation. This paperpresents examples of the process and concludes with observations and lessons learned.Figure 1. Informa0on visualiza0on pipeline. Jansen and
broaderrange of student responses).IntroductionThe landscape of educational data collection is rapidly evolving, with significant increases instudent enrollments and class sizes leading to an unprecedented growth in textual data fromacademic sources, such as assignments, assessments, and student feedback instruments [1] - [3].This proliferation of textual data presents a critical challenge: manual analysis methods areincreasingly untenable due to their time-intensive nature, highlighting the necessity forautomation in the assessment process, whether in whole or in part [4]. In response to this need, asignificant body of recent research has focused on the use of Natural Language to assess studentwork in the form of short answers, essays, or other
our analysis, we present acomparison of engineering school results to that of campus-wide results to uncover similarities(or dissimilarities) in extra credit accumulation patterns. The results reveal that althoughengineering and campus-wide students accumulate a similar number of extra credits, theircomposition is different. We would like to note that the methods used in this analysis, althoughapplied to the data from a specific university, are generally useful for credit-hour analysis.1 IntroductionCredit hours are a metric of time spent by a student in the classroom [4]: one credit hour equalsone hour in class every week for one semester [21, 11]. As per the requirements of all the regionalaccrediting agencies in the US, a bachelor’s
availableon campus.Our faculty-assignment optimization tool uses Linear Programming (LP) with the objectivefunction being the maximization of the overlap between the courses to be offered in a semesterand the faculty members’ preferences and skills. This maximizes the chances of every facultymember teaching courses they are interested in. A set of constraints is created to ensure the fullcoverage of all courses/sections to be offered and also to ensure that no faculty member is assignedto teach more than a pre-determined teaching load limit. The tool is embedded in a web-basedapplication and is available for the public to use.One of the greatest features of the tool is its objectivity. It generates the faculty-course assignmentsbased on the faculty
applications of Data Science technologies.Melika Akbarsharifi, The University of Arizona Melika Akbarsharifi is a Master’s student in Electrical and Computer Engineering at the University of Arizona, studying under Professor Gregory L. Heileman. Her research at the Curricular Analytics Lab focuses on using machine learning and data analysis to enhance educational outcomes. Key contributions include developing a cohort-tracking analytics platform that assists in improving graduation rates by addressing curricular barriers. Melika has co-authored papers presented at conferences such as the ASEE Annual Conference and Exposition, exploring the intersection of curriculum complexity and student performance. Her technical