Cyber-learning tool aids students in STEM courses

March 16, 2011

Jason Vaughn Clark (center), an assistant professor of electrical and computer engineering and mechanical engineering at Purdue University, is leading work to create a new online learning tool that has been shown to improve the performance of engineering students taking traditional in-class exams, raising their scores more than a letter grade. The tool, seen on the computer screen in the background, can be helpful for students taking courses in science, technology, engineering and mathematics, or STEM subjects. Clark is working with Multidisciplinary Engineering undergraduate student Manaz Taleyarkhan (from left), mechanical engineering graduate student Prabhakar Marepalli and Alejandra Magana, a visiting assistant professor in the Department of Computer and Information Technology and the School of Engineering Education. (Purdue University photo/Mark Simons)

Download image

WEST LAFAYETTE, Ind. — A new online learning tool created by a Purdue University researcher has been shown to improve the performance of engineering students taking traditional in-class exams, raising their scores more than a letter grade.

The tool can be helpful for students taking courses in science, technology, engineering and mathematics, or STEM subjects, said Jason Vaughn Clark, an assistant professor of electrical and computer engineering and mechanical engineering.

Results suggest the tool might be used to help improve the quality of STEM education, he said.

"It has the potential to increase the retention rate of STEM students who would otherwise drop out due to substandard performance," Clark said. "And it has the potential to eventually increase the number of students who consider going into STEM fields if they know that a helpful and personalized learning tool is readily available online."

Data show the cyber-learning tool increases exam scores and decreases study time, said Clark, who used preliminary versions of the tool in two engineering disciplines: thermodynamics in the School of Mechanical Engineering during the spring 2010 semester, and linear circuit analysis in the School of Electrical and Computer Engineering during the fall 2009 semester.

"For the 120 students who used the tool, their average exam grades were over a letter grade higher than the 480 students who did not use it," he said. "Improving exam grades is very important because timed exams are the most critically significant measure of a student's ability to identify, formulate, and solve engineering problems."

The learning tool was first described in a research paper presented in June during the International Conference on Information Technology in Education in Wuhan, China.

"The tool is geared to help level the playing field by automatically adjusting to the dynamic learning curves of each student, locating and strengthening areas of weakness to minimize the number of students who get left behind," Clark said. "Each student is different. Each has different educational and socioeconomic backgrounds, different points of view and study habits, different course loads and workloads, different personal and family issues."

As much as 90 percent of a student's grade can be based on exam scores. Although the main purpose of homework is to help students learn the material and prepare for exams, it often fails because it does not have to be completed within a certain period of time, Clark said.

"Students may spin their wheels for long periods of time and get nowhere, or they may have a TV on and be answering phone calls and e-mails," he said. "The traditional homework environment is nothing like an exam environment. During an exam, students have one hour to quickly do many problems that span several weeks' worth of concepts and textbook chapters."

To emulate exam conditions, students using the tool are timed for each problem. The students are allowed to redo variations of the same problem without penalty.

"This is to make sure the students know how to do the problems correctly after each use, to know that learning has taken place," he said. "This is usually not possible with traditional homework, where students are not sure if they've done the problems correctly."

The tool allows students to review how they've performed in past assignments and then focus on improving areas where they have shown weakness. Progress reports tell students how they are doing in all areas and what areas they need to strengthen before each exam.

"There is immediate feedback in the form of hints or solutions for incorrectly answered questions," Clark said. "Unlike static textbook problems, students can modify their problems to try what-if scenarios. For example, students can change a resistor into a capacitor and compute the solution of the modified circuit. This way they get more of an intuitive feel and get immediate answers to questions outside of lectures or office hours."

Findings indicate that students who used the tool in thermodynamics had an average in-class exam score 14 percentage points higher than students who did not use the tool that semester.

"This improvement in exam performance is well outside the uncertainty of 3.7 percentage points in the data," Clark said. "In an exit survey, students mentioned that the tool optimized and reduced their study time."

The research paper was written by a multidisciplinary team from Purdue: mechanical engineering graduate student researcher Prabhakar Marepalli; Alejandra Magana, a visiting assistant professor in the Department of Computer and Information Technology and the School of Engineering Education; undergraduate student researchers Manaz Taleyarkhan from the Multidisciplinary Engineering Program and Nikitha Sambamurthy from the School of Engineering Education; and Clark.

The tool, called SugarAid, is supported by and available on the NSF-supported nanoHub.org.

Writer:  Emil Venere, (765) 494-4709, venere@purdue.edu  

Source:  Jason Vaughn Clark, (765) 494-3437, jvclark@purdue.edu

Note to Journalists: An electronic copy of the research paper is available from Emil Venere, 765-494-4709, venere@purdue.edu  

ABSTRACT

SugarAid 0.2: An Online Learning Tool for STEM  

Prabhakar Marepalli1, Alejandra Magana2, Manaz R. Taleyarkhan3,
Nikitha Sambamurthy2, Jason V. Clark 1,4

Schools of 1Mechanical, 2Engineering Education, 3Multidisciplinary,
4Electrical and Computer Engineering, Purdue University

We present an online learning tool called SugarAid version 0.2 to assist in the education of students of science, technology, engineering, and mathematics (STEM). We have used the tool in both mechanical and electrical engineering courses with positive results, such as improved written exam scores. SugarAid may be used online at nanoHUB.org with remote computation; i.e. all that is required to use SugarAid is any device with Internet browsing capability. The tool is intended to replace or complement course homework, and to provide custom review material that adapts to each student's learning curve. The tool prepares students for in-class examinations by providing timed exercises, and allows the students and instructor to know which ABET (Accreditation Board of Engineering and Technology) concepts are being addressed. Immediate grading provides instant feedback to both instructor and students. SugarAid may be configured to display image files, function plots, multiple choice questions, detailed solutions, etc. We have implemented a weakness function in SugarAid that works by remembering exercises answered incorrectly and tests for the retention of such exercises at a later date. Reference material such as lecture notes and data tables may be displayed in SugarAid. The latest version allows students to modify exercises to explore what-if scenarios by, say, replacing a resistor with a capacitor in a circuit. In this paper we describe SugarAid and examine various metrics including a comparison of exam scores by students that did, and did not, use the tool. The exam results, usage data and survey suggest that SugarAid has a positive impact on students' performance.