Developing On-Line Homework for Introductory Thermodynamics

Journal of Engineering Education, Jul 2005 by Taraban, Roman, Anderson, Edward E, Hayes, Matthew W, Sharma, M P

ABSTRACT

Homework in engineering courses is used to develop problem-solving skills and to provide students with the practice they need in order to achieve mastery of essential concepts and procedures in their disciplines. We describe homework exercises that were developed for introductory thermodynamics and delivered to students via the Internet. Records of student use were created automatically by the computer server. The data revealed students' patterns of software usage in the context of the course; additional data from course instructors revealed the extent to which completing the on-line homework improved students' in-class test performance.

Keywords: homework, thermodynamics, on-line education

I. INTRODUCTION

Homework and other forms of applied problem solving are a major source of practice and the means to mastery of essential concepts and procedures for engineering and science students [1-11]. It is generally acknowledged that active learning and timely feedback are important elements of effective instructional materials [1-3, 5-7]. Engineering and science educators are increasingly turning to readily available computer and telecommunications technologies in order to incorporate these and other pedagogical innovations into instruction.

Beginning in the fall of 2001 to the present, two of the authors (Anderson and Sharma) incorporated computer-based homework into introductory thermodynamics at Texas Tech University (TTU) and the University of Wyoming (UWyo). These problems were written by one of the authors (Anderson) for the first ten chapters of the course textbook [12]. The instructional goal was to provide students with brief developmental exercises throughout the semester in order to review essential concepts and procedures and to prepare students for in-class evaluations.

The homework problems were made available to students over the World Wide Web. Each time students accessed the homework problems, their interactions were automatically recorded and stored on the computer server. The data in this paper are from the final semester in which we tracked students, a point at which the software had been refined and fully implemented into the course curriculum. We used the automatically-generated computer records (time-stamped key strokes and click trails) to answer three practical questions about the usefulness of the on-line homework to student learning:

* How much time did students spend doing on-line homework?

* How did students' incorporate on-line homework into day-to-day academic activities?

* Did doing on-line homework improve in-class test performance?

II. CASE STUDY

A. Participants

Data from the 29 students from UWyo and 72 from TTU who enrolled in and completed ES2310: Thermodynamics (UWyo) or ME2322: Engineering Thermodynamics I (TTU) are reported in this study. Self-reported demographic data indicated that all the students were engineering majors, predominantly male (84 percent), and Caucasian (86 percent). Half of the respondents had completed fewer than 19 engineering credits and half completed over 19 credits. The median course load was 15 credit hours.

B. Materials and Procedure

The on-line homework problems were drawn from a pool of 670 multiple-choice questions. Each question included a feedback screen that indicated the correct response. The questions were organized by topics into problem sets, each containing five questions. There were several problem sets that covered each topic; specific problem sets covering a topic were randomly assigned to students by the computer server. UWyo students were assigned 26 topics, and TTU students were assigned 35 topics, to complete during the semester, as part of their regular coursework. At TTU, students were expected to submit an on-line homework prior to most class meetings; at UWyo, on-line homework was not as uniformly assigned throughout the semester. At both institutions, students could repeat topics anytime during the semester without restriction. On-line homework counted for a small part (less than 10 percent) of students' course grades.

C. Results and Discussion

Four variables were generated for each student from the computer records. Total hits is the total number of problem sets completed by a student during the semester, and total time is the time the student logged onto the computer server while completing on-line homework; both variables indicate overall student effort. Unique hits is the number of topics a student covered in the on-line homeworks during the semester. On-line homework score is a student's average score for the on-line problem sets. Two independent measures of course performance were obtained from the course instructors for each student: average in-class test score and final exam score. Summary descriptive measures are shown in Table 1. By examining the total time, we find the answer to our first question: students spent less than 3 hours, on average, for the full semester at UWyo, and less than 5 hours at TTU. Completing on-line homework did not require an inordinate amount of time. On-line homework scores show that, on average, students scored 86-87 percent correct at the two institutions. Thus, the instructional goal of providing students with brief exercises that they could solve was fulfilled.