Davis Report - Results

More Student Learning, Less Faculty Work? The WPI Davis Experiment in Educational Quality and Productivity

V. Results Of Curricular Initiatives

In all, 11 courses in 7 disciplines were restructured using PLA-assisted CL (Table 3). Three somewhat different models of PLA use were employed in the 11 initiatives. The first, typical of the science and engineering courses, utilized PLAs to facilitate group process, group dynamics, and problem solving in small (typically 4-5 students) CL groups in intensive one-on-one settings. Group facilitation was the primary role of PLAs, and communication of course content was secondary. The second model, employed in all the PAC Mathematics initiatives, used PLAs to lead problem solving sessions for groups of 12 students. Since the 12 students typically worked in 3 groups of 4 students, PLAs did not have much time for one-on-one interaction with any specific group. In this model, the primary role of PLAs was communication of course content. The third model provides an intriguing glimpse of how the use of PLAs could be expanded to WPI's projects program. In Drama/Theatre Performance sufficiencies, PLAs were used to mentor project students through the preliminary stages of their writing tasks, and simultaneously to provide a layered organizational structure for a large group performance effort. As will be discussed later, the model used in the mathematics courses had different outcomes in terms of faculty productivity, although all models yielded improvements in student performance. Most participating faculty produced Davis project related publications and presentations, which are listed in Appendix C.

Initiatives for PLA-Assisted CL

Biology & Biotechnology
Prior to the introduction of the PAC course in 1993, a CL format NOT assisted by PLAs had been the standard for Prof. Miller's introductory biology course sequence (BB1010 and BB1020). Outcomes assessment had shown that students were benefiting from the CL experience (Goodwin et al., 1991), but the faculty time input required was not sustainable. In the spring of 1993, the pilot PAC version was offered side by side with a traditional lecture-based version of the same course. Students enrolled in the cooperative PAC format acquired the same or more factual knowledge, were significantly more satisfied with their experience, were better able to find information on their own, and learned teamwork skills when compared to students in the traditionally taught class. Total facultytime input into the PAC course was about 13.5 hours per week, about the same time as the traditional format required, and considerably less than 40 hours required by the CL format without PLAs.
Chemical Engineering
The introduction of PLAs into CM2002 enabled Prof. DiBiasio to add two new projects to the course and to continue an important laboratory project, in spite of an increase in class size. Educational outcomes included statistically significantly better examination scores, reduced attrition, and increased topic coverage in the PAC format. Decreased time spent in class, and most notably decreased time spent in laboratory supervision, resulted in an overall decrease of 24% in faculty time input.

Table 3: Davis Project Funded PAC Courses
Course	Title	PI	Duration	Funding
BB1010/1020	General Biology	Miller	1992-94 (2 yr)	$33,010
CE1030	Civil Engineering and Computer Fundamentals	Hart, Salazar, El-Korchi, Jayachandran, Albano, D'Andrea, DeFalco	1993-94	$45,750
CM2002	Intro. to Chemical Engineering	DiBiasio	1994-95	$23,083
CS2005	Techniques of Programming	Wills, Gennert, Ward	1993-94	$20,880
ES2001	Intro. to Materials Science	Demetry	1994-95	$14,141
Humanities Sufficiencies	Drama/Theatre	Vick	1995-96	$11,139
MA1023/24	Calculus III/IV	Ganter, Tang, Heinricher, Davis	1995-96	$47,631
MA1023/24 (Off-sequence)	Calculus III/IV	Fehribach, Lurie, Servatius, Tang	1996-97	$22,800 (plus add'l funding from WPI
MA2051	Differential Equations	Tang, Davis, Heinricher	1993-95 (2 yr)	$50,974
MA2071	Linear Algebra	Christopher, Bagchi, Ganter	1994-95	$21,833

Civil and Environmental Engineering
In a new team-taught course, CE1030 Introduction to Civil Engineering, students worked in PLA-assisted project teams on weekly computer projects that simulated real-life engineering projects. Surveys revealed a marked increase in the number of students taking this course who plan to major in Civil Engineering as a result of this experience. This increase in "market share" can be considered an indicator of faculty productivity. In addition to reduced lecture time and lecture preparation time, a significant savings in faculty time was realized in reduced office hours because the PLAs handled most of the student questions.
Computer Science
In CS2005 Techniques of Programming, PLAs were used to assist students with group programming projects similar to those encountered in the working world. Prof. Wills reported a reduced drop-out rate (defined as failing students who did not take the final or do the final projects) among students taking the PAC course, compared with the prior year when the course was taught by the same professor in a traditional format. There was no demonstrable decrease in student learning, and students learned valuable skills which are very difficult to teach in a traditional format and which prepare students for later MQP work. A 20% reduction of faculty time spent on instructional activities was realized, in part through reduced lecture time but more prominently in reduced office hours and miscellaneous "drop-bys". Few students needed to stop by with questions because almost all questions were answered by other group members or by the PLA. Like the faculty members, the TAs spent less time answering student questions, and the use of group projects resulted in many fewer projects for the TAs to grade. The CS Department was therefore able to allocate one less TA to the course than would normally be the case, and that TA was allocated to assist with other CS classes, resulting in second order improvement in faculty productivity in those classes.
Mechanical Engineering
In ES2001 Introduction to Materials Science, Prof. Chrysanthe Demetry introduced short group problem solving exercises into large lecture presentations, and used PLAs to assist with two problem solving and review conferences per week and a team materials selection design project. The educational outcomes of her experimental course, compared with the traditional format, included improved student knowledge and comprehension, more interest in materials science, greater student perception of improvement in oral and written communication skills and teamwork, and increased satisfaction with the course. Reduced faculty time input was due to reduced lecturing time, less use of faculty office hours and email consultation, and many fewer interventions required for group dynamics problems.
Mathematics
The Mathematical Sciences Department has experimented with three different models of large course teaching. The prevalent model when the Davis project began was to schedule students into sections of 25, each taught by a faculty member. The educational results of this model were adequate, but the low productivity was not sustainable. The department then began to convert some courses to a large lecture format, each with one weekly traditional recitation led by a teaching assistant. Inevitably, quality suffered. The PAC model as implemented in the math department provided for a compromise position of one large lecture taught by a faculty member, with two weekly small (12 students) group work sessions led by PLAs. In this section, comparisons with both the prior models are found.

In MA2051 Differential Equations, the introduction of PLAs allowed students who had previously been grouped into four small, cost-intensive classes (taught by four separate instructors) to be put into one large class with one instructor supported by PLAs in small problem solving groups. Student passing rates were improved, thereby requiring fewer students to repeat a failed course. In addition, there were more A's earned in the PAC course in comparison to its traditional counterpart. Prof. Tang reported that the PAC format resulted in a gross faculty productivity increase of 400% due to the reduction in number of faculty required.

In MA2071 Linear Algebra, Professor Christopher reduced large lecture class meetings from four to three per week. In the fourth class period, PLAs assisted teams of four students in tackling realistic applications requiring the use of computers, and in particular the use of the MAPLE symbolic manipulation package. In the first year of the experiment, faculty productivity decreased rather than increasing, due to additional time required to develop assignments and supervise the PLAs. Student attrition was reduced somewhat. Student satisfaction was mixed, with students generally happy with CL and with their PLAs, but generally less satisfied with the MAPLE package. In the PAC mode, the NR rate was 13%, whereas for the two previous years' offerings of the traditional course, the NR rates had been 16% and 20%, respectively.

The introductory calculus sequence had long been problematic in terms of student performance, satisfaction, and faculty productivity. The first of two initiatives to address the teaching of Calculus III and IV, Prof. Ganter's addressed the teaching of these two courses in the "off-sequence" term; i.e., Calculus III in B95 and Calculus IV in C96 (the "on-sequence" terms would be C96 and D96, respectively). Comparisons were made of the academic performance of three groups in each of Calculus III and IV during the 1995-95 academic year: students in large "off-sequence" courses utilizing PLAs, students in large conventional "off-sequence" courses without PLA assistance, and students in "on-sequence" courses taught in small faculty-intensive sections. Although the students in the "on-sequence" Calculus III course entered the course with higher mathematical skills than the other two groups, students in the large sections utilizing PLAs scored significantly better than the "on-sequence" students taught in small sections on performance indicators in Calculus III that were common to all three groups. In Calculus IV, however, students in the small sections performed better than students in the large PLA-assisted sections, in spite of the fact that these two groups of students entered Calculus IV with the same mathematical proficiencies as measured by the Calculus III final exam. Prof. Ganter calculated that the cost per student per term using the PLA model was only $91, compared with $116 for the large non-PLA model and $312 for the small MAPLE sections. The savings are largely from faculty salaries and TA costs, which were greatly reduced or eliminated in the PLA model.

Professor Fehribach extended the use of PLAs in off-sequence introductory calculus, in this case to those courses that enroll a mixture (and thereinlies the challenge) of advanced placement and "trailing" (those who have not passed one or more calculus courses) students. Data are still being analyzed, but preliminary results show student achievement and satisfaction similar to or slightly better than in large lecture courses not using PLAs. (For the off-sequence courses, small sections are not an option.)
Humanities
In an exciting departure from the course-based PLA model, Prof. Susan Vick, who runs a very active drama program in the Humanities and Arts Department, pioneered the use of PLAs in the advising of the Humanities Sufficiency. The Sufficiency, a project-based degree requirement with some analogies to the IQP and MQP, is an independent study/project experience culminating the study of five thematically related courses or projects in the Humanities. Theatre sufficiencies are particularly demanding of drama faculty effort in that they typically coincide with the major dramatic productions on campus which provide them with much of their subject matter. Vick used PLAs to mentor and advise students through the writing aspects of the early drafts of their sufficiency reports and toprovide a layered organizational structure for the production efforts, thus making better use of her time in advising on sufficiency content and the productions themselves. Student response was in general very positive, with the exception of a few problems that Vick attributed to her experiment in having experienced PLAs train new ones. She reported that the use of PLAs was correlated with better quality theater performances and increased her productivity.

Technology Initiatives
Five additional initiatives in four disciplines were funded in 1996 and 1997, to address quality and productivity by means of technology, not necessarily through the use of PLAs and CL. Because four of those initiatives are still in progress, the results from those initiatives are not included in the quantitative assessment of the project that comprises the main body of this report. Brief summaries of these initiatives are given in Appendix D.

Maintained by webmaster@wpi.edu
Last modified: Jun 21, 2005, 16:45 EDT