A Computer-Based Free Body Diagram Assistant ROBERT J. ROSELLI,1 LARRY HOWARD,2 SEAN BROPHY1 1
Department of Biomedical Engineering, Vanderbilt University, Nashville, Tennessee 37235
2
Institute for Software Integrated Systems, Vanderbilt University, Nashville, Tennessee 37235
Received 2 August 2005; accepted 12 March 2006
ABSTRACT: We developed an online Free Body Diagram (FBD) Assistant that allows students to construct 2-D FBDs and to receive constructive feedback for a wide range of practice problems. The system’s architecture allows for interoperability between learning management systems and interactive student simulations designed to improve both learning and assessment. The system gathers useful information about students’ decision processes as they construct FBDs. This information can be used with a diagnostic module we have developed to reliably, and dynamically, construct appropriate feedback that is specific to the diagram submitted by the student. Students use this asynchronous system to gain practice in constructing FBDs at their own pace, and their records can be analyzed to indicate progress over time. ß 2006 Wiley Periodicals, Inc. Comput Appl Eng Educ 14: 281290, 2006; Published online in Wiley InterScience (www.interscience.wiley.com); DOI 10.1002/cae.20088
Keywords:
free body diagram; FBD; mechanics; biomechanics
INTRODUCTION Students often have difficulty solving problems in mechanics, not because they are unable to apply Newton’s laws, but because they do not properly account for all of the forces and couples that are applied to the system being analyzed [1]. Free body diagrams (FBDs) are visual aids used to identify where forces and couples must be applied to the system they are analyzing. The correct representation
Correspondence to R. J. Roselli (
[email protected]). Contract grant sponsor: Engineering Research Centers Program; contract grant number: EEC9876363. ß 2006 Wiley Periodicals Inc.
of the system is essential when applying Newton’s laws to compute values for unknown forces and couples. We find that when students can construct a correct FBD, then chances are good that they will correctly compute the unknown reaction forces and couples. But mastery in the construction of FBDs takes time and practice [1]. This skill can be developed by consistently applying the same procedure to multiple mechanical systems found in biomechanics, civil engineering, physics, and physical therapy, to name a few. Our goal was to provide an automated method that gives students the opportunity to construct diagrams for a wide range of mechanical systems and to provide feedback that can be used to refine their diagrams. In this article we describe the rationale and 281
282
ROSELLI, HOWARD, AND BROPHY
development of a web-based practice environment within the context of representing forces and couples applied to physical systems. The FBD Assistant is just one example of a powerful modular architecture that allows for easily authoring and delivery of interactive learning environments capable of diagnosing students’ needs and providing them with productive feedback. Conceptually, the system emulates traditional intelligent tutoring systems. But operationally, the system offers an efficient method for constructing interactive learning environments around existing Java applets and FLASH animations. In addition, it illustrates how authors can construct lean, interactive environments separate from the diagnostic components. The separation of this functionality provides a development capacity that allows third party contributors to develop the diagnostic and feedback components of these interactive learning environments and refine them over time as more research is conducted on how students learn and interact with the system. The development of the FBD Assistant in parallel with the development of a general purpose authoring environment and Learning Management System has allowed us to explore a unique method for specifying automated learning activities. These activities can provide interactive lessons with formative feedback and can capture the students’ action for future evaluation. The modular architecture can provide a method for parallel and independent development of each component in the system. In this article we present the development of the FBD Assistant as a demonstration of modular architecture that utilizes software component interoperability to efficiently construct interactive learning environments and capture assessment data. The article concludes with a discussion of current trends and directions for these kinds of learning opportunities for students.
Tools for Analyzing Mechanical Systems The basic procedure for constructing a FBD consists of: (1) isolating the body of interest from all other bodies, (2) replacing contacting bodies with forces and/or couples (depending upon the type of interactions), (3) including other externally applied forces, such as the weight of the body, and then (4) generating the equations expressing the balance of these components based on Newton’s laws. Despite the simplicity of the procedure, students invariably have trouble constructing FBDs. They often fail to assign the correct reactions to specific types of supports. This recognition process improves significantly as students
are exposed to a variety of problems targeting various mechanical supports [1]. Therefore, we have focused on assisting students as they progress through the second and third stages of the FBD construction procedure. We are not aware of any computer-based biomechanics FBD tools available commercially. A number of web sites provide limited ability to construct FBDs. For example, we found four web sites that deal with non-biomechanical systems [25]. Huang [2] has developed a Java applet that demonstrates the force vectors for a free body diagram of a block on an inclined plane. The user can alter the weight of the block, angle of the plane, and frictional coefficient to explore how the magnitude and direction of these components change as the system changes. However, the vectors are already drawn, so students cannot gain practice in actually making decisions about where to place the forces and couples appropriately on the diagrams. A similar Java-based application that allows interactive FBDs to be constructed for blocks on ramps or blocks in free space was developed by a group of seniors as part of a design project at the University of Michigan [3]. When all forces acting on the block are placed on the diagram the block becomes an animated object. M. Casco Associates has developed an interactive FBD of a ladder [4]. Another Java applet for drawing FBDs of bicycle tires was developed for the bicycle module associated with Project Links [5]. These existing tools focus on how vectors can be used to visualize forces that are applied in various ways to a single system. Although it is important for students to go through the process of modeling how forces interact in a single system, a more general environment that encourages students to generate representations of forcecouple systems would be more beneficial. We need a system that helps students compare and contrast problems in multiple contexts so they can effectively notice subtle differences. In addition, we need a system that provides formative feedback on the appropriateness of a student’s representation. Although the applets identified provide interesting dynamic conceptualizations of physical phenomena, none of the applets provide the construction and feedback functions we needed to achieve all of our goals. Over the past 4 years we have developed, tested, and refined an interactive, web-based FBD Assistant that provides students with a means to gain experience in constructing 2-D FBDs [6,7]. The FBD Assistant was developed as part of the Vanderbilt-Northwestern-Texas-Harvard/MIT Engineering Research Center on Bioengineering
COMPUTER-BASED FBD ASSISTANT
Technology (VaNTH ERC). FBD modules were developed using VaNTH’s Courseware Authoring and Packaging Environment (CAPE) and delivered to students with VaNTH’s experimental Learning Management System (eLMS) [8,9]. In most problems the object of interest is in equilibrium, but some dynamic problems have also been included. Students use the tool to place all of the externally applied forces and couples on the isolated body. When they are finished, the system analyzes their representation with a set of predefined rules and provides them with feedback on their performance. If forces or couples are placed incorrectly or omitted, then students are given the opportunity to try again. If the student’s FBD is incorrect on the final attempt, the correct FBD is displayed. This initial prototype was developed with Macromedia Flash by a programmer at nTara, an industrial partner of the VaNTH Engineering Research Center [10]. The system was initially tested in a sophomore biomechanics class at Vanderbilt University [6]. Using student surveys and interviews we gathered students’ perceptions of the tool’s utility for learning and usability. Students reported that the system was beneficial in understanding how to construct a FBD. However, they requested more extensive feedback. Therefore, we revised the system to provide students with more specific information related to each misplaced and missing vector or couple. The revised system was tested the following year in a biomechanics class. Comparison of pretests and posttests showed the FBD Assistant was effective in assisting students in learning to construct FBDs [7]. However, comments from students indicated that the
Figure 1
283
feedback was sometimes confusing, especially when the diagnostics provided multiple possible interpretations of unidentified vectors or couples. These initial student evaluations led to the development of a system architecture that supports the development of a wider range of interactive simulations capable of interoperating with large enterprise systems like a learning management system (e.g., Blackboard). In addition, we developed a diagnostic method to assist in our ability to systematically interpret students’ actions without needing to rework the user interface design. The following sections describe the components of this modular architecture with the potential for constructing a large array of interactive learning environments that can be linked with the next generation of learning management systems.
ARCHITECTURE OF THE FBD ASSISTANT The construction and delivery of problems for students is accomplished with a series of components located on both an author’s local hard drive and a web server. Figure 1 outlines the interaction between these major components. First, the instructor or author develops a problem and provides a solution locally using the Solution Editor which was developed with Macromedia Flash. The Solution Editor provides a user interface for importing an image of the system to be analyzed, and tools to specify acceptable solutions for each vector and couple. This information is exported to an XML file, which can then be imported into CAPE. CAPE uses a visual editor metaphor, allowing authors to construct a sequence of learning
FBD Assistant architecture.
284
ROSELLI, HOWARD, AND BROPHY
activities for students. The modules developed in CAPE are packaged in a format that can be read by a learning management system. A special CAPE FBD module was developed specifically to display and analyze FBDs [8,9]. Next, the CAPE FBD module is published to the server running the eLMS [9], where the instructor uses the administrative aspects of eLMS to assign the module to the class. Students access the module by logging on to the eLMS and selecting the assigned CAPE FBD module. The problem is presented to the students via a Student Editor. This is another Macromedia Flash application that receives input from the students and submits it to the CAPE FBD module via an XML data structure. The CAPE module then provides diagnostics to the student. Details of each component are presented in the following sections.
Authoring Environment: The Solution Editor Construction of a FBD solution is made with the Solution Editor (Fig. 2) which was developed using Flash MX. The author uses this tool to provide the information used by the diagnostics module for
evaluating the appropriate construction of the diagram and providing context-specific feedback. From within the solution editor, the author selects a background image file, enters the problem statement, and places vectors and couples on the image. Vectors and couples are entered by clicking on the appropriate icon at the upper left corner of the workspace. Vectors or couples can be removed by first clicking on the object, then clicking on the recycle bin. The mouse is used to drag objects to the appropriate location on the diagram. The point of application is assumed to be at the tail end of the vector for a force and in the center of the icon for a couple. Cursor keys can be used to provide more precise movements. A vector can be oriented in its proper direction by clicking on the arrowhead and rotating it about the tail or by typing the angle in the object box in the upper right portion of the workspace. Clicking on the arrow of a couple will change its orientation from clockwise to counterclockwise, or vice versa. Checking the box in the ‘‘direction/tol’’ row will ensure that the student and solution vector or couple orientations are compared by the CAPE module. A tolerance of þ/x degrees can be specified for student/solution vector orientations. Removing the
Figure 2 Solution Editor for the construction of a free body diagram.
COMPUTER-BASED FBD ASSISTANT
check box turns off the orientation comparison. The magnitude comparison can be turned on or off in the same manner. The author is also responsible for providing additional information that is used when the student’s solution is checked for accuracy. A descriptive name or abbreviation can be entered for each couple and force. These are used by the diagnostic module to construct context-specific feedback by referring to the specific name when the solution is displayed to the student. A ‘‘type’’ is also selected for each vector and couple. The author can select multiple types for correct comparison of vectors and couples, or can eliminate the comparison completely by selecting the ‘‘undefined’’ type. Force ‘‘types’’ available in the current version include: undefined, weight, electrical force, magnetic force, applied force, tension, muscle force, normal force, friction force, general reaction force (GRF), horizontal component of GRF, vertical component of GRF, joint reaction force (JRF), horizontal component of JRF, and vertical component of JRF. The author can use the ‘‘major/minor axis’’ and ‘‘rotation’’ controls to select the size and orientation of an elliptical target area that will be used for comparison with the student’s point of application for the force or couple. The author is also responsible for providing a name for the point of application of the vector or couple, such as ‘‘the hand’’ or ‘‘the center of gravity.’’ This information is used in the final level of diagnostics. Once all objects have been placed on the diagram, the author saves the solution to a file in an XML-based representation by typing the file name in the box at the bottom and clicking the ‘‘save’’ button. Previously saved solutions can be accessed and edited by typing the appropriate file name in the ‘‘file’’ box and clicking on the ‘‘load’’ button.
Student Editor The Student Editor (Fig. 3) was also developed using Macromedia Flash MX, and is similar in operation to the Solution Editor. The initial screen (Fig. 3a) shows the isolated object and the problem statement appears in a pop-up window. The problem statement can be viewed at any time by clicking on the ‘‘instructions’’ icon at the left of the screen. Students add vectors and couples to the figure using the icons in the upper left portion of the screen. They drag them to the appropriate place on the diagram, name the objects, orient them, and assign types to each object. When finished, students press the ‘‘Submit’’ button to send their solution to the VaNTH eLMS, where it is persisted in the eLMS repository. They may save their
285
current state, without submitting the solution, by pressing the ‘‘save’’ button, allowing them to pick up where they left off at a later time. Students can also add comments to their FBD by clicking on the ‘‘C’’ icon on the left margin of the workspace. These become part of the permanent student record and can be accessed by the students, and instructor, when reviewing the module at a later time. The student’s solution is analyzed with the Diagnostic Module (details in the next section) and diagnostics are returned in the ‘‘feedback’’ pop-up window (Fig. 3b) in the students’ web browser window. In the example shown in Figure 3b, the student has made two mistakes. A couple should be applied to the hand, and the point of application of the reaction force should be located at the hand. The student revises the diagram by adding a couple, resubmits, and receives the feedback shown in Figure 3c. This assists the student in properly locating the reaction force, but the orientation of the couple is still incorrect (Fig. 3d). If the purpose of the assignment is to gain practice, the student is presented with the option to continue with the problem, view the correct solution to the problem, or move on to a new problem (Fig. 3e). If the assignment is for credit, the correct solution will be displayed at this point (Fig. 3f), followed by a grade for the assignment, based on the percentage of correct vectors and couples identified on the third and last trial.
Diagnostic Module and CAPE CAPE was developed by the VaNTH ERC specifically to deliver interactive and adaptive web-based courseware to students [8]. The FBD Diagnostic Module is written using the Python programming language and packaged within CAPE. We designed the Diagnostic Module to use a general set of evaluation rules and structured feedback that will work with any FBD problem generated by the Authoring Tool. CAPE has special wizards that link the diagnostic algorithms with the specific problems constructed in the Authoring tool, then uploads the problem to a server for delivery by the eLMS (see Fig. 1). These diagnostic algorithms were defined around the major decisions students need to make as they add or modify components on their diagram. Key decisions include whether or not a vector or couple is present, and if so, the location and the orientation of the components. The author could also specify if students need to include the type of the component in their specification. In earlier studies we realized that we could only diagnose the appropriateness of each vector with a limited level of certainty. Students may
286
ROSELLI, HOWARD, AND BROPHY
Figure 3 (a) Problem statement; (b) feedback at Trial 1; (c) increased feedback Trial 2;
(d) specific feedback Trial 3; (e) option to continue or see solution; and (f) display of the final solution defined in the Authoring Tool. include additional components making it difficult for the algorithm to differentiate the students’ intentions. We did not want to make specific assumptions. Therefore we made a table to define the level of certainty we could predict for rules that match the
components in their diagram. Table 1 illustrates the conditions for the rules based on what is included and the number of components that match the specific rule. Clearly, if only two components match the solution criteria, then one is not correct, but the
1
1
1
1
0
0
0
0
7
6
5
4
3
2
1
0
0
0
1
1
0
0
1
1
0
1
0
1
0
1
0
1
Typ Loc Dir
Attribute of component
Certain
Uncertain
Uncertain
Ambiguous
Ambiguous
Probable
Probable
Certain
Level of confidence in diagnosis
Table 1 Feedback Algorithm
>2 match These components are correct
1 match
hlabeli is needed, but is not Several of the components hlabel i is needed and in the quite complete right locations, but the at hAREA labeli may be direction is not right redundant The component hlabel i may be needed, but it is not located properly The component hlabeli may be needed, but its location and direction are not correct The h labeli hforce vector or The h labeli hforce vector or momenti appears to be momenti appears to be correct, but the type is not correct, but the type is clear not clear hlabeli at the hAREA The components at the The component at the Locationi could be correct, hAREA Locationi need hAREA Locationi could more information before but double check its direction be useful, but need to and type they can be fully know the type and evaluated double check its direction Component hlabeli might The component hlabeli be useful, but its type and might be useful, but location are not its type needs to be more consistent with the fully detailed solution Y of the components These components included is indeterminate are indeterminate hLabel 1i. . .
X components appear to be correct
1 match
Level 1
Feedback
>2 match
Components at hAREA labeli need more information before they can be evaluated: hlabel 1i hLabel2i. . .
These components at hAREA labeli may be redundant: hlabeli hlabeli
Level 2
COMPUTER-BASED FBD ASSISTANT 287
288
ROSELLI, HOWARD, AND BROPHY
system cannot always interpret which vector or couple the student meant. Table 1 details the level of certainty for each condition and the type of feedback it would provide to the learner. This logic table was used to construct the Python code executed by the Diagnostic Module. The diagnostic algorithm performs several comparisons as it evaluates a student’s submission. First, each component (force or couple) of the student solution is compared with the author solution and flags are set when matches are made with a component type, position, or orientation (direction). Also, the number of student components is compared with the number of correct elements. In the first round of comparisons, students are told which vectors are correct (perfect matches between student and author elements) and whether or not they have the correct number of vectors and couples on their diagram. After the second and third attempts, students are provided with additional information, depending on the placement of their elements. If they have too many elements on their diagram and some have no flags in common with a correct element, they are told that those elements probably do not belong on the diagram. If two of the three flags match an author solution, a hint based on the missing element is given. For instance, if the type and position of student vector ‘‘W’’ both match the author vector, the system will provide a diagnostic such as ‘‘The orientation of vector W appears to be incorrect.’’ If the position of an author element is not matched by a student element, the location of the missing element is revealed using the name of the point of application provided by the author (e.g., ‘‘the gymnast’s hands’’ in Fig. 3c). Every solution submitted by a student is saved in the eLMS student record database. Upon completion of the module the student or the instructor can access the module in ‘‘Review Mode’’ to examine the difficulties encountered while constructing the FBD.
DISCUSSION Many computations in engineering require the application of routine heuristics, which on the surface appear straightforward, but the application requires noticing subtle changes between systems before a learner becomes fluent with the procedures. The construction of FBDs is one example of this kind of heuristic. In our implementation of this system we have seen how students who use the system on problems targeting one set of mechanical supports outperform a second group of students who used the system to solve problems targeting a different set of
mechanical supports [7]. That study illustrates how carefully designed problem sets focused around fundamental principles can develop students’ abilities over time. It also highlights how learners can be bound to one context if that is the only context in which they are taught. The two groups in the initial study increased their fluency to recognize when and how to apply forces and couples as they practiced more problems and received additional feedback. We anticipate that students who use the FBD Assistant will learn how to construct FBDs for a range of isolated systems beyond those associated with the biomechanics course in which they used the system. In addition, we assume that these abilities will generalize to other problem types that will improve their ability to correctly solve FBDs without the use of the tool. Therefore, a tool that enforces a systematic approach for implementing procedures in homework exercises is valuable in designing effective and timeefficient learning environments. The CAPE/eLMS architecture illustrates how the components related to a learning activity can be put into a modular structure for efficient production of a learning activity. The important first step in our process was to identify the skills we wanted the students to have by the end of the biomechanics course and identify the critical points where learners will typically have difficulty. Then, we conducted an analysis of the task, which informed the actions we wanted students to take, the actions we needed to capture in our output representation, and the kinds of diagnostic capabilities necessary to assist learners in refining their representations. This ‘‘task analysis’’ and comparison between an expert model and a student model are common methods used in intelligent tutoring systems. However, those systems can be expensive to produce and often the structure of the components is highly interdependent. Furthermore, changes made to the system would require the efforts of a knowledgeable programmer. In contrast, the CAPE/eLMS architecture is a very open and flexible environment that provides a wide array of authoring capabilities necessary for achieving the practice environments we need to construct. We see this architecture as a highly efficient method for achieving interoperability between a learning management system and Flash animations and Java applet simulations (or commercial applications). If these components can provide output data in a standard format, like XML, and a known data structure, then a system like eLMS can maintain modules that add the diagnostic and feedback functionality to the learner. This module approach has many positive benefits. For example, the user
COMPUTER-BASED FBD ASSISTANT
interface typically has a wide range of challenges making the system intuitive for learners. We have had some success working with summer REU students and undergraduates to develop interesting and usable prototypes. However, time always runs out before we can develop the diagnostic algorithms and reporting systems that summarize learner’s performance. Often when the students leave, we lose the potential to reuse their code for a larger scale implementation of their work. However, if we now begin our work with them to define the ‘‘task analysis’’ and initial diagnostic scheme, then we can define the data structures that need to be shared between components. With this knowledge, our transient developers can construct useful components that we can build on over time. Typically, the diagnostic modules and the reporting systems require a longer development cycle that can be achieved by full time programming staff. We are also exploring the potential of operating with other open, and semi open, commercial applications like MATLAB. MATLAB provides a set of web services making communication between the eLMS and MATLAB possible. Therefore, we can use CAPE to design models that are run on the eLMS, which communicates with MATLAB to run the complex computations and relies on MATLAB to provide the output. The eLMS stores data related to students’ decisions as they construct their model and, through its data mining capabilities, can provide reports with detailed information. Therefore, this modular architecture distributes the development work into independent activities that can be accomplished by a team of programmers. In addition, it distributes the complexity of designing the interactive components among a larger team of developers. The challenge for interoperable systems is the achievement of a standard protocol for sharing information. The example provided in this article illustrates how specific activity sequences can be converted to a predictable set of tasks that can be programmed into the system. This procedure can be refined in a process that makes the construction of new problems easy and efficient. Another challenge is conducting research that identifies when and why students have difficulty and then constructing a learning activity that helps students increase their fluency for that skill over a wide range of problems.
CONCLUSION The FBD Assistant provides a powerful environment for students to practice their ability to construct FBDs for a wide range of exercises in biomechanics,
289
mechanics, structural mechanics, and other fields. The FBD Assistant provides an example of how the VaNTH eLMS can effectively communicate with software developed using web-based commercial packages (e.g., Macromedia Flash MX) using web services. Each element of the FBD Assistant, including the student editor, the author editor, and the CAPE diagnostic software, has been tested and found to perform properly. Each has been shown to communicate properly with the eLMS. This component architecture illustrates how we can construct interoperable components that result in an effective interactive environment for students, and stores useful information that can be used for assessment purposes. Our next steps include evaluation of sequential student submissions on eLMS for information that can help us diagnose the kinds of difficulties learners are having as they encounter specific mechanical systems. In addition, we are exploring other tasks, like the design of filters for bioinstrumentation systems that can be operationalized into an efficient practice environment for learners, but requires minimal development time by those authoring the problems. Finally, as we develop more of these systems we anticipate becoming more informed about how to standardize the communication protocol between components for the class of engineering problems we are designing.
ACKNOWLEDGMENTS This work was supported primarily by the Engineering Research Centers Program of the National Science Foundation under Award Number EEC9876363.
REFERENCES [1] J. L. Meriam and L. G. Kraige, Engineering mechanics, Wiley, New York, 2003, Vol. 1, pp 104111. [2] F. Hwang, Java Applet: Free-Body Force Diagram, 14 May 2001, http://www.phy.ntnu.edu.tw/java/ forceDiagram/forceDiagram.html (accessed 2/22/05). [3] The Physics Free Body Diagram Builder (FBDB) v1.0, http://www-personal.engin.umd.umich.edu/physics/ ACM_Intro.html (accessed 5/12/05). [4] M. Casco Associates, Innovation in Education, http:// www.mcasco.com/semhelp.html (accessed 5/12/05). [5] Project links: Mathematics and it’s applications in engineering and science, http://links.math.rpi.edu/ devmodules/bicycle/index.html (accessed 5/12/05). [6] R. J. Roselli, B. Cinnamon, P. R. Norris, S. P. Brophy, D. E. Eggers, and J. Brock, Development of an
290
ROSELLI, HOWARD, AND BROPHY
interactive free body diagram assistant for biomechanics, Proceedings of the Joint 2002 EMBS and BMES Meeting, Houston, TX, October 2002. [7] R. J. Roselli, L. P. Howard, B. Cinnamon, S. P. Brophy, P. R. Norris, M. P. Rothney, and D. Eggers, Integration of an Interactive Free Body Diagram Assistant with a Courseware Authoring Package and an Experimental Learning Management System, ASEE Annual Con-
ference (CD-ROM DEStech Publications) Session 2793: 10 pp, 2003. [8] L. Howard, Adaptive learning technologies for biomedical education, IEEE Eng Med Biol Mag 22 (2003), 5865. [9] L. Howard, CAPE: A visual language for courseware authoring, Second Workshop on Domain-Specific Visual Languages, Seattle, WA, November 4, 2002. [10] nTara, Inc., http://www.ntara.com/. Johnson City, TN.
BIOGRAPHIES Robert J. Roselli is a professor of biomedical engineering and chemical engineering at Vanderbilt University. As a VaNTH investigator he has been a pioneer in the development of challenge-based learning modules and has developed complete undergraduate courses in biomechanics and biotransport. He has conducted research aimed at testing the effectiveness of challenge-based learning, an interactive classroom communication system, and other in-class activities. He is currently developing online diagnostic learning modules which provide feedback to students while they are working homework problems. He received BS (1969) and MS (1972) degrees in mechanical engineering and a PhD (1975) in bioengineering from the University of California, Berkeley. Sean P. Brophy received his BS degree in mechanical engineering from the University of Michigan, an MS in computer science from DePaul University, and a PhD in education and human development from Vanderbilt University. He is the leader of the Learning Sciences Group in the VaNTH Engineering Research Center (ERC; www.vanh.org). He currently is an assistant professor in the Department of Engineering Education at Purdue. His research interests include reasoning with mathematics and models, technology-supported learning environments, conceptual change, and designing assessment for learning. A specific interest relates to using simulations and models to facilitate students understanding of difficult concepts within engineering as part of the VaNTH ERC. This work was conducted while Dr. Brophy was a research assistant professor of biomedical engineering at Vanderbilt University.
Larry Howard leads the Adaptive Learning Technologies Project at the Institute for Software Integrated Systems at Vanderbilt University. As a principal learning technologist for the NSF Engineering Research Center for Bioengineering Educational Technologies (VaNTH), he and his team pioneered advanced design environments and delivery platforms for online learning and their application in blended learning environments. He is participant in the Education and Outreach Program of the NSF Team for Research in Ubiquitous Secure Technologies (TRUST) Science and Technology Center supporting the creation and dissemination of online and classroombased learning materials for cyber-security. Prior to joining ISIS in 1998, he was a senior member of the technical staff at Carnegie Mellon University’s Software Engineering Institute, where his research concerned new design paradigms for flight simulation. Earlier at the State University of New York at Stony Brook he was a member of Nobel Laureate Paul Lautebur’s research group that conducted seminal investigations of nuclear magnetic resonance imaging.
