Paper ID #9578
Information Visualization for Product Lifecycle Management (PLM) Data
Ms. Chen Guo, Purdue UniversityChen Guo serves as a Teaching Assistant in the Department of Computer Graphics Technology at PurdueUniversity. She is currently pursuing her PhD in CGT from the College of Technology. Since 2011she has taught courses in Construction Graphics, Computer Graphics, Product Design, Simulation andVisualization. Her research interest includes in the area of Graphic Design, Information Visualization andInteractive Media.
Dr. Yingjie Victor Chen, Purdue University, West LafayetteDr. Yingjie Chen is an assistant professor in the department of computer graphics technology of Pur-due university. He received his Ph.D. degree in the areas of human-computer interaction, informationvisualization, and visual analytics from the School of Interactive Arts and Technology at Simon FraserUniversity (SFU) in Canada. He earned Bachelor degree of Engineering from the Tsinghua University(China), and a Master of Science degree in Information Technology from SFU. His research covers inter-disciplinary domains of Information Visualization, Visual Analytics, Digital Media, and Human ComputerInteraction. He seeks to design, model, and construct new forms of interaction in visualization and systemdesign, by which the system can minimize its influence on design and analysis, and become a true freeextension of human’s brain and hand.
Dr. Craig L. Miller, Purdue University, West LafayetteDr. Nathan W. Hartman, Purdue University, West Lafayette
Nathan Hartman is an Associate Professor in the Department of Computer Graphics Technology at PurdueUniversity, and Director of the Purdue University PLM Center of Excellence. Dr. Hartman is also Directorof Advanced Manufacturing in the College of Technology. His research focuses on examining the use of3D CAD tools in the product lifecycle, the process and methodology for model-based definition and themodel-based enterprise, geometry automation, and data interoperability and re-use. He currently teachescourses in 3D modeling, virtual collaboration, 3D data interoperability, and graphics standards and dataexchange. Professor Hartman also leads a team in the development and delivery of the online Purdue PLMCertificate Program and in the development of the next-generation manufacturing curriculum at Purduefocusing on manufacturing systems and the holistic product lifecycle.
Amy B Mueller, Purdue University, West LafayetteAmy B Mueller is a Clinical Assistant Professor in the College of Technology, Purdue University, WestLafayette campus. She received her BS in ME from Purdue University and her MBA in InformationSystems from the University of Toledo. Before joining the faculty in 2012, Ms. Mueller spent over 30years in industry and her career parallels the progression of CAD/CAM to PDM to PLM. She has heldindustry positions with Owens-Illinois, Parametric Technology, Cummins, Faurecia and Toyota IndustrialEquipment as well as a VAR and a consulting firm. She has held previous adjunct teaching positions withthe University of Toledo and Ivy Tech Community College. Ms. Mueller also worked as the Director ofMinds on Math for the Bartholomew County School Corporation which is an after school math enrichmentprogram for fourth graders. She is a member of ASEE, ACM and SWE.
Dr. Patrick E. Connolly, Purdue University, West LafayettePatrick Connolly is a Professor and Interim Head of the Department of Computer Graphics Technologywith Purdue University at West Lafayette, Indiana. He received his Bachelor of Science degree in De-sign and Graphics Technology and Master of Science degree in Computer Integrated Manufacturing fromBrigham Young University in Provo, Utah. He completed his Ph.D. in Educational Technology at PurdueUniversity. Dr. Connolly has been teaching at Purdue since 1996, and is active in several professionalorganizations. Prior to entering academia, he worked for twelve years in the aerospace and computer soft-ware industries and has extensive experience in CAD applications and design, CAE software support, andcustomer service management. His interests include solid modeling applications, virtual and augmentedreality, visualization techniques, innovative teaching methods, and distance learning.
c©American Society for Engineering Education, 2014
Information Visualization for Product Lifecycle Management (PLM) Data
Abstract
Enabling users to explore the vast volumes of data from different groups is one of product lifecycle management (PLM)’s goals. PLM must solve such problems as isolated “Islands of Data” and “Island of Automation”; the massive data flow of distanced collaborative design, manufacturing, and management; and the incapability of interpreting and synthesizing data from different perspectives.
This paper proposes a new approach from a different perspective: information visualization and visual analytics. An interactive information visualization approach was demonstrated in order to help designers gain insights into massive data and make appropriate decisions. Suggested are possible visualization methods for PLM data- structural visualization, temporal visualization, geospatial visualization, 3D model visualization, and multidimensional visualization. This idea is then demonstrated by a case study of developing an Internet-based information visualization system to visualize the Remote Control Helicopter.
Introduction
Product lifecycle management (PLM) is the process of managing the entire lifecycle of a product from its design and production to service support and retirement. Nowadays, PLM has become a mission-critical component for manufacturers, and it forms the information backbone of a product and its company1. However, facing the explosion of digital product data and different user requirements, the development of PLM is limited by (1) isolated “Islands of Data” and “Island of Automation,” (2) the massive data flow of distanced collaborative design, manufacturing, and management; and (3) incapability of humans interpreting and synthesizing data from different perspectives. The current state severely limits communication across different user groups and discourages collaborative management and concurrent product development.
3D models are used in almost all current PLM systems, which provide a realistic representation of the product in context. However, there are disadvantages in using these models. First, some parts may be invisible because they are covered by other components. People can choose section views to show interior details, but they may miss part of the external features of the object. Second, although photorealistic rendering makes the final product image look nice, the real material may be covered by the appearance of output. Moreover, some materials may have the same look in a 3D model, but they may have different weight and strength value in the real world. Lastly, 3D model visualization is unable to show metadata such as material, weight, and price.
Now PLM starts to combine 3D models with 2D visualization graphs. Teamcenter allows designers and engineers to view basic 3D measurement and 2D markup tools in a single environment. ENOVIA not only offers 3D visualization tools, but also provides 2D visualization services such as line charts and tree graphs. However, these provided 2D visualizations are still very simple in our view. The full potential of visualization has not been utilized. We believe it is essential to embed 2D visualization tools within 3D models. The integration will enable product
lifecycle participants to understand and analyze data quickly and accurately, resulting in shortened development times and lower lifecycle costs.
Growing out from the fields of Information Visualization (Infovis) and Scientific Visualization (Scivis), Visual Analytics (VA) promotes the development of science and technology in analytical reasoning, data transformation, and representations for computation and visualization2. VA has been shown to be efficient at handling massive, dynamic, and conflicting data. With the help of VA, people can synthesize information into knowledge, derive insights from data, and provide timely and understandable assessments. However, very few PLM tools currently provide sufficient visual capabilities to help users analyze abstract data. Therefore people have an absence of an exploratory “middle ground” to connect the PLM with VA technologies. Beginning in the 1990s, Internet-based PLM systems have provided a more flexible platform for users to share and work on data. The focus of this paper is to enable a new class of product data analysis tools by integrating VA technologies and Internet-based data communication into PLM. We envision that the innovative integration will accommodate communications across different groups, catalyze creative design ideas, support the exploratory data management process, and thus improve the full product lifecycle from design to manufacturing and beyond.
Current Visualization Attempts to Support PLM Data
PTC offers a robust set of 2D and 3D visualization solutions called Windchill Visualization Services (WVS) that enables users to view components by using Creo View3. Siemens provides two solutions for visually analyzing the product during its design process. The first one is NX that uses HD3D Visual Reporting from metadata to help designers understand design issues. With different color-coded tags and “see-through” settings, users can see the inside components of 3D models and comprehend data quickly4. With the integration of product views and 2D snapshots, Teamcenter’s lifecycle visualization can send CAD data to the stand-alone application viewer or the Lifecycle Viewer to provide a complete view of the whole assembly5.
Almost all these projects use spreadsheets, basic information diagrams, and tree widgets to display the product information. However, very few existing PLM systems adopt sufficient visualization technologies to support data interpretation and management. Some pioneer projects include visualizing product variations and configurations6; the use of VA approaches to predict the effects of different parameters in car engine design7; applying interactive visual analysis to support simulation runs in a hybrid-vehicle design8; and managing the flow of iron and steel associated with car production9.
Currently, product data management (PDM) technology has been used in many different manufacturing enterprises to organize design files and processes. With JT Open, WebGL, or HTML5, some researchers propose that PDM provides a collaborative environment by the means of dynamically exchanging and collaboratively visualizing 3D models. Some researchers have created an interactive visualization platform for large aircraft development10. The interactive platform provides evolutionary information in product lifecycle stages that enable the chief project engineer to accurately make decisions. Semantic mapping approach is also used in aircraft tooling design. With the use of Teamcenter Engineering (TcEng) programming technology, the semantic transmission between aircraft tooling and inventories is highly improved11.
Also, several projects use Internet-based product information sharing and visualization aiming to conquer the issue of “Islands of Data”12, 13, 14, 15. The Web can be used at different stages of the PLM cycle: such as sharing product information and knowledge during the design stage12, managing product data with the simultaneous development13, and monitoring the performance of the working system14. A combination of WebGL and X3D technology allows the successful visualization of CATIA models to the Web. It facilitates Web-based collaboration and 3D mediated communication in PLM15.
VA research has been growing rapidly in recent years and has started to transfer from research labs to real applications in industry. For example, Purdue’s VACCINE center has developed a system to analyze the historic response of U.S. Coast Guard search-and-rescue operations in the Great Lakes. This tool can help decision makers allocate resources for rescue resources16. Wang et al17 develops a VA system to help bridge managers analyze bridges and plan maintenances. Wong et al18 created a visualization system called GreenGrid to examine power system information through semantic encoding, multilevel graph visualization, and force-directed layout. Jigsaw19 and CZsaw20 enable users to make sense of a large collection of text. They offer a collection of visualizations to detect the connection among alternatives. With document view, scatter-plot view, history view, and dependency graph, these visualizations can help users examine the connection between entities and support analytical strategies. Such VA systems have been widely adopted in many domains. But it is still rare to see the application of VA on PLM.
Possible Information Visualizations for PLM Data
An effective PLM environment enables an enterprise to gain deeper insights into product data and make better decisions. Manually reading the massive amounts of data created in the product lifecycle is simply not viable. In the section below, we discuss several information visualization techniques based on Shneiderman’s information visualization taxonomy21 and its possible usage for visualizing PLM data. With these technologies, users would comprehend different kinds of data easily. It will also help users identify problems and guide the direction for future product improvement.
Structural visualization
Tree graphs for hierarchical structure: Tree graphs are a group of linked nodes, and each node (except the root) has a parent node and possible subtrees of child nodes (the first image from left in Figure 1). Many PLM systems use a tree to visualize the products’ assembly hierarchy. Teamcenter’s BOM (bill of materials) relation browser views BOMs as an expanding tree with layered nodes. Inside nodes are 2D screenshots of the parts or subassemblies. The product specification tree in CATIA displays the component structure as a tree with different icons. Aras EPLM provides a deep vertical tree layout for the BOM structure browser and product structure browser. With the tree graph of the product family, the user can easily see the hierarchical structure of the product.
Sunburst partition to visualize quantitative measurements: Extended from a general tree graph, a sunburst graph is a radial visualization technique to visualize hierarchical data. The root node is in the center of the graph. People can get the child data with different arcs by adding additional
layers (the second image from left in Figure 1). Each arc represents an assembly in a product’s hierarchical structure. Sunburst demonstrates hierarchies shaped like donuts, and one arc represents to its related value. The direct connections among nodes are not as clear as regular tree graphs. But the length of arcs provides an additional dimension to represent quantity measurement of the part/subassembly.
Network graphs to visualize network relationship: Many times the connections of entities are complex. Instead of a tree structure’s one (parent) to many (children) relation, network connections are many to many, just like the physical connections of many parts inside one product. One part may be connected to many other parts, and may have been connected by many parts, which forms a network. Various types of network graphs visualize such types of data. A dependency wheel is a powerful visualization tool to explore directed relationships among a group of entities (the third image from left in Figure 1). In the disc, each chord diagram represents a connection between two nodes. This visualization tool also demonstrates simple interactivity by using a mouse hovering on a chord to mask other dependencies and highlight the selected dependencies with different colors.
Matrix diagram to visualize strengths of relationships: Similar to a dependency wheel, a matrix diagram is another powerful tool to show the strengths of relationships among two or more groups. The matrix diagram is created in a table with rows and columns corresponding to the correlated items. The rest of the cells contain symbols or numerical values to indicate the strengths of relationships. Color or saturation can be used to denote the relative weight to the evaluation, and they make it much easier for users to comprehend the relationship (the forth image from left in Figure 1). Comparing the messy linkages in a dependency wheel, the connections may show unique visual patterns that reveal some important product assemble information.
Figure 1: Possible Structural Visualization Methods22
Structural visualization is useful to display hierarchy and network data in PLM. By differentiating node properties, such as color, size and shape in tree graphs, researchers are able to represent different part attributes such as weight, size, and material. For the proportional size of the nodes in a sunburst, they can display the quantitative metrics of data such as mass, lead-time, or cost. The thickness of the curve in a dependency wheel or different colors between nodes in a matrix diagram can designate the strength of the connection among components. Thus engineers can make appropriate design decisions, such as which parts have shorter lifespans or weaker links.
Temporal visualization
The temporal visualization method allows researchers to visualize the temporal distribution of objects. Arc diagrams are well suited to display the chronology of nodes (Figure 2, top). By drawing arcs between nodes, the visualization shows node-to-node relationships and makes it clearer for users to see how the information may evolve. With stacked layers, a stream graph can display time series data in a flowing river shape (Figure 2, bottom left). Constraining the thickness of the stacked graph also enables users to get easy access to different types of data. A connected scatter plot is another good choice to visualize data in real time (Figure 2, bottom right). A simple linear relationship may be used to represent the work-flow information related to the products.
Figure 2: Possible Temporal Visualization Methods22
Temporal visualization is useful to display the connection between time series data. Users can simulate product maintenance and see the cost changes over time, thus enabling them to plan ahead. These graphs can provide a set of prebuilt analytics that facilitate the management to maintain cost, quality, and lead time targets with temporal information. It would help designers reduce risks and raise product quality before the designs are used for full-scale manufacturing.
Geospatial visualization
Geospatial visualization helps users explore location-related data in a map view. Different kinds of color progression are used in choropleth maps to compare data values properly (Figure 3, left). By adding symbols or graphs such as circles, histograms and pie charts over an underlying map, users can create a proportional symbol map that enables them to visualize the proportion of each area (Figure 3, middle). A dot distribution map uses dot size and spacing to communicate the geographic distribution of events (Figure 3, right). Geospatial visualization is a natural choice for
detecting spatial relationships among geologically related data and helps users comprehend phenomena.
Figure 3: Possible Geospatial Visualization Methods22
Geospatial visualization tools provide users with the ability to visualize spatial relationships within large data sets. Most PLM data has a geographic location such as plant locations across the world, distributions of buyer values and seller costs, and sales territories. Oracle Business Intelligence Suite offers numerous geospatial visualization methods for PLM data. They deliver deeper analytical insights through thematic map visualization and add bar charts, graphs, and detailed reports to the map view. Anything that contains a physical location such as revenue, billed quantity, and shipped amount can be leveraged by geospatial visualization tools.
Multidimensional visualization
Multidimensional visualization is developed to deal with data of more than two attributes. The common visualization techniques for multidimensional visualization are bar chart, pie chart, parallel coordinate plot, scatter plot matrix, heat map and tree map. For example, each vertical axis in parallel coordinates corresponds to each of the dimensions, and its value represents the dimensional data (Figure 4, left). All the individual data elements are color coded and connected by lines depending on different characteristics. A scatter plot matrix is widely used for pairwise relationships. It shows ordered groupings of dimensions along vertical and horizontal axes (Figure 4, right).
Figure 4: Possible Multidimensional Visualization Methods22
Multidimensional data is everywhere in PLM. A multilevel product can consist of multiple subassemblies and parts1. Many PLM applications use BOM to show a detailed list of data
attributes. Most often the BOM is stored as a spreadsheet. It will be very hard to read if there are many parts in the BOM. Although the data can be put into a relational database to query limited information, it requires special training to use a database, and it costs more time and money. Thus multidimensional visualization techniques are suited to show the higher dimensions of BOM data. They can display the relationships among sales data, material types, warranty claims, and geometric information about parts. Moreover, through interactive filtering, zooming, and brushing, the visualization can provide more-focused analyses and touch different functions across the product lifecycle.
Obviously, information visualization provides various perspectives on PLM data through multiple visualization modules. It would enable any PLM user, including participants from design, engineering, manufacturing, and marketing, to interpret and share PLM data. The barriers of “islands of data” can be broken down, and different participants in the lifecycle can demonstrate their expertise and also inspire others with good problem-solving ideas.
A Case Study
This paper demonstrates the idea by a case study of developing an Internet-based information visualization system to help users interpret, manage, and analyze PLM data. Users can gain insight into the data via an overview of relationship, zooming, connecting and navigating. The representative data is collected from the Shuang Ma 9053 RC Helicopter (Figure 5).
Figure 5: The Shuang Ma 9053 RC Helicopter 3D Model
Framework of the Web-based product data visualization system
The framework for the Web-based visualization system is divided into three different layers according to Model-View-Controller (MVC) design23. The project constructed the 3D geometric model via CATIA and then extracted all the metadata for each assembly from CATIA to create a Bill of Material (BOM). The data include but are not limited to part number, file name, assembly level, volume, mass and link to different parts or components. Such data comprise the model layer. The controlling layer is responsible for service requests and query-task execution. With requests, such as finding a spare part or searching a subtree, the server can extract and display
PLM data. The view layer aims to provide rich interactive visualization interfaces for different roles of PLM users. Different users may be interested in different perspectives of the data. Industrial designers may be interested in the look and feel of the product, which is the 3D mode visualization. Some engineers may be interested in finding the weakest link in the product. Others may be interested in seeing the cost of parts in the product and looking for ways to reduce costs. Effectively combining visualization tools, the system can lead to better product understanding and help users make accurate decisions.
Product data visualization
The model was built according to the Shuang Ma 9053 RC helicopter specifications (Figure 5). Researchers implement three different visualization graphs for RC helicopter data with the D3js library (http://www.d3js.org). The platform enables users to click an individual node to see its 3D form and metadata structure in the webpage based on HTML 5’s WebGL technologies and three.js (http://www.threejs.org).
Figure 6: A Tree Graph to Visualize Hierarchy Data and Make Trade-offs
Visualizing general hierarchy of product data: The hierarchical tree visualization helps users see the product hierarchy, determine which parts will be required when assembling, and make it easier to find part replacements. The tree graph visualization tool enables designers to view a node’s “parent” and “sibling” (Figure 6). We provide a radial view with circular wedges to show the parent-child relationships. The root is in the center with different layers growing around it. The depth of each node refers to the path to its root and the link length represents the strength of
connection between nodes. Currently such trees must be widely adopted to visualize the Bill of Material (BOM) data of a product. There is a strong need in PLM to understand the connection between entities and manipulate sub-trees in the structure. Such analysis requires a combination of different visualization techniques. This tree graph uses the connection technique to help user explore hierarchical data from multiple views. By clicking on each node, users will navigate to a webpage showing the node’s corresponding part tree and related metadata information, making it easier to gain insight into sub-assembly data. Through connecting and navigating, users can interact with the tree structure and clarify the relationships in the data.
Making appropriate trade-offs between attributes: Engineering designers are always seeking appropriate solutions to product development. The nodes and edges in a tree can be utilized to display many attributes of the represented entity. The color of each node encodes the materials (Figure 6). Orange is assigned to multiple materials, gray to aluminum, blue to plastic and red to unavailable materials. Node size is related to part weight. Designers can make appropriate trade-offs between material, volume, and weight. Knowing the weight and bounding size of each part can help designers find the heaviest parts and stay within certain weight constraints. Other than color and size, a node can also use different shapes and boundaries to represent more attributes, for example, costs and lifespan. Within a limited display space, a static tree graph cannot accommodate extreme complex products that contain millions of parts and many hierarchical layers, for example, the Boeing 777, which has more than 6,000,000 parts. One direct solution is to create a collapsible tree graph and brings in interaction. By default, it only displays a certain level of the hierarchy without expanding to the end leaves. The user can interact with the graph to expand or collapse branches (Figure 7) by clicking on nodes. Also some visualization techniques have been proposed to visualize large trees, such as a botanical tree to visualize large information sets24, a focus+context (fisheye) technique for displaying huge hierarchical structures25 and SpaceTree to support aggregation and navigation in the large hierarchy with screen-optimized dynamic layout of nodes26.
Figure 7: A Collapsible Tree Graph to Expand or Collapse Branches
A zoomable sunburst partition is also created to demonstrate the quantity percentage of the attributes of parts (e.g., weight and costs). It uses radial space-filling visualization with labels aligned with each arc’s angle span to show part names. The color of each represents the material, and the proportional size of the node encodes the relative cost (or weight) of the material (Figure 8). This visualization technique also supports mouse hovering and clicking interaction. By hovering and clicking each node, users can smoothly zoom in and zoom out of the hierarchy. This simple interaction approach allows users to highlight certain items among thousands of elements. Thus, Designers can quickly see how much it would cost to use the material and thus have a better focus on improving the product, for example, spending more time to redesign to most expensive (or to the heaviest) parts to reduce the overall cost of the product.
Figure 8: A Zoomable Sunburst to Visualize Hierarchy Data and Relationships
Finding the strength of connection between components: Based on the helicopter BOM data, we created an L-matrix diagram to display the network relationship among parts. The strength of connection indicates the relationship between individual components. An example of a vulnerable connection would be the connection between the battery and the battery holder. An example of a strong connection would be the concentric constraint between the blade mount and the shaft of the Shaft B Subassembly. We put each part number into the corresponding row and column of cells in a spreadsheet. The color squares represent the strength of the connection
between the component in the x-axis and the other one in the corresponding y-axis. There is no sense in comparing a part to itself so the light blue squares mean that there is no correlation between them. Dark blue encodes a weak connection, and orange designates a strong connection (Figure 9). The closer the connection assignment is to the light blue diagonal, the closer the components are in the actual 3D model. For maintenance, repair and operations (MRO) of aircraft, the visual analytic tool will help them quickly find problems and make the right decision. The user can also see the importance of one part in terms of connectivity by looking at color squares in one row (or column). The more squares, the more connections the part is linked to other parts. Therefore, this part may merit closer attention for maintenance because its failure may cause the failure of other parts.
Figure 9: A Matrix Diagram to Visualize Physical Connections Among Assemblies
Mapping product structure into a 3D geometric model: The node tree graph conveys more-abstract information to users, and the 3D model shows more-realistic information. With HTML 5 and WebGL technology, the platform integrated the node tree and 3D models on the Web page (Figure 10). If users click a node, it will link to the Web page with an integration of a subtree graph, a table of product data, and a simplified part model. The subtree graph is on the top right of the following figure and it contains hierarchical information of the helicopter base. Detailed subassembly information is displayed on the top left of the following figure and users can view the metadata such as part number, material and assembly level of the base. The corresponding 3D base model is at the bottom of the figure. Users can rotate the model for 360 degree view with a mouse. With an integration of all these visualizations, customers will have a better understanding of the product development information, and various departments will have a more-effective communication to share ideas and thoughts for innovation and evaluation.
Figure 10: Web-based Product Visualization Platform for Helicopter
Conclusion and future work
We demonstrated using information visualization technologies to communicate product abstract data with vivid 3D models. This research is not intended to replace 3D models. A 3D geometric model is by far the most intuitive and popular way to provide a realistic representation of a product in context. It also delivers better insight into surface patterns of objects and enables designers to inspect for errors that might occur in the drawing process. In traditional PLM environment, designers are always working with 3D solid data, and it is not easier for them to deeply visualize hierarchical structure of product data or gain insights into important 2D information. Possible visualization techniques such as structural visualization, temporal visualization, geospatial visualization, 3D model visualization, and multidimensional visualization allow users to interactively explore large PLM data resources. Moreover, combining 2D graph data with 3D solid model will provide a faster and more intuitive way to make decisions.
While designing a visualization graph, for given type of data, there may exist several different visualization algorithms that the designer can choose. Also the choice of color, layout details, and graph elements vary greatly depending upon the nature of the data, the main purpose of product data communication, and the readers’ acceptance of different visualization methods. In this paper, we have presented our first approach of using information visualization to communicate product data. We can see that there is still a lot of work to be done in this area. The data in the use cases we gathered are directly from the 3D model of the RC helicopter. Figure 6 and Figure 8 are created based upon the same data, but displayed in totally different ways. The L-matrix diagram (Figure 9) can also be shown in different ways, such as the circular layout for networks (the third image from left in Figure 1). Compared with the circular layout, the L-matrix is wasting space, but is well organized and easier for users to read and understand the connections among different parts.
The ultimate goal of the research is to bring the power of visual analytic tools to mainstream PLM applications, e.g., Dassault’s ENOVIA27 or Siemens’ Teamcenter5. This will not only benefit engineering design and make economic sense, but it will also increase customer satisfaction. In the future, research will conduct user evaluations with internal controlled experiments and external usability surveys. Researchers will also iteratively conduct cycles of design and evaluation.
Acknowledgement
Special thanks to Zhenyu Cheryl Qian, assistant professor of Interaction Design, Nityeshranjan Bohidar, Edgar Flores, Carter E Grove, Cameron Bolinger Horton, undergraduate students from the Purdue University, for their contribution in this research.
Bibliography
1. J. Stark, Product lifecycle management: 21st century paradigm for product realisation. Springer, 2011. 2. K. A. Cook and J. J. Thomas, Illuminating the Path: The Research and Development Agenda for Visual
Analytics. 2005. 3. PTC, “Introduction to Creo View MCAD 2.0.” [Online]. Available:
http://dseserve.com/PTC/T3918_Exc_Guide/#/2/. [Accessed: 24-Dec-2013]. 4. Siemens, “HD3D Visual Reporting.” [Online]. Available:
http://www.plm.automation.siemens.com/en_us/products/nx/for-design/visual-analytics/reporting.shtml. [Accessed: 24-Dec-2013].
5. Siemens, “Teamcenter: Siemens PLM Software.” [Online]. Available: http://www.plm.automation.siemens.com/en_us/products/teamcenter/. [Accessed: 08-Oct-2012].
6. A. Pleuss and G. Botterweck, “Visualization of variability and configuration options,” International Journal on Software Tools for Technology Transfer (STTT), pp. 1–14, 2012.
7. W. Berger, H. Piringer, P. Filzmoser, and E. Gröller, “Uncertainty-Aware exploration of continuous parameter spaces using multivariate prediction,” in Computer Graphics Forum, vol. 30, pp. 911–920, 2011.
8. K. Matkovic, M. Ðuras, D. Gracanin, R. Splechtna, B. Stehno and H. Hauser, “Interactive visual analysis in the concept stage of a hybrid-vehicle design,” in EuroVis Workshop on Visual Analytics, 2013, pp. 61–65.
9. S. Nakamura, Y. Kondo, K. Matsubae, K. Nakajima, and T. Nagasaka, “UPIOM: a new tool of MFA and its application to the flow of iron and steel associated with car production,” Environ. Sci. Technol., vol. 45, no. 3, pp. 1114–1120, Feb. 2011.
10. H. Wang, G. Zhao, W. Wang, and C. Chen, “The design and implementation of the platform of interactive information visualization on aircraft product development,” in System Science, Engineering Design and Manufacturing Informatization (ICSEM), 2012 3rd International Conference (IEEE), 2012, vol. 1, pp. 180-184.
11. Y. Li, R. Yan, and J. Jian, “A semantics-based approach for collaborative aircraft tooling design. Advanced Engineering Informatics,” Advanced Engineering Informatics, vol.24, no. 2, pp. 149–158, 2010.
12. M. R. Cutkosky, R. S. Engelmore, R. E. Fikes, M. R. Genesereth, T. R. Gruber, W. S. Mark, J. M. Tenenbaum, and J. C. Weber, “PACT: An experiment in integrating concurrent engineering systems,” Computer, vol. 26, no. 1, pp. 28–37, 1993.
13. P. Trott, Innovation management and new product development. Prentice Hall, 2008. 14. H. L. Broberg and others, “Internet-based monitoring and controls for HVAC applications,” Industry
Applications Magazine, IEEE, vol. 8, no. 1, pp. 49–54, 2002. 15. P. Siltanen, and S. Valli, “Web-based 3D Mediated Communication in Manufacturing Industry,” in Concurrent
Engineering Approaches for Sustainable Product Development in a Multi-Disciplinary Environment, pp. 1181-1192, 2013.
16. A. Malik, R. Maciejewski, B. Maule, and D. S. Ebert, “A visual analytics process for maritime resource allocation and risk assessment,” in Visual Analytics Science and Technology (VAST), 2011 IEEE Conference on, 2011, pp. 221–230.
17. X. Wang, W. Dou, S. E. Chen, W. Ribarsky, and R. Chang, “An interactive visual analytics system for bridge management,” in Computer Graphics Forum, 2010, vol. 29, pp. 1033–1042.
18. C. P. Wong, K. Schneider, P. Mackey, H. Foote, G. Chin, R. Guttromson and J. Thomas, “A novel visualization technique for electric power grid analytics,” Visualization and Computer Graphics, IEEE Transactions on, vol. 15, no. 3, pp. 410-423, 2009.
19. J. Stasko, C. Görg, and Z. Liu, “Jigsaw: supporting investigative analysis through interactive visualization,” Information visualization, vol. 7, no. 2, pp. 118–132, 2008.
20. N. Kadivar, V. Chen, D. Dunsmuir, E. Lee, C. Qian, J. Dill, C. Shaw, and R. Woodbury, “Capturing and supporting the analysis process,” in Visual Analytics Science and Technology, 2009. VAST 2009, IEEE Symposium on, 2009, pp. 131–138.
21. B. Shneiderman, “The eyes have it: A task by data type taxonomy for information visualizations,” in Visual Languages, 1996. Proceedings, IEEE Symposium on, 1996, pp. 363–343.
22. d3js.org, “Data-Driven Documents.” [Online]. Available: http://www.d3js.org. [Accessed: 4-Jan-2014]. 23. E. G. Krasner and T. S. Pope, “A description of the model-view-controller user interface paradigm in the
smalltalk-80 system,” Journal of Object Oriented Programming, vol. 1, no. 3, pp. 26–49, 1988. 24. E. Kleiberg, H. Van De Wetering, and J. J. Van Wijk, “Botanical visualization of huge hierarchies,” In
Information Visualization, IEEE Symposium on, 2001, pp. 87-87. IEEE Computer Society.
25. J. Lamping, R. Rao, and P. Pirolli, “A focus+ context technique based on hyperbolic geometry for visualizing large hierarchies,” In Proceedings of the SIGCHI conference on Human factors in computing systems, 1995, pp. 401-408. ACM Press/Addison-Wesley Publishing Co..
26. N. Elmqvist and D. J. Fekete, “Hierarchical aggregation for information visualization: Overview, techniques, and design guidelines,” Visualization and Computer Graphics, IEEE Transactions on, vol. 16, no. 3, pp. 439-454, 2010.
27. Dassault, “ENOVIA - Collaborative Innovation Application - Dassault Systèmes - Dassault Systèmes.” [Online]. Available: http://www.3ds.com/products/enovia. [Accessed: 08-Oct-2012].
Paper ID #9627
Impact of Optional Supplemental Course to Enhance Spatial VisualizationSkills in First-Year Engineering Students
Dr. Deborah M. Grzybowski, Ohio State University
Dr. Grzybowski is a Professor of Practice in the Engineering Education Innovation Center and the Depart-ment of Chemical and Biomolecular Engineering at The Ohio State University. She received her Ph.D.in Biomedical Engineering and her B.S. and M.S. in Chemical Engineering from The Ohio State Uni-versity. Prior to becoming focused on engineering education, her research interests included regulationof intracranial pressure and transport across the blood-brain barrier in addition to various ocular-cellularresponses to fluid forces and the resulting implications in ocular pathologies.
Ms. Olga Stavridis, Ohio State University
Ms. Olga Stavridis, Ohio State University Olga Stavridis is a Lecturer in the College of Engineering atOhio State University, teaching First-Year Engineering for Scholars (Humanitarian Section) classes in theEngineering Education Innovation Center. She also teaches Spatial Visualization, Engineering GraphicPresentation for non-engineers and Computer Graphics SolidWorks courses. Olga earned her bachelor’sdegree in Industrial and Systems Engineering from Ohio State University and her Master’s in IndustrialEngineering from Arizona State University.
Ms. Lisa A. Barclay M.S., The Ohio State UniversityDr. Lisa Abrams, The Ohio State University
Lisa Abrams is currently serving as the Interim Director of Diversity and Outreach for the College ofEngineering at The Ohio State University. She oversees the Women in Engineering and Minority En-gineering programs promoting a culture of diversity in the College through recruitment, retention, andadvancement of underrepresented groups at all levels. Lisa received her Bachelor’s and Master’s Degreesin Mechanical Engineering and PhD degree in Industrial Engineering from Ohio State. She has sevenyears of industry experience in the areas of Design and Consulting. She was previously the Director ofWomen in Engineering Program at Ohio State and the Assistant Dean of the School of Engineering andApplied Science at Miami University. She mostly recently held the position of Assistant Professor ofPractice in the Department of Mechanical and Aerospace Engineering at Ohio State where she taught awide variety of engineering courses in First Year Engineering and Mechanical Engineering. In the lastseveral years, she has received four teaching awards including the 2013 Boyer Award for Excellence inUndergraduate Teaching Engineering Innovation and the Charles E. MacQuigg Award for OutstandingTeaching.
Dr. Sheryl A. Sorby, Ohio State University
Dr. Sheryl Sorby is a Professor Emerita of Mechanical Engineering-Engineering Mechanics from Michi-gan Technological University. She is currently serving as a Fulbright Scholar at Dublin Institute of Tech-nology. She recently served as a Program Director within the Division of Undergraduate Education at theNational Science Foundation. Her research interests include graphics and visualization. She has been theprincipal investigator or co-principal investigator on more than $9M in external funding and is the authorof numerous publications and textbooks. She was the recipient of the Betty Vetter research award throughWEPAN for her work in improving the spatial skills and ultimately the success of women engineeringstudents. Dr. Sorby currently serves as an Associate Editor for ASEE’s online journal, Advances in Engi-neering Education. In 2007, she received the Distinguished Service Award from the Engineering DesignGraphics Division of ASEE and in 2009 she was elected to Fellow status in ASEE.
Jessica Thomas, Ohio State UniversityDr. John A Merrill, The Ohio State University
c©American Society for Engineering Education, 2014
Paper ID #9627
John A. Merrill is the Associate Director of the Engineering Education Innovation Center at The OhioState University, which includes the First-Year Program. Approximately 2300 students annually takecourses in fundamentals designed to ensure student success through rigorous academics in a team-basedenvironment. His responsibilities include operations, faculty recruiting, curriculum management, studentretention, and program assessment. Dr. Merrill received his PhD in Instructional Design and Technologyfrom The Ohio State University in 1985, and has an extensive background in public education, corporatetraining, and contract research. He has made presentations at conferences held by the American Societyfor Engineering Education (ASEE) and its affiliate conference, Frontiers in Education (FIE). Dr. Merrillcurrently serves as an advisor for Engineers for Community Service (ECOS), a student-run organization atOhio State. He teaches a Service-Learning course for Engineering students, which also involves projectson behalf of a rural orphanage and vocational school in Honduras. He is a two-time recipient of theCollege of Engineering’s Boyer Award for Excellence in Teaching.
Address: The Ohio State University, 2070 Neil Ave., 244E Hitchcock Hall, Columbus, OH 43210-1278;telephone: (+1) 614.292.0650; fax: (+1) 614.247.6255; e-mail: [email protected].
c©American Society for Engineering Education, 2014
Impact of Optional Supplemental Course to Enhance
Spatial Visualization Skills in First-Year Engineering
Students Abstract
The impact of spatial visualization skills on retention and performance in undergraduate
engineering schools has been studied extensively. The National Science Foundation funded a
five-year program called “Engaging Students in Engineering” or ENGAGE. One strategy in
ENGAGE is to improve students’ spatial visualization skills. With this goal in mind, we have
developed an optional one-credit hour non-graded spatial visualization skills intervention course
at The Ohio State University which is offered to incoming first-year engineering students based
on their performance on the Purdue Spatial Visualization Test: Rotations (PSVT:R). All entering
engineering students have taken this test during summer orientation. We are now assessing the
impact of this course on student visualization skills and asking the question “Can an optional
one-credit hour spatial visualization intervention class be used as an effective instructional tool
that enables students to enhance their spatial visualization skills?”
The research design is an experimental design with control and treatment groups. The treatment
group are the students who score an 18 or below on the PSVT:R and opt to take the intervention
course. The control group are students who score an 18 or below on the PSVT:R but opt out of
taking the intervention course and enroll directly in the first-year engineering courses. The
intervention course is a one-credit hour non-graded visualization class that offers additional
representations of objects to be depicted in assigned engineering graphics drawing problems.
To date the intervention class has been given to approximately 180 students while being offered
to 358 qualifying engineering students. Only data from the first year of the study is fully
available for analysis with the course being offered to 193 students in 2012 and 88 opting to
enroll. Based on the literature, we hypothesize that students who take the intervention class will
perform better in their first-year engineering classes as well as overall academically when
compared to the control group. This paper details our findings to date.
Introduction
The utilization of spatial visualization skills plays a vital role in the STEM fields, from the
ability of biologists to understand the relationships of objects they see under a microscope, to
heavily design-oriented fields such as engineering, where one must use drawings to effectively
communicate schematics and solutions to various problems4, 5
. Accordingly, in a longitudinal
study of 400,000 American high school students, students’ spatial visualization abilities were
found to be strongly associated with the acquisition of higher degrees and occupational
credentials in STEM, and in fact its importance grows greater with higher academic degrees11
.
Moreover, this same study found that current methods for identifying students with great
potential, which often focus on verbal and mathematical abilities, often miss individuals with
high capabilities in spatial ability, creating an overlooked population with talent11
.
A great deal of research has been done on various methods of improving the visualization skills
of engineering students, as this is an important skill for both design, and communicating ideas
within the field3. Not only must engineers be able to represent their ideas about a three
dimensional object, design, or mode of manufacturing in two dimensions, but increasingly with
today’s technological advances, they must also be able to translate their ideas into a three-
dimensional model using design programs.9
Efforts to improve students’ spatial abilities are especially important for underrepresented groups
in engineering, such as women and minority groups, who are at particular risk of leaving the
field while still in school2. It has been demonstrated that females’ overall score are lower on
measures of spatial visualization skills than their male counterparts for reasons including
environmental factors, differences in learning style, and testing methodology that favors male
students9. Additionally, minority and female students can be impacted by the negative effects of
stereotype threat. Since this is not necessarily due to a lack of interest in these fields, it
strengthens the argument of making support of these students especially important2.
Strategies to successfully teach spatial skills include the use of multiple concurrent approaches
and mediums to teaching, allowing students to make cognitive connections between multiple
representations1,6
. For example, Mayer and Anderson found that the problem solving ability of
students who watched a concurrent animation and narration of the mechanical workings of a
bicycle tire pump or automotive braking system was significantly better than those who had
received narration before or after watching the animation6. Significantly, Sanchez and Wiley
found that the addition of animations to the explanation of a scientific concept (plate tectonics)
eliminated the differences in spatial ability, interest, and learning7. Supplemental classes for
students who scored lower on spatial visualization testing have been found to lead to statistically
significant gains in a post-course test evaluation, even if only a segment of the supplemental
material was taught10
.
The National Science Foundation funded a five-year program called “Engaging Students in
Engineering” or ENGAGE. One strategy in ENGAGE is to improve students’ spatial
visualization skills. With this goal in mind, we have developed an optional one-credit hour non-
graded spatial visualization skills intervention course which is offered to incoming first-year
engineering students based on their performance on the Purdue Spatial Visualization Test:
Rotations (PSVT:R). We are now assessing the impact of this course on student visualization
skills and asking the question “Can an optional one-credit hour spatial visualization intervention
class be used as an effective instructional tool that enables students to enhance their spatial
visualization skills?”
The course, ENGR 1180 (Spatial Visualization Practice and Development), implemented at The
Ohio State University, is based on a similar course taught at Michigan Technological University
over the past two decades. The question to be answered is whether a successful course at a small
town technological university with approximately 7,000 undergraduate and graduate students can
be successfully transferred to a large public land-grant university with a total enrollment of
around 57,500 students. One significant difference between the two courses is that the course at
Ohio State was offered as a pass/fail course with no grade associated with it; whereas, the course
at Michigan Tech was a fully graded course. It is important to determine the impact, if any, that
this change in course offering has on its overall impact in addition to the differing demographics
at each school.
Research Design
The research design is an experimental design with control and treatment groups. The treatment
group (TG) are the students who score 18 or below on the PSVT:R and opt to take the
intervention course. The control group (CG) are students who score 18 or below on the PSVT:R
but opt out of taking the intervention course and enroll directly in the first-year engineering
courses. The intervention course is a one-credit hour non-graded visualization class that offers
additional representations of objects to be depicted in assigned engineering graphics drawing
problems. All entering engineering students are encouraged to take the PSVT:R during summer
orientation.
Assessments of the course include a follow-up PSVT:R. To date the intervention class has been
given to approximately 180 students while being offered to 358 qualifying engineering students.
The course was offered to 193 students in 2012 with 88 opting to enroll. Based on the literature,
we hypothesize that students who take the intervention class will perform better in their first-year
engineering classes as well as overall academically when compared to the control group
(students who scored 18 or below on the PSVT:R but did not take the intervention course).
Methods
Design of Spatial Visualization Intervention Course - ENGR 1180
Autumn semester 2012 was the first time the spatial visualization intervention course was
offered. This 10-week long course (ENGR 1180) met once per week for 80 minutes and was
developed as an optional one-credit hour non-graded (pass/fail) spatial visualization skills
intervention course. It is offered to incoming first-year engineering students based on their
performance on the Purdue Spatial Visualization Test: Rotations (PSVT:R). The course is
strongly encouraged for students who scored 18 or below on the PSVT:R. The PSVT:R is taken
by all entering engineering students at summer orientation. The maximum enrollment of
students for each class section was 36. Attendance was required and was a component in the
grading scheme. The grade earned by students was either Satisfactory/Unsatisfactory.
During Autumn 2012, there were four sections of ENGR 1180 offered. The students sat four to a
table where a computer was available to each student at the table. The workbook utilized was
Developing Spatial Thinking Workbook by Sheryl Sorby and software by Anne Frances
Wysocki.8 A Tactile Modeling Set (linking cubes shown in Figure 1) were provided to students
to build the objects based on a given coded plans (Figure 2) detailed in the workbook. These
colorful linking cubes were available during class and for students to take home to assist with
homework.
Figure 1. Tactile Modeling Set (linking cubes) available for assignments both inside and
outside of class
Figure 2. Sample assignment for use with the Tactile Modeling Set
During each class, a new module was introduced during a brief lecture and then class continued
according to the structure in Table 1.
Table 1. Sample Class Structure
Class Structure: Time:
A. Instructor-led lecture and
demonstration (done on the document
camera)
15 - 20 minutes
B. In-class exercise completed as a team
at each table
10 minutes
C. Computer module completion 15 minutes
D. Open lab time to work on all
homework problems
35-40 minutes
After the students completed the in-class exercise, students (voluntarily) came up to the front of
the room to share with the class how his/her team created the drawing. Also, while each table
was working on the in-class exercise, the instructor and two Teaching Assistants (TAs) walked
from table to table to offer guidance and/or check that the students’ drawings were done
correctly.
Each module typically had four worksheets that represented each concept of that module.
Students were assigned the entire module as homework but did not know which problems would
be graded until the following class when homework was collected for grading per the criteria in
Table 2.
Table 2. Sample Grading Criteria
Points: Grading Criteria:
1 All the problems were completed and done correctly (few
errors).
0.5 Most of the problems were completed and there were more
serious errors showing lack of effort.
0 Homework was not submitted.
The following course goals and learning objectives were described in the course syllabus:
Course Goals: Through instruction and exercises to develop spatial visualization skills in
preparation for engineering coursework and/or advanced coursework, students will learn how to
visualize objects in 3D and communicate that same object on 2D medium by developing their
spatial thinking
Course Learning Objectives: Students will learn how to do the following:
Create isometric and orthogonal sketches based on given data
Create sketches of solid objects by combining them with other solid objects or revolving
them about one or more axes
Represent a 3-D object by “unfolding” it and sketching a flat pattern on paper or
computer screen
Create the sketch of an object reflected and shown as a “sectioned” view
Results
To date, pre-test and post-test data have been collected and analyzed for the 2012 academic year.
Summary tables of incoming engineering student data for the 2012 academic year by pre-
PSVT:R Score and gender (Table 3), ethnicity (Table 4), and calculated composite ACT score
(Figure 3) are shown below. Calculated ACT is used for students who only submit SAT scores.
The ACT calculated is the highest known composite ACT equivalent of an SAT score on record
for a student. Otherwise, if the composite ACT score is available, this score is used as the
calculated ACT. The average calculated ACT score for incoming first-year engineering students
who scored 18 and below on the PSVT:R was 27.9 and for those scoring 19 and above was 29.4.
The threshold (or cut-off) value for offering the visualization skills intervention course was 18 on
the PSVT:R.
Table 3. 2012 Academic Year Incoming First-Year Engineering Students by Gender and
Pre-PSVT:R Score
Num of
Students
%
Gender
Num of
Students
%
Gender
Total
Students
(%) % Total
Gender 0-18 0 - 18 19-30 19 - 30 Total (%) Total
F 82 21.4 302 78.6
384
(22.2%) 100.0
M 111 8.3 1233 91.7
1344
(77.8%) 100.0
Total 193 29.6 1535 170.4
1728
(100%) 200.0
Table 4. Incoming First-Year Engineering Students by Ethnicity and Pre-PSVT:R Score
2012 Academic Year Pre-PSVT Scores
Num of
Student
%
Ethnicity
Num of
Student
%
Ethnicity
Num of
Students % Total
Ethnicity 0-18 0 - 18 19-30 19 - 30 Total
American
Indian/Alaska Native 1 10.0 9 90.0 10 100.0
Asian 19 12.8 130 87.2 149 100.0
Black 20 24.7 61 75.3 81 100.0
Hispanic 6 11.5 46 88.5 52 100.0
Native Hawaiian/Oth
Pac Island 2 33.3 4 66.7 6 100.0
Other 9 24.3 28 75.7 37 100.0
White 136 9.8 1257 90.2 1393 100.0
Grand Total 193 126.4 1535.0 573.6 1728.0 700.0
Figure 3. Incoming First-Year Engineering Students by Calculated ACT Score and Pre-
PSVT:R Score for the 2012 Academic year. (Average calculated ACT score for 18 and
below on the PSVT:R was 27.9 and for those scoring 19 and above was 29.4)
All of the following data focus on the students who scored 18 or below on the PSVT:R and either
opted to take the spatial visualization intervention course (treatment group (TG)), or opted not to
take the course (control group (CG)). Table 5 is a summary of the TG and CG by ethnicity.
Table 6 shows the percentages of students by gender that opted in/out of taking the intervention
course.
After completion of ENGR 1180 the students were instructed to re-take the PSVT:R. These
post-PSVT:R scores by gender are shown in Table 7, and by ethnicity in Table 8. Since four
students did not pass the intervention course with an “S”, these four students are identified in
Tables 7 and 8 as “U” for “Unsatisfactory” and are omitted from any further analyses. In
addition, the seven students who failed to take the post-PSVT:R are identified in Tables 7 and 8
as “No Post.”
The calculated composite ACT for both TG and CG further broken down by post-PSVT:R scores
for the TG are shown in Table 9. The CG did not take a second PSVT:R. A t-test between the
calculated composite ACT of the TG and CG indicate a significant difference (p = 0.013)
between the two groups suggesting a self-selection bias to opt out of the intervention course
based on ACT score. Table 10 shows the average GPA for the TG and CG and the number
retained in engineering after three semesters.
Figure 4 is a histogram of post-PSVT:R minus pre-PSVT:R scores from the TG showing an
average increase in PSVT:R score of 5.3 points across the group. The average pre-PSVT:R
score for the TG was 16.01 (53.4%) and the average post-PSVT:R score was 21.4 (71.3%) which
was significant at p < 0.0001.
Table 5. 2012 Academic Year Treatment Group (TG) and Control Group (CG) by
Ethnicity
Count % Opt In Count
% Opt
Out Overall Pop
Overall Pop
% by
Ethnicity
Ethnicity TG TG CG CG Total Total %
American
Indian/Alaska
Native 1 100.0 0 0.0 1 0.5
Asian 4 21.1 15 78.9 19 9.8
Black 12 60.0 8 40.0 20 10.4
Hispanic 4 66.7 2 33.3 6 3.1
Native
Hawaiian/Oth
Pac Island 0 0.0 2 100.0 2 1.0
Other 2 22.2 7 77.8 9 4.7
White 65 47.8 71 52.2 136 70.5
Grand Total 88 45.6 105 54.4 193 100.0
Table 6. 2012 Academic Year Treatment and Control Groups by Gender
Num of
Student
% Opt in
by Gender
Num of
Student
% Opt
Out by
Gender
Overall
Pop
Overall Pop %
by Gender
Gender TG TG CG CG Total %
F 44 53.7 38 46.3 82 42.5%
M 44 39.6 67 60.4 111 57.5%
Grand Total 88 45.6 105 54.4 193 100%
Table 7. 2012 Academic Year Treatment Group Post-PSVT:R Scores by Gender
Treatment Group
Gender
Number
TG < 18
%
TG < 18
Number
TG > 18
%
TG > 18
*No
Post
%
*No
Post ** U
%
** U
F 11 25 29 65.9 3 6.8 1 2.2
M 8 18.8 30 68.1 3 6.8 3 0
*No Post = Students that did not take post-PSVT:R test
**U = Received Unsatisfactory grade in intervention course
Table 8. 2012 Academic Year Treatment Group Post-PSVT:R by Ethnicity
Treatment Group
Ethnicity
Number
TG < 18
%
TG < 18
Number
TG > 18
%
TG > 18
* No
Post
%
* No
Post ** U
%
** U
American
Indian/
Alaska
Native 0 0 1 0
0 0 0 0
Asian 2 50 2 50 0 0 0 0
Black 4 40.0 6 60.0 2 16 0 0
Hispanic 0 0.0 4 100.0 0 0 0 0
Other 1 50 1 50 0 0 0 0
White 12 21.1 45 78.9 4 6.1 4 6.1
* No Post = Students that did not take post-PSVT:R test
**U = Received Unsatisfactory grade in intervention course
Table 9. 2012 Academic Year Calculated ACT Composite by Group
Calculated
ACT Number
TG < 18
Number
TG > 18
Number
CG
20 1 0 0
21 0 1 0
22 0 1 3
23 0 1 0
24 2 7 3
25 1 0 4
26 4 5 6
27 3 6 8
28 5 14 19
29 1 12 20
30 1 2 8
31 0 7 8
32 0 1 3
33 1 2 4
34 0 0 2
Average Calc
ACT 26.8 27.8 28.5**
**p = 0.013 between TG and CG
Table 10. Overall Group Average GPA and Retention after Three Semesters
Group Avg GPA
Retention
(%)
TG 2.95 78/84 (92.9%)
CG 2.82 97/105 (92.4%)
Figure 4. Treatment Group Post-PSVT:R Minus Pre-PSVT:R
Discussion
The premise of this study is that students who score an 18 or below on the PSVT:R test can have
improved spatial visualization scores upon re-test after taking an optional one credit hour spatial
skills intervention course in addition to improved performance in the subsequent first-year
engineering courses that focuses on engineering graphics. An additional objective of this study
is to determine whether other factors such as gender, ethnicity, or composite ACT also play a
role in determining proficiency with visualization skills.
Looking at Table 3, one can see the overall percentage of female students in the general
population of the 2012 incoming first-year class was 22%. 21.4% of these female students
scored an 18 or below on the PSVT:R at orientation compared to 8.3% of males. In other words,
more than double the number of female students present at first-year orientation were without the
necessary level of visualization skills. Looking at the overall incoming 2012 class broken down
by ethnicity, there are approximately 25% Black, 11.5% Hispanic, and 9.8% white scoring an 18
or below on the PSVT:R. This is encouraging for the Hispanic population, but very discouraging
for the Black population which scores below 18 at more than twice the rate of the white
population. In addition, overall, 193 students (11%) in the 2012 academic year incoming first-
year students “qualified” to take the optional intervention course based on their score of 18 or
below on the PSVT:R.
Figure 3 looks at the calculated ACT scores for students who score 18 or below compared to
those above 18. The average calculated ACT for the < 18 group was 27.9 and 29.4 for those who
scored above 18. This difference may indicate that in fact we have different populations and
other factors are at play that help determine student scores on the PSVT:R.
Tables 5 & 6 show how the students self-selected into the intervention course by ethnicity and
gender. Overall, 45.6% of the qualifying students self-selected into the intervention course.
0
5
10
15
20
25
30
-6 - -2 -3 - -1 0 - 2 3 - 5 6 - 8 9 - 11 12 - 14 15 - 17
Fre
qu
en
cy
Post minus Pre-PSVT:R score
Treatment Group Post Minus Pre-PSVT:R
Post-Pre PSVT:RScore
When broken down by ethnicity, approximately 60% Black, 67% Hispanic, and 48% white self-
selected into the course while 54% of the females and 40% of the males opted in. This seems to
indicate that the message given by the advisors at orientation was that the intervention course
was designed to help students who did not have good results from the PSVT:R. Additionally,
they were told that since spatial visualization skills were important for success in engineering,
they were strongly encouraged to take the intervention course. Even though nearly 46% of the
qualifying students self-selected to take the course, the course instructor noted that a significant
amount of time during the first few sessions was spent trying to ease students’ perceptions that
they somehow did not “have what it takes” to become a successful engineer. It appeared that
being placed in this optional course stigmatized the students and diminished their self-confidence.
This was especially true for those students who shared dorm space with fellow engineering
students who did not get encouraged to take the intervention course.
This realization that students were stigmatized by taking an intervention course necessitated a
change in the approach for the advisors at orientation for the incoming 2013 class. The emphasis
was placed heavily on the fact that these students did not have as much experience with graphics
skills prior to orientation, specifically, having the opportunity to take an engineering graphics
course during high school. Consequently, these students would need to strengthen their skill set
similar to the study of a foreign language in that without any prior instruction in that language, it
would follow that they would need additional instruction to catch up to those who already had
experience building their foreign language skills.
Tables 7 & 8 focus on the breakdown of the post-intervention course post-PSVT:R scores by
gender and ethnicity. If a successful outcome is defined as scoring above 18 after successful
completion of the intervention course, then, overall, 72% of the female students and 79% of the
male students were successful. Broken down by ethnicity, 60% of the Black students, 100% of
the Hispanic, and 79% of the white students were successful.
In an attempt to understand why some students were successful in the intervention course, Table
9 looks at the post-PSVT:R scores by calculated ACT. It appears as though students with higher
calculated composite ACT self-select out of taking the intervention course, creating significantly
different populations (p = 0.013). Table 10 shows that the average GPA for the TG is slightly
higher than the CG, although not statistically significant even though the average ACT was
statistically higher for the CG. The retention rate for both groups is nearly equal.
Another measure of success in the intervention course is the gain in PSVT:R scores by each
student. Figure 4 shows the actual gain for each student in the intervention class (TG) with the
average gain being 5.3 points across the group. The average pre-PSVT:R score for the TG was
16.01 (53.4%) and the average post-PSVT:R score was 21.4 (71.3%) which was significant at p
< 0.0001.
Why nearly 20% of the students who opt to take the intervention course are not able to score
above an 18 on re-test may indicate that this group may need a longer, more intense intervention
in order to bring them up to the same level as the other groups. It may also have to do with the
self-confidence level of these students feeling stigmatized by taking an intervention course.
Another possible reason is that the Ohio State course is a non-graded course meaning that the
students are not as motivated to complete all assignments and to perform their best on the post-
test. Students may not have taken the post-test very seriously since it was not graded as
evidenced by six students not even taking the post-test, and hence may have resulted in lower
than expected scores.
Table 11. Summary of Pre-/Post-PSVT:R testing at Ohio State and Michigan Tech
Average Pre-Test Average Post-Test Significance of Gain
Ohio State (n=81) 16.01
(53.4%)
21.4
(71.3%)
p<0.0001
Michigan Tech (n=33) 15.4
(51.3%)
23.6
(78.7%)
p<0.0001
Table 11 presents a summary of the pre-/post-PSVT:R test scores at Ohio State and Michigan
Technological University. It appears that the students in the Ohio State TG do not experience
average gains in spatial skills (18.0%) at a level comparable to the Michigan Tech students
(27.3%). The most likely reason for this difference is that the Ohio State course is a non-graded
optional course so the students may not be as motivated to complete all assignments and to
perform their best on the post-test. At Michigan Tech, the PSVT:R post-test counts as 15% of
the final exam grade in the course, so students are highly motivated to do their best on this test at
the end of the course. In addition, the student populations at Michigan Tech and The Ohio State
University are different populations with varying demographics, which may affect student
performance in the intervention course.
Conclusions
This study is an important first step in verifying whether a spatial skills intervention course
offered at Michigan Tech over the past two decades can be replicated at a large public land-grant
university with similar results. It appears that the gains in spatial skills at Ohio State are not as
large as those obtained at Michigan Tech. This could be due to the fact that the Ohio State
course is offered on a pass/fail basis rather than as a graded course. The performance of these
students will continue to be tracked into the future to determine if there are any long-term
benefits of the spatial skills training in this new setting.
References
1. Ault, H.K., John, S. “Assessing and Enhancing Visualization Skills of Engineering Students in Africa: A
Comparative Study.” Engineering Design Graphics Journal Vol. 74, Issue 2, 2010.
2. Beasley, M. and Fischer, M. “Why they leave: the impact of stereotype threat on the attrition of women and
minorities from science, math and engineering majors.” Soc Psychol Educ, 15:427–448, 2012.
3. Crown, S.W. “Improving Visualization Skills of Engineering Graphics Students Using Simple JavaScript
Web Based Games” Journal of Engineering Education, Vol. 90, Issue 3, p. 347–355, July 2001.
4. Eberbach, C. and Crowley, K. “From Everyday to Scientific Observation: How Children Learn to Observe
the Biologist’s World.” Review of Educational Research, Vol. 79, issue 1, pp. 39-68, March 2009.
5. Jones, M.G., Gardner, G., Taylor, A.R., Wiebe, E., and Forrester, J. "Conceptualizing Magnification and
Scale: The Roles of Spatial Visualization and Logical Thinking". Research in Science Education. 41, no. 3:
357-368, 2011.
6. Mayer, R.E., Anderson, R.B. “The Instructive Animation: Helping Students Build Connections between
Words and Pictures in Multimedia Learning.” Journal of Educational Psychology, Vol. 84, N.4: 444-52,
Dec 1992.
7. Sanchez, C.A. and Wiley, J. “Sex differences in science learning: Closing the gap through animations.”
Learning and Individual Differences, 20, pp. 271–275, 2010.
8. Sorby, S.A. and Wsocki, A.F. “Developing Spatial Thinking Workbook and Software.” Cengage Learning,
1st edition, 2011.
9. Sorby, S.A., Leopold, C., and Górska, R. “Cross Cultural Comparisons of Gender Differences in the Spatial
Skills of Engineering Students.” Journal of Women and Minorities in Science and Engineering, Vol. 5, pp.
279-291, 1999.
10. Veurink, N. L., A. J. Hamlin, Kampe, J. C. M., S. A. Sorby, D. G. Blasko, K. A. Holliday-Darr, et al. 2009.
"Enhancing Visualization Skills-Improving Options and Success (EnViSIONS) of Engineering and
Technology Students". Engineering Design Graphics Journal. 73 (2): 1-17.
11. Wai, J., Lubinski, D., and Benbow, C.P. “Spatial Ability for STEM Domains: Aligning Over 50 Years of
Cumulative Psychological Knowledge Solidifies Its Importance.” Journal of Educational Psychology, Vol.
101, No. 4, 817–835, 2009.
Paper ID #9841
Assessment of Students’ Changed Spatial Ability Using Two Different Cur-riculum Approaches; Technical Drawing Compared to Innovative ProductDesign
Dr. Mark E Snyder, Illinois Institute of Technology
Architectural Engineering Faculty at IIT. Creating and testing innovative classroom pedagogy for the last10 years. Evaluating the link between visualization and improved abstraction skills to specific classroomactivities. Investigating the connection between ethical judgement and academic motivation to improvethe learning environment.
Prof. Matthew Spenko, Illinois Institute of Technology
c©American Society for Engineering Education, 2014
Assessment of Students’ Changed Spatial Ability Using Two Different Curriculum Approaches; Technical Drawing Compared
to Innovative Product Design
Introduction
Improving student performance on academic tasks in mathematics, science and engineering appears to occur when students’ spatial visualization skills have been improved. Studies have found improving spatial visualization can increase success in chemistry (Carter, et.al, 1987)1, reduce math anxiety (Maloney, et.al., 2011)2, improve calculus grades (Sorby, et.al., 2012)3, and increase retention and success in science and engineering curricula (Potter, et.al., 2006)4. Creativity in the sciences and engineering seems directly related to spatial visualization based on work by Kozhevnikov, et.al. (2013)5.
A widely used pathway to improve students’ spatial abilities employs three-dimensional software programs that enable a student to create an object and manipulate the object in real-time. Yet, studies seem to suggest that this approach is not always successful when spatial ability is tested using a standard spatial ability test. Work by Frey et.al. (2000)6 and Towle et.al. (2005)7 indicate that sole use of three-dimensional imaging software does not improve spatial ability in a significant way. Other work by Study et.al., (2011)8 and Veurink et.al, (2009)9 suggest that a mixed approach using hand drawing techniques and software may be more effective in increasing spatial abilities.
Finding an effective approach to improve spatial ability is considered an important research and pedagogical imperative for the profession of engineering. Research by Charyton et.al., (2011)10 explored the relationship between spatial visualization and creativity in engineering design tasks and found convergent validity between assessments for creativity and the Purdue Spatial Visualization Test-Rotations; this infers that improving spatial abilities may improve student creativity which, in turn, may help students meet today’s engineering challenges. Seminal work by Sheppard et.al., (2009)11 in Educating Engineers, Designing for the Future of the Field found design projects that could foster an “approximation to professional practice” were important to invoking a sense of real-world practice, improving teamwork and creativity but entered the curriculum too late to give students a sense of what professional practice entails. Therefore, increasing students’ spatial abilities in the context of design projects early in their academic career may be important for preparing students to enter practice with the mindset needed for the profession.
Study Purpose
This study attempts to determine to what degree a product design course involving creative real-world problem-solving, limited hand drawing, three-dimensional software modeling and building models improves students’ spatial ability compared to a traditional technical drawing class. The authors view this study as a pedagogical exploration that may establish an improved approach to significantly increase student’s spatial visualization capability and identify how changes are distributed over cohorts and genders.
Study Design
The Purdue Spatial Visualization Test – Rotations (PSVT-R) will be used to quantify a before-and-after change in student exposure to the course’s differing content and approach. The PSVT-R (Bodner and Guay, 1987)12 is used significantly in the literature as a standard test for spatial ability (Carter, et.al., 1987)1. The test utilizes a set of line drawings of an object that the test taker must manipulate mentally to arrive at a solution that mimics the rotations of an example object. Objects are not repeated and each question contains different example objects and test objects the student must manipulate mentally. The test is timed to prevent the test taker from analytically solving each problem; that is, draw in axes and determine the rotations needed to obtain the example object’s final orientation. The test is highly correlated to test takers scores on spatial tasks (Kovac, 1989)13 and has high construct validity (Branoff, 2000)14.
The PSVT-R was given to students before the beginning of each course and near the end of the semester. Data was gathered on student gender and class cohort at the time of testing. Only paired data sets (each student’s pre and post test score) are included in the final analysis to provide an overall assessment of change in spatial ability.
The two courses are referred to in this paper as MMAE 232 and CAE 100. The MMAE 232 course is the engineering design course and CAE 100 is the technical drawing course. Mechanical engineering and material science majors normally take MMAE 232 in their sophomore year. Civil engineering majors take CAE 100 in the freshmen or sophomore years, but many students take it when it fits easily in their curriculum plans.
MMAE 232 – Design for Innovation is a sophomore-level design course for mechanical engineers. It is the first in a series of three design courses. Although it is a sophomore level class, several juniors and seniors take the course, especially transfer students. Students take the second course in the series, which focuses on machine elements, in their junior year. The third and final design course is the capstone mechanical design course which students take their senior year.
The mechanical engineering department has taught Design for Innovation for three years, beginning in the fall of 2011. The course has three main objectives: 1) introduce design thinking and open-ended problem solving earlier in a student’s career, 2) teach technical writing, and 3) improve student use of three-dimensional CAD software.
Students begin the class with two-weeks of lecture on isometric hand-drawings, engineering drawings, and the basics of CAD software. Students use Autodesk Inventor for this class. Coupled with lectures are weekly 3-hour lab sections for the students to become familiar with the CAD software or work on their projects. In the lab sections, students follow a tutorial in a book (Autodesk Inventor Essentials by Thom Tremblay, Wiley Publishing). Instructors are present to help answer students’ questions. Coupled with this portion of the class are two individual assignments in which students must create part files, assembly files, and engineering drawings.
After the completion of the CAD assignments in the first three weeks, students form groups to focus on three open-ended design problems in which they must design and fabricate a device. In
the past two years, the problems have included a chair made entirely from foam-core board without any fasteners or glue, a trebuchet in which the main axle is made from ¼” diameter acrylic rod (this requires the students to perform stress and deflection analyses), and a bio-inspired robot fabricated using Arduino microprocessors and RC servos.
Each project focuses on a particular aspect of the design process. The chair represents sustainable design techniques such as light-weighting, whole-system design thinking, and lifecycle thinking. The trebuchet project encourages students to focus on the analysis portion of the design process, a step that many students overlook. The bio-inspired robot project introduces students to the bio-inspired design process, mechanism design, actuators, and mechatronics.
For each of the projects, students must create isometric sketches of their conceptual designs and engineering drawings of their final designs; as well as building the actual object and demonstrating that it works. Final grades are based on the quality of the technical communication as well as engineering drawings associated with each design phase.
CAE 100 – Introduction to Engineering Drawing and Design - is a freshmen level technical drawing class. Students who take this course may come from any cohort (freshmen, sophomore, junior or senior) since the class is not a pre-requisite for later courses. The text used in CAE 100 is Technical Drawing with Engineering Graphics, by Giesecke, et.al.,14th edition (2012)15. The course covers the basics of free-hand sketching and the use of instruments (triangles, t-squares and compass), lettering, isometric projection, orthographic projection, and two-dimensional drawings using scale. No computer software is used in this class.
CAE 100 spends a significant amount of time (almost a third of the semester) improving students’ free-hand sketching ability. Natural and mechanical objects are used as subjects of drawing. As students improve they are introduced to the basics of isometric drawing using hand drawing. This leads easily into perspective which is explored in the students’ collection of drawings. Instruments are introduced starting with lines, line thickness, lettering and the use of the compass.
Once orthographic perspective is introduced students perform orthographic hand drawing with simple objects. Measurement is introduced (types of measures, use of engineering and architectural scales) and related back to hand drawing of orthographic perspective of objects.
Use of the T-square and triangles are introduced in order to create orthographic drawings quickly and to proper scale. Continued practice with instruments and engineering objects give the students practice in creating accurate engineering drawings with appropriate dimensioning and are related to engineering drawings in various engineering disciplines. Grades in CAE 100 are based on completing a number of drawings and the quality of student’s orthographic drawings to meet industry standards.
General Results
The following tables show the overall data collected for both courses. Data are broken out into overall course scores on the PVST-R (out of 20 possible points) then by cohort and gender for CAE 100 and MMAE 232.
Table 1 shows results for the CAE 100 course. PSVT-R scores are means followed by standard deviation (SD). The column for increased and decreased scores is the number of such changes followed by the mean percentage change in PSVT-R score.
CAE 100 Number
of Students
PSVT-R Pretest score mean(SD)
PSVT-R Posttest
score mean(SD)
Number of increased
scores (mean percent
increase)
Number of decreased
scores (mean percent
decrease)
Number of unchanged
scores
All students 42 14.14(3.92) 15.69(2.78) 24(42%) 12(12%) 6(0%) Freshmen 19 15.05(3.14) 15.79(2.99) 8(27%) 7(9%) 4(0%) Sophomores 12 14.17(4.04) 16.67(2.41) 10(37%) 2(15%) 0(-) Juniors 6 13.00(4.90) 14.33(2.16) 2(127%) 3(17%) 1(0%) Seniors 5 12.00(5.15) 14.60(3.36) 4(42%) 0(-) 1(0%) Males 31 14.65(3.83) 16.19(2.74) 18(37%) 10(11%) 3(0%) Females 11 12.73(4.08) 14.27(2.59) 6(52%) 2(16%) 3(0%)
Table 1. General results for the CAE 100 course.
MMAE 232
Number of
Students
PSVT-R Pretest score mean(SD)
PSVT-R Posttest
score mean(SD)
Number of increased
scores (mean percent
increase)
Number of decreased
scores (mean percent
decrease)
Number of unchanged
scores
All students 38 15.26(3.43) 16.10(2.76) 19(26%) 12(13%) 7(0%) Sophomores 11 15.09(2.54) 16.54(2.46) 6(35%) 4(15%) 1(0%) Juniors 18 15.66(4.10) 16.00(3.26) 9(18%) 6(12%) 3(0%) Seniors 9 14.66(3.12) 15.77(2.16) 4(25%) 2(11) 3(0%) Males 33 15.12(3.48) 16.09(2.91) 16(28%) 10(13%) 7(0%) Females 5 16.20(3.19) 16.20(1.64) 2(14%) 2(13%) 1(0%)
Table 2. General results for the MMAE 232 course.
The general results show that the CAE 100 course had higher mean percentage increase in PSVT-R scores from pre to post test compared to the MMAE 232 course. Decreased and unchanged PSVT-R scores were similar for both groups. The CAE 100 male students started with lower PSVT-R scores then the male MAE 232 scores which is consistent with previous discipline specific PSVT-R studies that showed that mechanical engineering students (the MMAE 232 students) have a slightly higher PSVT-R scores then civil engineering students (Veurink, et.al., 2012)16. Interestingly the same study showed civil engineering females (CAE 100) should have PSVT-R scores similar to the MMAE 232 females, which is not the case for this study.
Analysis of Student Data A statistical analysis was undertaken to determine if the differences within the CAE 100 and MMAE 232 overall score changes for cohorts and gender indicated any statistically significant changes. A two-tailed T-test for paired data was performed for the CAE 100 course (all students). The calculation showed a significant difference in the scores for pre-test (m=14.14, SD=3.92) and post-test (m=15.69, SD=2.78) conditions; (t(41)=3.05, p=0.004). Cohen’s d statistic gave a value of 0.457 which makes the effect medium in scale (Cohen, 1988)17. An examination of the CAE 100 cohorts (freshmen, sophomore, etc) showed no statistically significant difference in pre- and post-test scores for the freshman, junior or senior cohorts. The sophomore cohort did show a statistically significant change in scores from pre-test (m=14.17, SD=4.04) and post-test (m=16.67, SD=2.41) conditions; (t(11)=2.51, p=0.029). Cohen’s d statistic was 0.759 which is considered a medium to large change. Gender differences for the CAE 100 course showed females did not have a statistically significant change in scores from pre- to post-test with the PSVT-R. Males did show a statistically significant change in scores from pre-test (m=14.65, SD=3.83) to post-test (m=16.19, SD=2.74) conditions; (t(30)=2.73, p=0.010). Cohen’s d was 0.459 which is considered a medium effect. Performing a two-tailed T-test for paired data for the MMAE 232 course (all students) showed no statistically significant change in scores from pre- to post-test with the PSVT-R. Examination of the MMAE 232 cohorts (sophomore, junior, and senior) also showed no statistically significant change from pre-to -post-test scores for the PSVT-R within each cohort. MMAE 232 gender grouping did show a statistically significant change from pre- to post-testing of the PSVT-R. MMAE 232 males showed a change from pre-test (m=15.12, SD=3.48) to post-test (m=16.09, SD=2.91) conditions; (t(32)=2.04, p=0.049). The Cohen’s d statistic was 0.302 considered a small to medium effect. There were not enough female pairs in MMAE 232 class to perform a T-test for females only. Discussion of Analysis The statistically significant moderate to large change in PSVT-R scores for the CAE 100 class and lack of an equivalent change in the MMAE 232 course indicates the use of product design coupled with three-dimensional software and building prototypes is not as effective at increasing students’ spatial ability compared to a focused technical drawing curriculum. Comparing the graded assignments of each course suggests to improve spatial ability student effort must be coupled with direct drawing assignments completed by all students to ensure an improvement in PSVT-R scores.
The CAE 100 course required each student to complete all drawing assignments for a grade while the MMAE 232 course utilized groups that handed in one set of drawings; potentially some students did not create drawings reducing the chance to improve their spatial visualization skills. Data was not collected on which students prepared drawings in MMAE 232 but faculty observation indicated that one student in each group accomplished most drawings handed in for assignments. The analysis indicated that the CAE 100 sophomore cohort showed a significant change in PSVT-R scores over the course of the semester. The sophomores involved in the study were in four separate sections so it does not appear that the score improvement was related to grouping the students together. Further testing and background information would have to be accomplished with all CAE 100 cohorts to establish common variables that may have influenced the improvement in PSVT-R scores. The percentage of students that showed no change or a decrease in PSVT-R scores for CAE 100 and MMAE 232 was 43% and 50% respectively. This sizable number suggests that some aspect of both courses is failing to influence students in a positive manner and should be explored to understand it as a confounding factor. Study Limitations Student perceptions of academic tasks were not assessed. This evidence, although anecdotal, would have provided a sense of; which assignments changed student perceptions of their spatial ability, did students in both classes receive the same number of impactful assignments, and whether the approach in either class was affecting the intended educational objectives. The affect of maturation on student PSVT-R scores is unclear. The authors believe the data gathered for this study are not adequate to make a statement concerning maturation effects and highlight the need for such information in future work to account for potential influence. Psychological assessments related to maturation via student motivation or perseverance on tasks may provide a means to assess a relationship between maturation and the PVST-R. Nearly half of the students in this study had no change or a decrease in PSVT-R scores regardless of the curriculum approach. The authors believe this statistic requires explanation. Unfortunately previous research does not appear to consider this aspect or provide explanations for its potentially confounding influence. The authors intend to examine this aspect in future work. The amount of student-faculty engagement was not quantified for this effort. The quality and degree of student-faculty engagement could be a significant variable. The MMAE 232 classes were typically large (over 90 students) with limited time for the faculty member to interact with groups. In CAE 100 classes were typically smaller than 15 students to each faculty member but did not include group projects. It is possible this difference in engagement may have affected PSVT-R changes in both courses.
Conclusion The use of a product design methodology coupled with hand drawing, three-dimensional software and prototyping of solutions had no statistically significant increase in PSVT-R scores. In contrast, a fairly traditional technical drawing class had a moderate to large significant increase in PSVT-R scores. Future work with these two classes should assess student perceptions of what assignment aspects seem to improve their spatial ability, quantify student maturation and attempt to assess and equalize the degree of student-faculty engagement in both classes. With these additional factors controlled the results of the current pedagogical exploration can be quantified with known or measurable classroom or cohort attributes that may facilitate the transferability of successful practices to other institutions that want to increase student’s spatial visualization skills. 1. Carter, C.S., Larussa, M.A., and Bodner, G.M. (1987). A Study of Two Measures of Spatial Ability as Predictors of Success in Different Levels of General Chemistry. Journal of Research in Science Teaching, 24(7), 645-657. 2. Maloney, E.A., Waechter, S., Risko, E.F., and Fugelsand, J.A. (2012). Reducing the Sex Difference in Math Anxiety: The Role of Spatial Processing Ability. Learning and Individual Differences. 22, 380-384. 3. Sorby, S., Casey, B., Veurink, N., and Dulaney, A. (2012). The Role of Spatial Training in Improving Spatial and Calculus Performance in Engineering Students. Learning and Individual Differences, 26, 20-29. 4. Potter, C., Van Der Merwe, E., Kaufman, W., and Delacour, J. (2006). A Longitudinal Evaluative Study of Student Difficulties with Engineering Graphics. European Journal of Engineering Education. 31(2), 201-214. 5. Kozhevnikov, M., Kozhevnikov, M., Yu, C.J., and Blazhenkova, O., (2013). Creativity, Visualization Abilities, and Visual Cognitive Style. British Journal of Educational Psychology. 83, 196-209. 6. Frey, G., and Baird, D. (2000). Does Rapid Prototyping Improve Student Visualization Skills. Journal of Industrial Technology. 16(4), 2-6. 7. Towle, E., Mann, J., Kinsey, B., O’Brien, E et.al. (2005). Assessing the Self Efficacy and Spatial Ability of Engineering Students from Multiple Disciplines. 35th ASEE Frontiers in Education Conference, October. Session S2C. 8. Study, N., (2011). Long-Term Impact of Improving Visualization Abilities of Minority Engineering and Technology Students: Preliminary results. Engineering Design Graphics Journal. 75(2), 2-8.
9. Veurink, E.L., Hamlin, A.J., Kampe, J.C., Sorby, S.A., Blasko, D.G., et.al., (2009). Enhancing Visualization Skills-Improving Options and Success (EnViSIONS) of Engineering and Technology Students. Engineering Design Graphics Journal. 73(2), 1-17. 10. Charyton, C., Jagacinski, R.J., Merrill, J.A., Clifton, W., and DeDios, S., (2011). Assessing Creativity Specific to Engineering with the Revised Creative Engineering Design Assessment. Journal of Engineering Education. 100(4), 778-799. 11. Sheppard, S.D., Macatangay, K., Colby, A. and Sullivan, W.M., (2009). Educating Engineers, Designing for the Future of the Field. The Carnegie Foundation for the Advancement of Teaching, Princeton, NJ., Jossey-Bass Publishers. 12. Bodner, G.M. and Guay, R.B., (1997) The Purdue Visualization of Rotations Test. The Chemical Educator. 2(4), 1-17. 13. Kovac, R.J., (1989). The Validation of Selected Spatial Ability Tests Via Correlational Assessment and Analysis of User-Processing Strategy. Educational Research Quarterly. 13(2), 26-34. 14. Branoff, T.J., (2000). Spatial Visualization Measurement: A Modification of the Purdue Spatial Visualization Test – Visualization of Rotations. Engineering Design Graphics Journal. 64(2), 14-22. 15. Giesecke, F.E., Mitchell, A., Spencer, H., Hill, I., Dygon, J., Novak, J. and Lockhart, S. (2012). Technical Drawing with Engineering Graphics, 14th Edition. Pearson/Prentice Hall, Columbus, Ohio. 16. Veurink. N. and Sorby, S.A., (2012). Comparison of Spatial Skills of Students Entering Different Engineering Majors. Engineering Design Graphics Journal. 76(3), 49-54. 17. Cohen, J., (1988). Statistical Power Analysis for the Behavioral Sciences, 2nd Edition. Lawrence Earlbaum Associates, Hillsdale, NJ.
Paper ID #9738
Enhance Creative Thinking by Collaborating with Designers
Dr. Yingjie Victor Chen, Purdue University, West Lafayette
Dr. Yingjie Chen is an assistant professor in the department of computer graphics technology of Pur-due university. He received his Ph.D. degree in the areas of human-computer interaction, informationvisualization, and visual analytics from the School of Interactive Arts and Technology at Simon FraserUniversity (SFU) in Canada. He earned Bachelor degree of Engineering from the Tsinghua University(China), and a Master of Science degree in Information Technology from SFU. His research covers inter-disciplinary domains of Information Visualization, Visual Analytics, Digital Media, and Human ComputerInteraction. He seeks to design, model, and construct new forms of interaction in visualization and systemdesign, by which the system can minimize its influence on design and analysis, and become a true freeextension of human’s brain and hand.
Dr. Zhenyu Cheryl Qian, Purdue University
c©American Society for Engineering Education, 2014
Enhance Creative Thinking by Collaborating with Designers
Abstract
This paper reports a case study of training technology graduate students to think more
creatively in a visual analytics system-design competition by collaborating with designers. In the
summer of 2013, two faculty members led a team organized by technology and design graduate
students to work on a visual analytics system design and won the only two awards of
“Outstanding Creative Design.” To investigate and learn from this interdisciplinary collaboration
experience, we purposely tracked and collected the design process information such as meeting
minutes, white-board discussion photos, and development files at different stages from the very
beginning of the collaboration for analysis and review. The paper reviews this 7-week design
process and specifically focuses on observing how the technology students were influenced by
their design peers, started to learn and adopt design methods, then accepted and generated “wild”
design ideas by themselves. Furthermore, we also discuss and report faculty’s roles in this
process and the types of strategies that drove the collaboration and fostered the creativity.
Introduction
In the book of “a whole new mind why right-brainers will rule the future,” Daniel Pink [1]
claims that just as information workers surpassed physical laborers in economic importance, the
workplace terrain is again changing, and power will inevitably shift to people who possess
different kinds of minds, such as artists, inventors, and storytellers – creative and holistic “right-
brain” thinkers. This argument sounds a little overbearing; however, it is very true that
humanity’s future relies on the creative mind. As educators, we look forward to inspiring,
motivating, and fostering students’ creativity. Most technology and engineering students tend to
quickly focus on the technical aspects of a project, and design students tend to focus on usability,
quality, innovation, and the aesthetics of products [2]. In this paper, we discuss our collaborative
practice to integrate these two tendency directions and inspire creativity in the practice.
As a category of the IEEE VIS conferences, the VAST (Visual Analytics Science and
Technology) challenges aim to push the forefront of visual analytics tools using benchmark data
sets and establish a forum to advance visual analytics evaluation methods [3]. By participating in
the VAST challenges, researchers are expected to gain understanding of how their system would
be used in dealing with real data analytic tasks. The 2013 VAST challenge presented three
typical challenges problems [4]. The mini-challenge 2 (MC2) was a design-focused problem that
asked participants to design an innovative large display to support situation awareness in a large
computer network control center [5] . Participants of this task are expected to act not only as
problem-solvers, but also as innovative designers who can change the boring work environment
in the network control room. In the summer of 2013, we led an interdisciplinary team of
technology and design graduate students in the Purdue University to work on this competition.
The technology students are from the department of Computer Graphics Technology (CGT). The
design students are from Interaction Design (IXD) of the department of Art and Design (A&D).
Overall, the collaborations were very successful: the two totally different designs created by the
same group won the only two awards. A variety of interesting moments occurred in the process.
To understand and evaluate the collaboration, and improve the strategies for future education, we
collected the process information from meeting minutes, white-board photos, and design
sketches, and we interviewed students during the design process. This paper reviews and
analyzes the collected data, reports this interdisciplinary collaboration process, and suggests a
pedagogical approach to enhance the creativity development of technology students.
Theoretical Foundations of the Study
Creativity and Creative Thinking
Among the varied definitions of creativity [6]–[8], there is a common agreement on what
creativity involves: “bringing something into being that is original (new, unusual, novel,
unexpected) and also valuable (useful, good, adaptive, appropriate) [9].” Researchers have all
agreed that to be creative, creative thinking must take place. To be of value, that thinking needs
to be critical. Creative thinking should integrate fundamental aspects associated with thinking in
general, which combines recalling and imaging; classifying and generalizing; comparing and
evaluating; and analyzing and synthesizing, deducting, and inferring [10]. The role of design and
creativity is well established in art and design domains. However, in computing technology
disciplines, specifically in the development of software systems and information technology, the
computing educational community is struggling to include creativity and design in its teaching
and research [11].
Adopt Creative Thinking from Designers
Today’s design strongly seeks ways to change itself into a more competitive and innovative
discipline, taking advantage of the emerging advanced technologies as well as their profound
effects on emerging design theories, methods, and technologies. Several reform programs have
been initiated by research institutes, universities, and design practices. The Interaction Design
program at Purdue University, which focuses on developing new approaches to explore the
interaction possibilities in the context of industrial design, is one of them [12].
In the literature of Engineering Education, employing Industrial Design (ID) collaborators to
enhance design thinking and creation has been studied for more than three decades [13], [14].
Studies have been conducted on the educational perspective [15], [16] and design methodology
[17]. Esko Kurvinen [18] outlined critical settings and situations that should be taken into
account when industrial design is introduced to engineering-oriented product development. In
biomedical engineering, Jay Goldberg [19] argued that although engineers and designers tend to
emphasize different aspects of design, both disciplines place heavy emphases on identification of
customer needs, manufacturing methods, and prototyping. Several recent research efforts [15],
[16], [20] focused on analyzing the collaboration between Industrial Design students and
Engineering students in various projects. In recent years, researchers interested in studying
problems involving complex interactions on human-machine systems have risen. Numerous
researchers highlighted the importance of user interface features in design because the design
will help users to predict what will happen toward the system [21] [22]. Although many studies
have been made on Interaction Design and complex technology products, few studies highlight
emerging themes in Interaction Design, such as the role of methods and theories, interaction
design processes, and design criteria [23].
Design Collaboration Models
Many design process models show a cycle that repeats itself and is populated by several
design strategies [24]. The illustration in Fig.1 shows the most typical design strategy. A
standard starting point is to read and understand the specifications and constraints of the design
challenge, followed by researching the idea and brainstorming for possible solutions. Ideas then
get prioritized, and the best are selected to be built. The prototype then undergoes an evaluation
based on the product specification checklist. Designers experience the cyclical nature of their
work when their designs cycle through many iterations before the final version is completed.
First they develop one idea, build it to try it out, notice changes that need to be made, make those
changes, and evaluate the new product. The cycle then repeats itself. The act of repeating these
steps is an iteration in the product design cycles.
Fig.1. Cyclic Design Model
Based on this typical design process model, we want to explore how technology students can
effectively learn from and collaborate with design students. Mattessich and Monsey’s survey in
collaboration literature [25] has drawn a clear distinction among cooperation, coordination, and
collaboration. Cooperation is the informal relationship without a clearly defined common
mission, structure, or effort. Coordination shares the understanding of compatible missions, but
authority still rests within the individual organization. Collaboration suggests a more durable
and pervasive relationship, and the authority is determined by the collaborative structure. We
aim to establish a true collaborative relationship in this design competition task. To judge the
collaboration type of design, identifying its mission, authority, and relationship is important.
Kvan [26] suggested that collaboration is also episodic and cyclical. Collaborators interact
periodically, but they work independently and parallel during portions of the design. Kvan’s
model is demonstrated by Fig. 2. There are generally four stages in an iterative cycle: meta-
planning, negotiation, expert work, and evaluation.
Find/Revise Problem
Research
Brainstorming
Prioritize/Select Items
Build/Prototyping
Evaluation
Fig. 2. Kvan’s model of Design Collaboration [26]
Conceptual Justification of the VAST Competition Project
The design problem developed by the 2013 VAST mini-challenge 2 is how to help the
computer network administrators to obtain accurate knowledge about a large computer network
and analyze issues in an efficient manner [5]. This challenge was looking for design talents than
can task risks and envision creative solutions [4]. We sliced this problem into three layers: (1)
visualize all the dimensions of the real-world security network; (2) help the analysts to quickly
be aware of emerging issues; and (3) provide an effective but not disturbing environment in a
control room. To meet these requirements, we look into literature in three areas: big data Visual
Analytics system design, situational awareness, and ambient information display.
To represent, communicate, and analyze complex information, previous work in Visual
Analytics (VA) has explored a variety of approaches to handle ‘big data’ [27]. As ‘big data’ are
often complex, multidimensional, and multivariate, many VA works have discussed different
methods of visualizing large datasets. Pixel-based visualization in different forms are popular to
fit the huge data space into a limited screen space [28]. Another popular method of visualizing
network data is to use graph-oriented visualizations where machines are mapped to nodes, and
links connecting those nodes with different characteristics, such as thickness and color, represent
relations among nodes [29].
Situation awareness (SA) refers to the perception of elements in the environment within a
volume of time and space, the comprehension of their meaning, and the projection of their status
in the near future [30]. The importance of SA as a foundation of decision making and
performance spans many fields, such as emergency medical call-outs, search and rescue, forestry
service, air traffic controllers, driving, power plant operations, maintenance, and military
operations. In a review by Livnat et al. [31], a growing body of research aims to validate the role
of visualization as a means to solve SA issues in complex data problems.
In this challenge task, the analysts should become aware of issues and their severities as soon
as these issues emerge, but they still should spend most of their time on investigating and solving
identified issues. Therefore the display should not be disturbing when not needed. Ambient
Information Displays provide an alternate method of displaying information that does not require
the constant attention of the user [32]. This method has been widely used to encourage and
facilitate communication. Researchers have developed systems that use a multitude of everyday
items to display information such as lights, sounds, shadows, and artificial flowers. Pousman and
Stasko [33] proposed a taxonomy of ambient information systems as a new information
visualization subdomain that complements the focus on analytic tasks, and also provides
analytics, awareness, social, and reflective sights.
A Collaborative Design Process
The mini-challenge of “Situation Awareness Display Design” started in the beginning of
May and its submission deadline was on July 8, 2013. We organized a team with two Computer
Graphics Technology (CGT) students, three Interaction Design (IXD) students and two faculties.
The seven team members started to actively work on these challenges from the middle of May.
The two faculties, one come from CGT, one come from IXD, served in multiple roles –
supervisor, teacher, collaborator, and researcher to study the collaboration process. The whole
collaborative design and development period lasted for about seven weeks. To effectively
integrate design thinking into the process, we as instructors purposely introduced design methods
and models to the collaborative process. Although most of the students are unfamiliar with the
cybersecurity and visual analytics techniques, we encouraged them to speak out and suggest wild
ideas bravely. We recorded and collected all the meeting minutes, white-board notes,
brainstorming sketches, and related documents to review this collaboration. In the following text,
we integrate the traditional cyclic design model (Fig.1) and Kvan’s model of design
collaboration (Fig. 2) to describe and analyze this 7-week collaboration process.
Week 1. Meta-planning Stage: Design Research to Identify Problem
The goal of the design project is to create a large display in an operation control room that
can monitor an enterprise network consisting of several hundred thousand computers. To
motivate the creativity, the organizers listed no particular requirement. The description specifies
only three network features (health, security, and performance) and four types of conditions
(normal activity, routine issues, non-routine issues, and crises). It is hard to find descriptions of
the setup and workflow of a computer network control room. The problem described is close to
the cybersecurity dataset provided in VAST 2012 MC1 [34]. The two faculties have accumulated
some successful experience while competing in that challenge [35]. But to avoid influencing
students’ creative thinking, the faculties only introduces the VAST 2012 data and solutions
briefly, and encourages the students to explore the possibilities by themselves. Design students
were used to conducting ethnography studies to inspect the context and investigate the problem.
During the very first week while everybody was still puzzled by the challenge description, one
IXD student self-initiated to interview a network security analyst.
The interviewee described their main responsibilities, expertise, and the work flow in the
control room. Two students coded this interview audio recording. Based on the interview coding,
team members started to understand how control rooms function to monitor a network for certain
conditions and how to avoid degraded service by controlling it in real time. Generally, the
functions analysts are monitoring status boards, route information, internal networks, external
networks, and social media in real time. Given the urgent nature of work, the organizational
structure tends to be very hierarchical. This hierarchical nature applies to both the personnel and
the procedure. Employees usually have been distinguished and utilized based on their types and
levels of expertise. One level of analysts will be assigned to solve a certain type of problem. At
the same time, management is strongly centralized and directive via formal lines of
communication. As a result, a central information display should be comprehensive and easily
legible, which should facilitate connectivity across organizational connectivity.
Based on these interview outcomes, our design problem has been defined more clearly. We
aimed to develop a large and multidimensional network display that monitors the four types of
conditions of three features in real time and will help the managers to be aware of urgent
conditions and assign tasks to suitable personnel immediately.
Weeks 1 ~ 2. Meta-planning Stage: Brainstorm with Sketching
Based on the identified problems, we encourage students (both IXD and CGT) to sketch their
ideas freely without constraints. Designers think visually. Visual thinking involves the
interaction between mental (imaging), graphical (drawing), and perceptual (seeing) images [36].
Although cognitive psychologists have mixed views of relations between mental imagery’s
nature and representation, it has been examined and confirmed that drawing is valuable for
creating new problem-solving ideas [37].
For CGT students, they are not used to representing ideas through sketching. In Fig. 3, they
either drew ideas carefully on the grid paper or made annotations on existing images. They
tended to limit everything into a reasonable organized arrangements and start the brainstorming
from a constrained perspective of designing system interface. In the middle sketch, the student
tried to sketch the steps of interaction with a couple of smaller windows and directional arrows.
Fig. 3. Idea Sketches by Computer Graphics Technology Students
IXD students are accustomed to communicating ideas with sketches, but they are unfamiliar
with system design and visual analytics concepts. They felt free to draw their ideas on any type
of paper (the middle image of Fig. 4 was on a kind of cream-colored cardboard), flexibly
occupying the paper corners. They also inserted notes and small figures freely to explain their
ideas of how analysts interact with the visualizations (right sketch in Fig. 4). There are usually
several components of one idea’s representation. Any useful piece would be added if necessary.
Fig. 4. Design Sketches by Interaction Design Students
Weeks 3 ~ 4. Negotiation Stage: Prioritization and Brainstorm
After the first round of idea sketching, the students quickly generated more than thirty
sketches. In week 3, we began to review, compare, discuss, and filter ideas. The challenge task is
to design a large visualization that would grant analysts awareness of the current condition and
help them to understand its effects in light of their pertinent goals. Ambient displays seek to
convey a continuous feed of live information subtly in the background, without alerts to
unnecessary effects. In our search for delicately complex designs we looked into the nature and
wanted to bring some scenes that can be soothing for those who have to face the monotonous and
detailed data on a daily basis.
Fig. 5. Discuss, Filter, and Integrate Ideas (photos of white boards)
Based on this overall goal, the team listed all the ideas on the white boards to discuss and
filter them. The left image of Fig. 5 shows the results of the first round of negotiation. Some of
the ideas were ticked and grouped into either the artistic approach or the scientific approach.
Students did another round of sketching after the 3rd week’s group meeting. After looking at the
design student sketches, CGT students were obviously relaxed and started to sketch in a more
flexible way in the second round. In week 4, we compared and selected a couple of ideas as
shown by right image of Fig. 5. Generally, there were two main directions of these visual design
ideas: circular-based visualizations that can show the hierarchical structure of network, and line-
based visualizations that display most of the network properties. The team members were also
excited by the waterfall scene idea of artistic approach and the galaxy scene idea of scientific
approach. We found it difficult to give up either of these directions or ideas. So we decided to
make two submissions to the challenge: integrate the circular visualization with the galaxy idea
and the line visualization with the waterfall idea. The group of five students split into two teams
and each team has at least one CGT student and one IXD student. Two faculties were
participating in the design and development of both teams.
Weeks 5 ~ 6. Expert Work Stage: Detailed Design and Prototyping
During the expert work stage, both teams focused on the visualization design details and
building interactive prototypes to demonstrate the design. For example, the circular visualization
SolarWheels was inspired by the Solar System and the wheels commonly seen in our daily life.
We applied the metaphors of orbiting planets, solar coronas, planetary hierarchies, along with the
circular shape and spoke element of wheels. Together the display, as a collection of modified
form of Solar Systems, is able to interactively visualize a global computer network. In the
detailed design, the team members discussed how different dimensions of data can be visualized
by properties in the circular graph (right image in Fig. 6). The designers illustrated the decisions
in the figures (left image in Fig. 6) and the technology students created interactive prototypes to
demonstrate how circles split, connect, scale, and integrate. The two components were under
construction in parallel. We used different kinds of communication tools (e.g. Dropbox, Skype,
and WeChat) to discuss and update progresses on a daily basis.
Fig. 6. Process of Designing the Display of SolarWheels.
Weeks 5-7. Evaluation Stage: Review and Packaging
During the final stage, the two teams evaluated each other’s projects in the big group weekly
meetings. This type of peer evaluations turned out to be very effective and productive. Since
these two ideas are concentrated from early ideas contributed by all the team, their names should
be listed in both submission entries but varies in sequence. As a result, all students were
motivated to carefully inspect problems and tried to contribute innovative and practical ideas to
both entries. The relationship between two teams are more like collaborators than competitors.
For example, suggested by other team members, the circular team added a corona-like ring to
display the network speed. The line team changed the metaphor from a pouring waterfall to a
scene of calming rain (Fig. 7). Many changes like that were made, based on the peer critiques
and suggestions in the big group meetings.
We submitted two entries to the 2013 VAST mini-Challenge 2: SolarWheels and SpringRain.
Each of these submissions had to include a digital system prototype, a summary Web page, and a
high-resolution 4-minute video to demonstrate the design. During the final two days, all students
and faculties worked intensively in the same lab. Everybody was busy working on the parts in
which they are most expertized. For example, the student with a good voice was responsible for
all the video narrations, an experienced movie editor worked with a novice on both videos. The
professors drafted the submission summaries, solved key technical problems, and inspected all
the submission components.
Fig. 7. Two Challenge Entries: SolarWheels and SpringRain
Discussions
Our two submissions won the only two awards of 2013 VAST mini-challenge 2 while
competing with teams from many prestigious research centers. Both of these awards were
entitled “Outstanding Creative Design” and received many compliments from the reviewers and
from the conference presentations. Students were very excited and motivated by the international
reviewing committee and thus built up confidence to explore more in the domain. While looking
over this whole design experience, as the supervisor and educator of the team, we were interested
in investigating how creative thinking was seeded, nourished, and flourished. The collaboration
of technique and design was definitely a winning start. However, we have had experience with
projects having such interdisciplinary collaboration that were not very successful [2]. To
understand more about this process, we not only analyzed the process data collected, but also
reviewed the entire process with all of our student participants after conference presentations.
Based on a discussion by Amoussou et al. about the fundamental aspects of creative thinking
[10] and our team review discussion, we identified a few related design strategies to be referred
by technology students:
Define criteria and constraints for ill-defined design problems. For technology students,
they used to solve well-defined problems with professional skills: the problem has a
clearly defined given state, a finite set of operators or rules to apply on a given state, and a
clear goal state. Many times only a limited number of optimistic solutions are available.
Facing open-ended design problems like this challenge, they are easy to get confused and
lost because of so many viable solutions, or they will focus deeply on their first idea and
spend all their time to make that one perfect. Adopting some ethnographical research
methods to investigate, analyze, and define the constraints are very crucial for solving the
design problem in a creative way.
Use drawings to imagine, represent, recall, and communicate ideas. Sketching is a tool,
but not a task. A creator should allow the mind to drive the pencil and escape the limits of
paper size to record emerging ideas, using free-style drawings to recall and discuss ideas.
Even though the sketch may not be professional and aesthetically pleasing, the wild ideas
behind it are invaluable and traceable.
Classify, deduct, infer, and synthesize the idea pool to elaborate. For a design problem,
the number of solutions is basically infinite. The strategy to find a good solution is not to
select an idea, but to group, compare, filter, integrate, and even reconstruct several ideas.
While facing a large group of choices, analyzing the pool to identify the best and creative
solution requires collaborative intelligence and efforts.
Compare and evaluate to enhance the solution. Based on the circular design model,
evaluation is always the decisive stage to review and detect flaws. We luckily had two
teams working in parallel to solve the same problem with different approaches and were
able to evaluate and contribute in an effective way. Using a proper evaluation method and
recruiting the suitable evaluators are very determinative for the success of this project.
As educators, participants, and observers of this interdisciplinary collaboration experience,
we continuously encouraged technology students to try and learn “wild” design methods and
metaphors from their design major peers. The goal of this collaboration is not to receive creative
ideas from designers like many previous approaches [12] [13], but learn to be as creative as
designers. At the same time, our students also taught us a great deal and motivated us to explore
more in this journey of searching for creativity in education.
Acknowledgment
Our participation in the 2013 VAST Challenge was partially funded by the College of
Technology, Purdue University.
Bibliography [1] D. H. Pink, A Whole New Mind Why Right-brainers will Rule the Future. Princeton, N.J.: Riverhead Trade,
2006.
[2] Z. C. Qian, C. Zoltowski, and W. C. Oakes, “Collaborating Interaction Design in Engineering Projects in
Community Service(EPICS),” in Soaring to New Heights in Engineering Education, Seattle, WA, 2012, pp.
1141–1146.
[3] J. Scholtz, M. A. Whiting, C. Plaisant, and G. Grinstein, “A reflection on seven years of the VAST
challenge,” in Proceedings of the 2012 BELIV Workshop: Beyond Time and Errors-Novel Evaluation
Methods for Visualization, 2012, p. 13.
[4] M. Whiting, K. A. Cook, C. L. Paul, G. Grinstein, B. Nebesh, K. Liggett, M. Cooper, and J. Fallon, “VAST
Challenge 2013: Situation Awareness and Prospective Analysis,” in Visual Analytics Science and Technology
(VAST), 2013 IEEE Conference on, Atlanta, GA, 2013.
[5] VAST, “VAST Challenge 2013: Mini-Challenge 2,” 2013. [Online]. Available:
http://vacommunity.org/VAST+Challenge+2013%3A+Mini-Challenge+2. [Accessed: 13-Sep-2013].
[6] E. P. Torrance, “The nature of creativity as manifest in its testing,” Nat. Creat., pp. 43–75, 1988.
[7] R. J. Sternberg, E. L. Grigorenko, and J. L. Singer, Creativity: From potential to realization. American
Psychological Association, 2004.
[8] M. Csikszentmihalyi, Creativity: Flow and the Psychology of Discovery and Invention. New York:
HarperPerennial, 1996.
[9] R. Ochse, Before the gates of excellence: The determinants of creative genius. CUP Archive, 1990.
[10] C. Kamii, “Teaching for thinking and creativity: a Piagetian point of view,” Lawson AE Ed, pp. 29–58, 1980.
[11] G.-A. Amoussou, E. Cashman, and S. Steinberg, “Ways to Learn and Teach Creativity and Design in
Computing Science,” in Proceedings of the 2007 Symposium on Science of Design, New York, NY, USA,
2007, pp. 12–13.
[12] Z. C. Qian, S. Visser, and Y. V. Chen, “Integrating user experience research into industrial design education:
interaction design program in Purdue,” in Open Catalyzing Innovation, Washington, D.C., 2011.
[13] C. Lorenz and T. J. Peters, The Design Dimension: the new competitive weapon for product strategy and
global marketing. Blackwell, 1990.
[14] M. Yamamoto and D. R. Lambert, “The impact of product aesthetics on the evaluation of industrial products,”
J. Prod. Innov. Manag., vol. 11, no. 4, pp. 309–324, 1994.
[15] N. Hoekstra and J. Morris, “Solving industry design problems using engineering technology and industrial
design student teams,” in 39th IEEE Frontiers in Education Conference, 2009. FIE ’09, 2009, pp. 1–6.
[16] M. L. Martinez, G. Romero, J. J. Marquez, and J. M. Perez, “Integrating teams in multidisciplinary project
based learning in Mechanical Engineering,” in 2010 IEEE Education Engineering (EDUCON), 2010, pp.
709–715.
[17] R. M. Goff, M. R. Vernon, W. R. Green, and C. R. Vorster, “Using design - build projects to promote
interdisciplinary design,” in Frontiers in Education, 2004. FIE 2004. 34th Annual, 2004, pp. S3C– 27–32 Vol.
3.
[18] E. Kurvinen, “How industrial design interacts with technology – a case study on design of a stone crusher,”
vol. 16, pp. 373–383, 2005.
[19] J. Goldberg, “Senior design - biomedical engineering/industrial design collaboration in senior design
projects,” IEEE Eng. Med. Biol. Mag., vol. 26, no. 3, pp. 75–76, Jun. 2007.
[20] K. Sugiyama and H. Osada, “Importance of Industrial Design and its collaboration process with Engineering
Design,” in Technology Management in the Energy Smart World (PICMET), 2011 Proceedings of PICMET
’11:, 2011, pp. 1–6.
[21] D. Benyon, P. Turner, and S. Turner, Designing Interactive Systems: People, Activities, Contexts,
Technologies. Harlow, Eng.: Pearson Education, 2005.
[22] B. Schneiderman and C. Plaisant, Designing the user interface. Addison-Wesley Longman., 1998.
[23] D. I. Rosli, R. A. Alias, and A. A. Rahman, “Interaction Design Issues: A literature review,” in User Science
and Engineering (i-USEr), 2010 International Conference on, 2010, pp. 133 –138.
[24] B. Lawson, How designers think: the design process demystified. Routledge, 2006.
[25] P. W. Mattessich, M. Murray-Close, B. R. Monsey, and A. H. W. Foundation, Collaboration: What makes it
work. Amherst H. Wilder Foundation St. Paul, MN, 2001.
[26] T. Kvan, “Collaborative design: what is it?,” Autom. Constr., vol. 9, no. 4, pp. 409–415, 2000.
[27] J. J. Thomas and K. A. Cook, Illuminating the Path: The Research and Development Agenda for Visual
Analytics. Los Alamitos, CA, United States(US).; Pacific Northwest National Laboratory (PNNL), Richland,
WA (US): IEEE Computer Society, 2005.
[28] D. A. Keim, “Designing pixel-oriented visualization techniques: Theory and applications,” Vis. Comput.
Graph. IEEE Trans. On, vol. 6, no. 1, pp. 59–78, 2000.
[29] I. Herman, G. Melançon, and M. Marshall, “Graph Visualization and Navigation in Information Visualization:
a Survey.,” IEEE Trans. Vis. Comput. Graph., vol. 6, no. 1, pp. 24–43, 2000.
[30] M. R. Endsley, “Measurement of situation awareness in dynamic systems,” Hum. Factors J. Hum. Factors
Ergon. Soc., vol. 37, no. 1, pp. 65–84, 1995.
[31] Y. Livnat, J. Agutter, S. Moon, and S. Foresti, “Visual correlation for situational awareness,” in Information
Visualization, 2005. INFOVIS 2005. IEEE Symposium on, 2005, pp. 95–102.
[32] C. Wisneski, H. Ishii, A. Dahley, M. Gorbet, S. Brave, B. Ullmer, and P. Yarin, “Ambient displays: Turning
architectural space into an interface between people and digital information,” in Cooperative buildings:
Integrating information, organization, and architecture, Springer, 1998, pp. 22–32.
[33] Z. Pousman and J. Stasko, “A taxonomy of ambient information systems: four patterns of design,” in
Proceedings of the working conference on Advanced visual interfaces, 2006, pp. 67–74.
[34] K. A. Cook, G. Grinstein, M. Whiting, M. Cooper, M. Havig, K. Liggett, B. Nebesh, and C. L. Paul, “VAST
Challenge 2012: Visual Analytics for Big Data,” in IEEE Conference on Visual Analytics Science and
Technology, Seattle, WA, 2012, pp. 151–155.
[35] V. Y. Chen, A. M. Razip, S. Ko, Z. C. Qian, and D. S. Ebert, “Multi-aspect Visual Analytics on Large-scale
High-dimensional Cyber Security Data,” Inf. Vis., vol. In Press, no. Special Issue on VAST challenges, p. 10
pages, May 2013.
[36] R. H. McKim, Experiences in visual thinking. Monterey, Calif.: Brooks/Cole Pub. Co., 1972.
[37] F. L. McKoy, N. Vargas-Hernández, J. D. Summers, and J. J. Shah, “Influence of design representation on
effectiveness of idea generation,” in ASME Design Theory and Methodology conference, Paper No. DTM-
21685, 2001, pp. 10–3.