Dr. Foreman recently received his PhD from the Department of Neurobiology and Anatomy at The University of Utah, School of Medicine (SOM). He is a faculty member with the Divisions of Physical and Occupational Therapy at The University of Utah, College of Health, and teaches gross anatomy and neuroanatomy.
Dr. Musolino is assistant professor in the Division of Physical Therapy and director of Clinical Education, College of Health, at The University of Utah. Her research area is in curriculum design, development and evaluation, with focus areas in self-assessment, service learning, and cultural competence for the interdisciplinary health professions.
Dr. Albertine is professor of pediatrics, medicine, and neurobiology and anatomy at Utah's SOM. He is training director of the Children's Health Research Center and director of the Research Microscopy Facility at the Health Sciences Center.
Use of computer-assisted instructional (CAI) material as an adjunct to teaching human gross anatomy at health sciences schools is increasing (Berube et al., 1999; McNulty et al., 2000). Contributing reasons for the growing use include the visual nature of the topic, the desire by students to have self-study tools outside of the cadaver laboratory, the declining number of qualified gross anatomy teachers, and the declining number of basic science course hours (Cottam, 1999; Drake et al., 2002).
CAI plays a role in the delivery of anatomical material to students. For example, CAI is used to supplement dissection (Guy and Frisby, 1992; Predavec, 2001; Bukowski, 2002), as well as to deliver instructional material (Toth-Cohen, 1995; Barker, 1998; Boucher and Hunter, 1999). An advantage of CAI is the use of digital images to illustrate instructional material, which is preferred by users (Cheng et al., 2003). Furthermore, CAI provides interactivity with the instructional content (Chou, 2003). For the purpose of this study, we defined interactivity as selectable views and content that permits the user to tailor self-study and review. Interactivity includes five dimensions that fulfill communication needs. The five dimensions are playfulness, choice, connectedness, information collection, and reciprocal communication (Ha and James, 1998). The dimension of playfulness incorporates the presence of educational material, which stimulates curiosity. An example would be a question-and-answer format and/or games. Choice is defined as the amount of information users have access to. Connectedness is measured by the presence of information of interest to the user. Information collection provides tracking use. The reciprocal communication dimension provides communication between users and authors, such as by electronic mail (Chou, 2003). However, CAI is not without limitations. One limitation is suboptimal use of the screen area. Examples are use of a fraction of the available screen area or lack of control of the image size by the user. Other examples are when annotations are placed around the margins of an image, requiring the image to be reduced to incorporate both image and annotations in the viewable screen area, or when annotations and lead lines decrease clarity of the image. Lastly, CAI tools often lack interactivity by the user.
We designed a CAI tool that optimized advantages and minimized limitations. We hypothesized that students would recognize the value of a CAI tool in neuroanatomy if the CAI tool was designed for easy navigation and facilitated self-study and review. Therefore, we designed the CAI tool to optimize use of screen area, provide user control of the image size and orientation, use annotations that did not obscure the image, and provide user interactivity with the content. The interactive component incorporated playfulness, choice, and connectedness. Forty-three physical and occupational therapy graduate students enrolled in human neuroanatomy evaluated the CAI tool. Their evaluation addressed navigation, clarity of the images, benefit of the CAI tool for self-study and review, and rating the CAI tool compared to traditional learning tools.
Software and Authoring
Fireworks MX was used to edit and size the images and drawings. Because we had no control over screen area of the computers or the Web browsers that the students used, we determined the optimal image dimensions and resolution as well as image file format so that the downloaded images would nearly fill any viewable screen area. Our goal was to minimize the margins and maintain clarity of the images. Clarity was defined as images that were not obstructed by annotations and would not degrade (pixelate) when zoomed. To meet that definition, the original images were initially obtained in a tagged image file format (TIFF), either as digital photographs or scanned images. However, because of the large file size of TIFF files, we compressed the images into joint photographic experts group (JPEG) file format, at 125 pixels per inch. In addition, the images were sized to a 600 pixel height and the proportionate width. This allowed for fast download time. Because the majority of monitors have a viewing capacity of 72 or 96 pixels per inch, importing the image into the template at 125 pixels per inch retained clarity of the images with magnification (zooming). To accommodate for a variety of screen sizes, we embedded the CAI tool in an HTML file, within which we coded the page to be displayed at 100% height and width. Therefore, the CAI tool nearly filled any available screen, regardless of browser. All of the imported digital images were original, thus further avoiding copyright infringement.
Flash MX was used to design and program the CAI tool. First, we imported the JPEG-formatted images into Flash MX. We imported images of complete brains and brainstems, as well as coronal, horizontal, and sagittal slices of those structures. In addition, images of head surface anatomy, skull anatomy, and cerebral vasculature were added. The Web site had 30 images (to view samples of the Web site, go to http://medstat.med.utah.edu/diganat/anatomy). Buttons and annotations were created using scalable vector graphics (SVG) and positioned over the structures to be identified. In addition, navigational buttons were created for zooming, panning, and rotating the image (Fig. 1). The rationale for creating components of the CAI tool using SVG is that SVG do not lose clarity when enlarged (zoomed in), contrary to pixel images, which degrade when enlarged. Figure legends were composed to provide descriptive text. We coordinated the images with navigation, rollovers, and figure legends using ActionScript scripting language within Flash MX. ActionScript is designed to write scripts that enable interactivity with the CAI tool via the keyboard or mouse. Applications created using Flash MX are compatible with most operating systems (i.e., Windows®, Microsoft; Macintosh®, Apple Computer) and browsers (i.e., Internet Explorer®, Microsoft; Netscape®, Netscape Communications; Safari, Apple Computer; Firefox™, Mozilla). However, a free plug-in (Macromedia Flash Player) is required to view the content. The Macromedia Flash Player is available for download and installation from the Macromedia Web site (http://www.macromedia.com/shockwave/download/download.cgi?P1_Prod_Version=ShockwaveFlash). Dreamweaver MX was used to organize and modify the Web site. Dreamweaver MX software was also used to code the HTML pages, as well as upload and synchronize the Web site to the server.
Thirty-eight physical therapy and 20 occupational therapy students were enrolled in a laboratory course in human neuroanatomy (44 hr of laboratory session; 88 hr of course contact) in the spring of 2003, for a total of 58 students, 35 (60%) males and 23 (40%) females. Fifty-six of the students were in their first year of training in a master's degree program. Two of the students were in their second year. Each student was asked to participate voluntarily in the study. The study was approved by the University of Utah Institutional Review Board (IRB). The need for consent was waived.
On the final day of class, students completed an IRB-approved questionnaire that evaluated navigation, clarity of the images, benefit of the CAI tool, and rating of the CAI tool compared to traditional learning tools. The six evaluation statements are presented in Box 1. The questionnaire used a Likert scale (Portney and Watkins, 2000). Students were also asked to give a brief written description of their response to the CAI tool. Results are shown as mean ± one standard deviation, as well as minimum and maximum values (Microsoft Excel 2000, Microsoft).
From the Web site, the students used the CAI tool to choose among the images. The CAI tool provided nearly full-screen-sized digital images that could be interactively zoomed in to view small structures. Furthermore, the CAI tool enabled the user to pan and rotate (360°) in any direction.
Buttons, created as SVG objects, were placed over structures to be identified (Fig. 1). When a button was rolled over, SVG overlays were opened that highlighted the underlying structure. Simultaneously, related text was loaded into the figure legend (Fig. 1). Rollout from a button removed the highlighted area and removed the related text from the figure legend. The figure legend could be displayed (enabled) or hidden (disabled) by the viewer. This design feature allowed the user to choose an unobstructed view or an obstructed view, respectively. In addition, the buttons and highlighted areas were overlays that were not embedded (flattened) with the digital image. Because we used Flash MX software and ActionScript programming language, the digital image and its overlays moved together in register. Therefore, the highlighted area moved with the image when the image was zoomed, panned, or rotated (Fig. 2).
Forty-three of the 58 students (74%) completed the questionnaire. The mean age of the students who used the CAI tool was 25 years (range, 21–37 years).
Student assessment of the CAI tool was by a questionnaire distributed after the neuroanatomy course ended. The questionnaire was based on a Likert scale of 1 (strongly agree) to 7 (strongly disagree) for five of the six statements. The sixth statement was based on a Likert scale of 1 (much better than traditional tools) to 5 (much worse than traditional tools).
Graphical representation of the questionnaire responses is shown in Figure 3. Eighty-eight percent of the participants determined that the CAI tool was easy to navigate, with a mean score of 1.40 ± 0.82. Eighty-four percent of the participants also determined that the images were clear, with a mean score of 1.72 ± 0.85. Eighty-four percent of the participants determined that the CAI tool provided education in structure identification, with a mean score of 1.65 ± 0.92. Seventy-six percent of the participants determined that the CAI tool was beneficial for self-study and review, with a mean score of 1.81 ± 1.10. Eighty-eight percent of the participants determined that the CAI tool was overall beneficial, with a mean score of 1.49 ± 0.77. Finally, 95% of the participants determined that the CAI tool was much to somewhat better than traditional tools (atlases), with a mean score of 1.56 ± 0.67.
We designed a CAI tool for self-study and review of neuroanatomy by first-year physical therapy and occupational therapy students. The study was designed to subjectively assess navigation, clarity of the images, benefit of the CAI tool, and the rating of the CAI tool compared to traditional learning tools (specifically, neuroanatomy atlases). We used a Likert scale to evaluate the CAI tool. Student evaluation indicated that our Web-based CAI tool was easy to navigate, the images were clear, the CAI tool facilitated self-study and review, and the CAI tool was rated higher than traditional learning tools.
We used software designed by Macromedia (Fireworks MX, Flash MX, and Dreamweaver MX) to compose a CAI tool for interactive use via the Internet/intranet. We used commercial versions of the software because Macromedia software products licensing agreement permits licensing of original compositions when the commercial version of their software programs is purchased (http://www.macromedia.com/software/eula/tools/). This broader use is not permitted by the licensing agreement for educational versions of Macromedia software products.
The design of the CAI tool allowed the user to choose among 30 digital images. Each image could be zoomed in/out, panned, and rotated. In addition, SVG buttons were placed over structures of interest. Rollover of an SVG button highlighted the underlying structure. Simultaneously, related text was displayed in the figure legend. The user had the choice to display (enable) or close (disable) the figure legend to control the screen content.
From this study, we learned the importance of evaluating the design of CAI tools. Design is an important factor in the context of learning (Chou, 2003). Design is important because it provides the structure and method to deliver content. Furthermore, we learned the importance of interactivity, which is crucial in acquiring knowledge (Sims, 1997). Interactivity plays an important role because it engages the learner with the educational material. However, evaluation of CAI tool design and interactivity in health science students' education has not been reported. Instead, evaluation has been on outcomes data, such as examination scores (Erkonen et al., 1992; Stanford et al., 1994; Devitt and Palmer, 1999; Garg et al., 1999; Bukowski, 2002; Fleming et al., 2003). Paradoxically, the impact of CAI tools on student performance on tests may be influenced by design features of the CAI tool. For instance, if the design does not provide easy navigation and clear images, then students may not use the CAI tool. In that case, test performance might be expected not to improve.
Design of CAI tools should be structured around a technical framework. The technical framework should incorporate interactions between learner/interface, learner/content, learner/instructor, and learner/learner (Chou, 2003), because without interaction, delivery of instructional content could be discouraged. The CAI tool that we designed incorporated interactions between learner/interface and learner/content. The learner/interface interaction occurred through a single-menu page in which the user could select an image from the image library. The learner/content interaction occurred by providing flexibility for the user to highlight structures of interest and manipulate the image (e.g., zoom, pan, and rotate). Our CAI tool did not include design for learner/learner and learner/instructor interaction, such as electronic mail to other learners or instructors, respectively. Those types of interactions were not included because that was not the goal of our study. Nonetheless, learner/learner and learner/instructor interactivity could be beneficial to encourage communication.
The computer equipment used by the students, and the speed at which they access Web-based CAI tools, also should be assessed when developing CAI tools. Assessment is warranted because access and willingness to use CAI tools can be limited by accessibility. Accessibility issues may be compromised by either the type of computer equipment or the students' network connection. We did not assess these technological issues in the present study. We are in the process of a technological needs assessment to characterize the types of computer equipment, peripheral equipment, and the type, speed, and method of Internet connection used by students.
Limitations of our study may have affected the results. One limitation is the small sample size (n = 43) from a single institution. In addition, learning styles of students were not assessed. Therefore, the applicability of our results to other health professions schools is not known. Because of these two limitations, and given that instructional methods vary among schools, a multicenter study in which all of the participating schools use the same CAI tool would provide more comprehensive results.
Our study also relied on responses from students who completed the questionnaire. We cannot address the assessment of the 15 students who did not complete the questionnaire. Of these students, six indicated that they had insufficient time to use the CAI tool. Seven other students did not provide a reason. Two students indicated that they had difficulty accessing the Internet server. One of those students used a different Web site; the other student declined use because they heard that the CAI tool did not display a comprehensive library of images.
Another limitation of our study is it assessed student subjective perceptions of the CAI tool. Our study did not assess knowledge gain. Assessment of such tools has been performed, primarily characterizing their frequency of use (McNulty et al., 2000). Therefore, further research needs to be conducted to evaluate knowledge gain and retention between students who use CAI tools versus traditional learning tools (e.g., atlases).
Recommendations from students included the desire for higher-resolution images. This recommendation would improve the clarity of the images for zooming in; however, the larger file size would have the undesirable consequence of longer download times. An alternative approach would be to distribute the CAI tool on compact disc (CD) or digital versatile disk (also called digital video disk or DVD) media, in which greater-resolution images and three-dimensional views could be incorporated into the CAI tool. However, that type of distribution would limit the flexibility of Web-based CAI tools because Web-based CAI tools can be easily and repeatedly updated, whereas CD or DVD-based CAI tools cannot. Another recommendation was for addition of more structures to the Web site. Experience taught us to anticipate this criticism and we intend to expand the library of digital images, which is easier to accomplish on a Web-based CAI compared to a CD/DVD-based CAI tool.
In conclusion, our study shows that the design features of our CAI tool accomplished the goals of easy navigation, clear images, and benefit for self-study and review. Because of these design features, the Web-based content was perceived as better than traditional atlases for self-study and review. Thus, our study emphasizes the importance of design to create CAI tools for students in health professions schools. A next step will be to perform learning outcomes research. Such research should focus on objective assessments of learning efficiency and knowledge retention.
Supported in part by grants from the Educational Computer Committee and funds from the Department of Neurobiology and Anatomy, the School of Medicine, and The Division of Physical Therapy, at the University of Utah.