Mable B. Kinzie
Associate Professor, Instructional Technology, University of Virginia
Valerie A. Larsen
Graduate Research Assistant, Instructional Technology, University of Virginia
Joseph B. Burch
Computer Systems Engineer, Information Technology & Communications, University of Virginia
Steven M. Boker
Graduate Research Assistant, Center for Developmental & Health Research Methodology, Department of Psychology, University of Virginia
To go to "Net-Frog,"
2.0 Purpose of The Interactive Frog Dissection
3.0 The Importance of the WWW for Distribution and Use of Educational Materials
4.0 Design of Net-Frog
5.0 Use of Net-Frog: Quantitative and Qualitative Data
6.0 Evolution of Net-Frog and Educational Use of the Internet
In addition, controversy has emerged surrounding environmental issues (specimens are collected in the wild), economic issues (specimens are expensive), and educational issues (students are not well-prepared to learn from dissection). (See  for a more complete discussion of these issues.) While some science educators and scholars advocate replacing the traditional dissection lab with the study of live organisms [1,3], others, such as Berman , Hoskins , and Igelsrud [6,7], express support for dissection. Their support, however, is conditional upon the use of great care and planning in the design of lessons using animal specimens.
Currently available alternatives include books, charts, computer programs, models, filmstrips, slides, transparencies, videotapes, and videodiscs. Many of these, however, can be faulted for their lack of realism and opportunities for student involvement. We initially developed the videodisc-based Interactive Frog Dissection to provide both realistic imagery and opportunities for student practice on frog dissection and anatomical identification. We also hoped the program would serve as both a substitute for laboratory dissection, where necessary, and as a supplement that would better prepare students for the dissection experience . Versions of the program were created for the PC [8, 9] and the Macintosh computer .
Our research with these videodisc-based materials suggests that students using them instead of dissecting will perform as well as students who dissect, when tested on knowledge of frog anatomy and on dissection procedures [2, 8, 11]. Further, we found that students who used the materials prior to conducting a dissection performed the subsequent dissection more effectively than students receiving no preparation and more effectively than students viewing a videotape as preparation . Students who dissected after using the materials as preparation also learned more about frog anatomy and dissection procedures than those who dissected without preparation .
Despite these positive findings, we experienced difficulties getting the videodisc-based materials into classroom use. While educators were enthusiastic about using them, we ran into problems on two fronts: hardware and software. In order to use the materials, teachers had to have access not only to a computer, but to a videodisc player and video monitor, at the least. If they were to use the PC materials, the difficulties increased because specific video overlay and videodisc controller cards were required, along with a touch screen. Software availability was also an issue. The materials had been developed with internal research funds. Once our initial pressing of several hundred videodiscs had been given away, there were no funds for additional pressings. Being a university school of education, we lacked the mechanisms for bankrolling, marketing and distributing educational software.
As a research university, we are no longer limited by non-existent budgets for multiple platform development, and by our lack of software marketing and distribution functions. We can focus on what we do best: instructional research and development. Barriers for users, such as possession of a videodisc or access to specific hardware components, have been removed. In the first eight months since its Internet release in August, 1994, Net-Frog has been used more each week (an average of 1,943 visits/week) than the videodisc-based program had been used in its previous five years of existence. Note: This is a conservative estimate. For information on how this and other data were determined, see section 5.0.
Net-Frog provides a laboratory dissection experience on-line (sans the smell). Both preserved and pithed specimens are depicted with 60 in-line color images to highlight the visual similarities and differences in the frog anatomy. QuickTime movies (n = 17) are used to demonstrate dissection techniques, and provide information unavailable from still photographs, such as how to hold the skin with forceps when making incisions, or how the lung inflates. Interactive practice involves users in the experience, asking them to identify critical locations for various dissection procedures and to find various internal organs. Feedback is provided, and the user can always review before attempting a practice activity. In this way, Net-Frog goes beyond just providing information, as illustrated in these user comments:
We worked hard to make Net-Frog easy to use, through clear menu design and provision of navigational options. The photographic images and text are automatically presented. Users are allowed to choose whether or not to download the QuickTime movies or to engage in interactive practice activities. Common user responses include comments such as:
Since the release of "Net-Frog", the instructional worth of the materials has been suggested in comments submitted by biologists at other institutions:
Next, a program was written in the Splus statistical language to read the qualified data files. Table 1 contains our summary statistics.
Table 1: Net-Frog Summary Statistics, 8/4/94 -- 3/31/95
We elected to take a conservative approach to determining number of unique machine visits. (By unique machines, we refer to the number of machines for which the address string [whether, alpha, numeric or alphanumeric] is unique.) We considered all access requests received from a single machine address in a given day to be one machine visit. We recorded 62,201 such visits within the 8 month period.
We found that a number of the clients making these visits were repeat customers. We recorded 35,355 unique machine addresses served, suggesting an average of 1.8 visits each.
A total of 1,145,719 files and over 48.5 billion bytes were served, resulting in an average of 18.42 files and 780,107 bytes served per machine visit.
HTML files consisted of 26.3% of the files served (HTML files make up 37.9% of the total program), while 71.8% of files served were GIF images (GIF images make up 43.4% of total program files). QuickTime movies were 1.6% of the total files served (these movies make up 6% of the total program files).
One facet of the Net-Frog statistics that warranted further exploration was the number of unique network domains represented in the data. Our HTTP server software (NCSA HTTPD 1.3) attempts to resolve IP addresses into more human-readable domain names by polling a name-server at the time of access. IP addresses not having a corresponding domain name remain unresolved at this point. Only about 75% of our data was resolved to domain name, leaving about 25% unresolved.
We were curious as to whether these unresolved IP addresses could be further identified. To explore this possibility, we compared the unresolved IP addresses with entries obtained from a standard network reference ("nets.unl.now" obtained from ftp.merit.edu on April 6, 1995). If a match was found for the network identifier, we extracted its corresponding domain-style identifier and inserted this into the access logs in much the same way that the original server resolution had been accomplished. By this means, we were able to resolve about 95% of the previously unresolved addresses.
This additional data resolution resulted in the identification of 2 additional unique domain networks (previous figures suggested 68 unique domain networks; our resolution procedures resulted in a total of 70 unique domain networks.
Our clients were located primarily in the United States; 75.8% of the requests that could be tracked to a domain were US-related, with 33.6% from U. S. Educational institutions, 25.2% from U. S. Commercial addresses, 8.0% from Network addresses, 4.5% from the U. S. Government, and 2% or less each from the "United States" domain (2.0%), Non-Profit Organizations (1.5%), and the U. S. Military (1.3%).
Globally, requests from the United Kingdom (5.6%), Canada (5.5%), Germany (2.0%), Netherlands (1.7%), Sweden (1.6%), France (1.2%), and Australia (1.1%) were notable. See Table 2 for a summary.
Table 2: Unique Domain Networks Accessing Net-Frog: Major Players
Note: These proportions were computed following the IP address resolution efforts described in the text.
It is interesting to note that the address resolution procedures reported above did not result in substantially altered proportions (as compared to the proportions that had previously been obtained). The most obvious change was in the number of unique clients attributable to addresses in the United States--this figure increased by 2.5%, up from 73.3%. This increase was influenced by the additional resolutions of IP addresses from U. S. Government (1.1%), U. S. Commercial networks (up .8%), and "Network" addresses (up .45%). Slight decreases in foreign access were concomitant and equally small.
While usage in schools is still limited due to lack of direct Internet connectivity in the classroom, we are finding that Net-Frog is an ideal tool for parents and children to use together, and for anyone at all to use who is interested in learning:
We hope that these experiences will continue to encourage innovative and effective instructional practice using the Internet as a powerful tool.
Valerie Larsen is a doctoral student in Instructional Technology at the University of Virginia's Curry School of Education. She is engaged in research and development on effective use of interactive technologies in adult and teacher education. She is editor of Interface, a national newsletter sponsored by IBM which covers topics relating to technology infusion and teacher education.
Joseph Burch is Computer Systems Engineer with the UNIX Systems Group of Information Technology & Communications, University of Virginia. He is an obsessive C and Perl programmer and is actively exploring web technologies, especially statistical techniques.
Steven M. Boker is a Ph.D. Candidate in the Quantitative Area of the Department of Psychology at the University of Virginia where he studies dynamical systems, the perception of time, and neural modeling. He is the author of several commercial software packages and consults on statistics, data management and software interface design.